April 20, 2026 • 61 mins
Most organizations in regulated industries aren't slow on AI because of compliance. They're slow because no one has decided to be brave enough to move.
In this episode, Chad sits down with Ondar Tarlow — a marketing executive who led AI adoption inside financial services organizations before most CMOs were willing to have the conversation. Ondar's team deployed propensity modeling and machine learning to identify the next best product for existing customers, drove a 5x improvement in campaign performance, and cut production time by 75% using generative AI tools layered into their creative workflow.
The conversation goes well beyond marketing tactics. Ondar makes the case that AI should never be compartmentalized inside IT — or any single department — and walks through a practical framework for how middle market leaders can get started: map the workflow first, identify high-leverage use cases, and treat AI as an assistant that helps your team do more in less time, not a replacement for the people already doing the work.
You'll also hear Ondar and Chad dig into what it actually looks like to navigate the tension between moving fast and managing compliance risk in a regulated environment — and why the leaders who manage that tension best are the ones willing to align marketing, legal, and risk teams around a shared strategy before they start testing.
Ondar Tarlow is a CMO and consultant with more than 20 years of experience in financial services, motorsports, and lifestyle brands. His hands-on experience with AI adoption in highly regulated environments makes his perspective directly applicable for middle market leaders who are ready to move past the conversation and into action.
Walk away from this episode with a clearer starting point for mapping AI into your existing workflows — and a sharper sense of what's actually holding your organization back.
Watch the full episode: youtu.be/VOyAn_BX3Qw Connect with Ondar: ondartarlow.org
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:03):
I'm Chad Harvey and this is AI for the C-suite, the show for senior leaders who know AImatters and need to figure out what to do about it.
Each episode, we dig into what it all actually means for middle market organizations, howit's changing decisions, strategy, leadership, and the nature of work.
Let's get into it.
Today, I'm joined by Andar Tarlow, a marketing exec with more than 20 years of experiencebuilding brands and financial services, motorsports, and lifestyle industries.
(00:33):
Andar served as chief marketing officer at Connecta Federal Credit Union and PacificPremier Bank, where he led campaigns that generated over $1.7 billion in product volume
and earned him Diamond Awards and a Top 100 Marketer Recognition.
Andar's perspective is especially relevant for our listeners because he didn't just watchAI happen from the sidelines.
(00:56):
He rolled up his sleeves and put it to work.
His team used propensity modeling and machine learning to transform how they targetedcustomers and then layered in generative AI tools to overhaul their creative workflows.
We're talking five times improvement in campaign performance and 75 % reductions inproduction time.
(01:17):
Andar, welcome to AI for the C-suite.
Hey, Chad, how are you?
I am doing really well and I gotta say I am super excited to have you on today because thenatural approach point for a lot of organizations is through marketing and then I also get
a lot of questions from folks about highly regulated industries and you were in that aswell.
(01:38):
So I'm really looking forward to our conversation today.
Good good.
Well every superhero has an origin story so let's talk about your AI origin story maybe.
Take me back to the moment that AI stopped being abstract for you and started beingsomething that you needed to act on.
I'm sure that you saw something in the world of marketing that made you take notice andsay, we really need to lean into this.
(02:05):
So what was that?
What did that look like?
How did your experience unfold there?
Yeah, well, first of all, AI has moved so quickly over the last few years that you'retalking about something a month ago and it could have already advanced tremendously beyond
that and the way you're using it today is totally different.
But going back to the origin question was really working with some very, very smart folksin our business intelligence team who were utilizing AI and ML machine learning for some
(02:38):
time and identifying, how can we take a look
our book of business, existing customers, existing members on the credit union side andfigure out how we can do better for them.
How can we identify what types of products they might need on the financial side, whetherthat be a credit card, whether that be a personal loan, whether that be something on the
(03:01):
wealth management side is to figure out exactly what made the most sense.
And so by being able to dig in and be able to do things like A-B testing between
utilizing the preventively modeling in comparison to general filters and then figuring outyou know what actually was working and why it wasn't working.
So once we started seeing pickup in terms of simple campaigns like email marketing andsimple and simple factors in terms of moving forward like email open rate increasing
(03:33):
dramatically and then conversion conversion rates driving up as well.
We knew there was something to that and so
that we continue to kind of push on the accelerator to figure out exactly where we couldgo from there.
So that was kind of like the aha moment, but obviously everything that was happeningaround us at the time, and obviously continues exponentially now as far as AI, right?
(03:57):
Both on kind of the AI learning and modeling as far as data is concerned, and then also onthe gen AI side, which a lot of teams in marketing started to use a couple of years ago or
even earlier.
ah very immature versions of GEN.AI, whether it's for image creation, content creation, orother.
(04:18):
And now, taking a look at where it is today.
Also being able to figure out exactly when you say AI, what does that mean?
And then figuring out not so much, hey, we're going to go use AI for X, Y, and Z, but morealong the lines of starting off of what is the strategy overall as far as the company is
concerned?
is, how is marketing supporting that strategy?
(04:41):
And then how can we integrate AI into the entire workflow that we have?
think that's really key.
And you had a lot of marketing
including ourselves that started off on the content side, right?
So to help become a little bit more efficient.
So starting off with ideation and then working into things like social media and otherelements that content creation comes into.
(05:04):
But then at the same time, alongside that, and you always have to keep this in mind as faras regulated industry, so I'm glad you brought that up, is the fear factor, right?
So there's the fear factor in terms of, okay, what does AI mean?
What are the issues as far as IP rights are concerned in terms of content?
as well as in terms of images that are used and so we were basically doing a lot ofresearch at the same time and figuring out exactly how we could use it effectively and
(05:34):
then kind of fast forward to today where I'm doing a lot of consulting and utilizing AI ina manner of which I can become more and more efficient for clients and asking the right
questions and not just looking at AI as a search engine which a lot of people
still do, right?
It's like, yeah, no, I just use chat GPT, you know, versus Google.
(05:56):
Well, I mean, you're really not even scratching the surface as far as the power of how youcan use it as really an assistant.
And then of course, keeping in check.
whether you're in a regulated industry or not, always keeping in check with what are thesources of the information.
(06:16):
And so I constantly hammer on that because you'd still see it.
It happens, I would say, less from where I am than it did a year or two ago.
But you still see those hallucinations from time to time as far as GenAI is concerned.
And you need to make sure that you are double checking all of the work.
You're not just copying and pasting.
(06:37):
They really understand the sources.
and then you're utilizing it as a way, as I mentioned, to become more efficient and to beable to do more in a shorter period of time that it has more impact and then be able to
utilize and test it along the way.
I'm very interested in what the conversation looked and sounded like when you initiallyhad it.
(07:00):
And let me just give you some context for where I'm asking this question from.
ah You've.
been there, you've done that.
You have been as Dean Acheson famously wrote in his book title, President Creation, right?
You worked for an organization that leaned into this technology early on.
I can tell you, because I also do a lot of consulting, I've gotten called into six, maybeeight different banks and credit unions over the last couple of years.
(07:25):
And with the exception of some very broad stuff, almost none of them want to move forward,which is something I see a lot of in the middle market right now.
An awful lot of foot dragging.
think 2026 is going to be
wake up year for a lot of people in many ways positive and negative.
Be that as it may.
The highly regulated industries the financial services sector in particular seems veryreluctant to move forward.
(07:49):
So did you have to sell this?
Did you have to pitch it to the leadership team or was it pretty much.
OK, hey, the guys over in B.I.
or the gals did what they needed to do.
They've kind of cracked open the door.
This is aligned with our strategy.
Away we go.
What did the conversation look like?
I'm fascinated by that just because of all the roadblocks I see out there withorganizations.
(08:09):
Yeah, you're absolutely correct.
mean, in the industry, whether it's financial services or others, there's such a varyingdegree in terms of experience.
You have people who are getting close to being advanced.
I'd say it's difficult to really call anyone in the field an expert nowadays, consideringhow quickly it's moving and how you're utilizing it in your day-to-day or strategic work.
(08:33):
But basically, the conversations were to your point, which is you had the BI team that hadopened it
in terms of the analytics end of things, which was great.
But again, that's different than the Gen.ai side.
And what I mean by that is, now we're going to see if we can utilize it for imagery to beable to become more efficient and faster in terms of creation of advertising, etc., as
(08:54):
well as content creation.
So the conversations were continual, uh but also there was leadership in the C-suite andin other areas of uh the operation that could see that we needed to lean into it to really
be able
figure it out but also be able to put guardrails around it so that we didn't step too farto one side or the other where there would be issues later on.
(09:17):
And so I believe that the organization did the right thing in terms of understanding thatthey needed to not necessarily be on the bleeding edge of it but they definitely needed to
be more on the leading side than to your point kind of dragging along the way.
So the conversations dealt everywhere from more the legal implications from an IP side, uhall the way to the data side.
(09:39):
And so there were committees that were set up, right?
Typically always having some things that a committee that are set up in order to identifyexactly what this is technology wise and how we potentially use it and benefit from it.
And so that approach is an important one.
(10:01):
than not only at the C-suite level.
but at the levels across the organization, they really are forming teams to basically say,eh there are definitely opportunities for us to become more efficient as an organization.
There are opportunities for us to be able to leverage these tools across all the softwarewe're using, whether it's CRM, the Microsoft Suite, whatever it is, for us to get better
(10:24):
at what we're doing.
But we need to also be smart at the same time and understand what we need to do as far asthe guardrails are concerned.
So as we're testing things out, we're not launching something without understandingexactly what the implications may be.
Mm-hmm.
If.
You were going to if you were going to counsel or coach somebody that's in a comparablerole to where you used to be as a CMO for an organization, maybe in a regulated industry
(10:53):
or not.
What's one thing that you would tell them to be prepared to address head on?
And what's maybe one tip that you could give them if they're looking to try and bringgenerative AI tools into the organization?
Then we'll start moving forward with the conversation here.
But I just I feel like we've got such a wide range of experience sets out there with
organizations in terms of maturity about where they are on their AI journey.
(11:17):
I know a lot of people are still having conversations that you had three years ago.
So what can you share with us there?
Yeah, so first of all, should not be compartmentalized.
And what I mean by that is should not be within one specific department, which typicallythings that seem technical in nature are typically in the IT world to begin with, which is
a mistake.
(11:37):
The reason it's a mistake is because AI integrates with everything across the board.
And so if you house it within one specific area because you feel comfortable with it,you're going to lose out in terms of what opportunities there are across your
organization.
And so even if it seems a little concerning, right, and it does seem technical in nature,you need to dig in.
(12:00):
So what I would say to anyone that I'm consulting with is they have to dig in, they haveto learn it.
They can't sit on the sidelines and look at it and say, oh, you know what, let's have ourCIO lead this or let's have our CMO lead this.
And they really need to dig in and understand themselves.
They need to be able to.
really understand by utilizing it and not just utilizing a chat GPT as an example as asearch engine, but looking at what they're doing day to day and figure out by testing,
(12:32):
figure out how they themselves can become more efficient because it's across the board.
Obviously it's with customer service.
It's with customer experience.
It really is soup to nuts nowadays.
And so really the, you know, the genies out of the bottle.
And so
if you believe that it can be contained within one area or the next, you're going to bemissing out tremendously.
(12:54):
So you have to look at it kind of from the standpoint of...
what tools should we use, right?
It's chat, it's Claude, or it's more developed tools that are very specific forindustries, uh legal, et cetera, and really look at it from a standpoint as a starting
point of efficiency.
So where are we leaking time, right?
(13:16):
Where do we spend a lot of time?
Let's look at our entire workflow, department to department.
Where do we spend a lot of time and how can AI as our assistant, if you will, help withreducing time?
to market for various things, whether it's products and product development, whether it'stechnology, whether it's marketing across the board.
(13:39):
Love it.
If this interview stopped right there, think you just gave our listeners some solid gold.
So I appreciate that very much.
Thank you, Andar.
It's all by trial and error, trust me.
Continually.
Continually.
Failing forward, that's right.
All right, I want to go back to one of the things you touched on before and dig a littlebit deeper here.
(14:00):
I want to talk about the propensity modeling and the email marketing segmentation that youreferred to.
I'm interested in maybe having you walk us through a little bit more how that worked andwhat were you trying to predict versus what the data actually told you, because I've had
some conversations with folks where what they thought they were going to end up with, theydidn't actually end up with.
(14:20):
And maybe that was your experience.
it wasn't.
So walk us through what worked and what were you trying to predict and what did the datatell you?
Sure.
Yeah, so what I can share is basically what we were looking at is kind of, if you will,kind of the next best product, so to speak, right?
So if you're dealing with a client of financial services, let's say in banking, we werebasically looking at and saying, okay, here are the set of products that they have.
(14:47):
And then based on the accounts that they have, based on their behavior, everything we cancollect about them, plus additional data that we had available.
What would make the most sense in terms
of their need.
So as I was speaking about earlier, is it looking as though they may need a personal loanbecause they're drawing down a lot from their deposits or other activities are occurring
(15:11):
or is it something where a credit card might be helpful for them?
digging into that and being able to run models was really what we were working with on theBI side.
And so on the BI team, working with a lot of experts as far as data analytics andidentifying based on what we were looking to do.
we basically, again, we had to start at the strategy is, where did we need to generate notonly business, but that business then could support our customer set, right?
(15:43):
So it's really all in one.
Starting off with that point to then take a look and say, okay, what do we can what can wemodel and then what can we test?
And so as an example, let's say that we were we were looking at it and saying, okay wewant to make sure that our customers are Healthy as far as finances are concerned, right?
And what does that look like?
(16:04):
And so being able to then work with the BI team to be able to model that come back andthen look at the entire customer set and say, okay here we have identified these
tranches of customers that need different types of products and then being able to testthat out.
And so again, it's trial by error or trial and error uh in terms of trying to figure out,what is going to work?
(16:28):
So an easy uh method for that, as far as marketing medium is concerned, or easier, shouldsay, was email marketing.
So first off, putting out opportunities for our existing customers and figuring out whatresonated.
So based on looking at it on one side of kind of that A-B testing, here's the standardfilters that we would select.
(16:51):
And then let's look
at that against a list that was essentially created by utilizing those propensity modelsand then comparing the two.
And what we saw then was that uh much larger increase in open rate, which means thatthere's higher interest and then all the way down the funnel to conversion as well.
So how long did it take from say the initial model build uh to getting meaningful results?
(17:16):
Cause I think you walked through a couple of different phases there.
I'm just interested in maybe it's months, maybe it was a year.
What, how long did that take?
Yeah, so just to start off with it was probably more than six months because this is alsoearlier in kind of the AI days.
So things weren't as advanced.
Things didn't move as quickly.
Even now I can see out in the marketplace is taking much less time as long as you as longas you have access to the data.
(17:42):
The data is still very key and then being able to also merge other data sources as well.
So it definitely took a fairly significant amount of time.
But that amount of time has definitely been
decreased over the past year or so just with the advancement in tools and the speed ofutilization of AI overall.
(18:04):
And the improvement that you saw with the click to open rates and and some of the other uhpositive indicators that you saw Were you able to uh to put that into real dollar terms
for the business or as a percentage of uh net profit or how were you?
uh Measuring success and what kind of success did you see there in terms of bottom lineimpact?
(18:27):
Yeah, so what I can share is that we were taking a look at it overall in terms of productvolume across specific products and then also products per customer or products per
member, as you would say, kind of in the CU world, identifying where we're at at one pointin time to the next point in time and then what did that look like as far as the bottom
line.
So we looked at it in multiple ways to be able to figure out is it actually beingeffective or not.
(18:52):
And what we saw is that it was being effective and so we were utilizing it more and more.
Then you kind of fast forward also in terms of being able to utilize AI within other toolsas well, which also help with the overall, not only awareness, but also overall conversion
as far as those products are concerned.
So utilization for members of the customer service team, right?
(19:17):
Whether utilizing a CRM or utilizing other tools to help be able to push that forward,right?
So not just utilizing email to make that effective, but
really across the board with every touch point that you can.
Got it.
Well, and I think that's a perfect uh segue into something I wanted to ask you about.
uh When you and I chatted before our call, you talked about the key not being uh specificAI tools or identifying specific AI tools, but figuring out where they fit in existing
(19:48):
workflows.
And so I'm wondering if we can dig into that a little bit since you touched on that with acouple of different departments there.
What does a workflow first AI adoption look like in practice?
Yeah, so first off, again, you need to take a look at what is your workflow.
As you're working on, let's say, a campaign on the marketing side, how are you startingoff?
(20:12):
Is it starting off with you are looking at generating content to begin with as a startingpoint after you've got the strategy and the business case, et cetera?
so that if you're looking at beginning with content, what tools can you use in order tomake that process not only more efficient, but also be able to leverage it just from a
(20:35):
power perspective.
So what tools can you utilize to be able to fit in?
So it's really mapping out the workflow and then figuring out at this point, let's trythis.
At this next point, let's try this.
But not looking at it as a replacement to begin with, but looking as an assistant alongthe way.
It
you can empower your team to be able to utilize those tools to be able to do more, moreeffectively in a shorter period of time.
(21:02):
Did you find that people uh attempted or maybe desired to use those tools outside of adefined workflow since they had this shiny new toy?
For sure.
mean, uh any good marketer is going to be curious.
But you need to also, of course, have those, as we talked about, those guardrails.
But you need to be able to also be able to test and be able to figure out things on yourown as well.
(21:27):
I think the interesting thing is that maybe it was roughly around two years ago.
was in uh several different meetings at conferences.
And the big debate was people being so concerned about AI replacing their jobs.
which is still a concern today.
What I did see with various marketing teams that I would interact with plus my own is alot of curiosity in terms of how they can figure out to essentially help them to be more
(21:54):
empowered in their role.
Hmm.
And so while I believe that that fear and in certain instances, rightly so, still existsout there, I think that the people who are doing very, very well are also looking at it as
how they can basically empower themselves to do more.
So whether it's a designer that is utilizing tools within the creative suite for Adobe andutilizing AI within Photoshop, which has gotten much, much better, still not perfect,
(22:27):
but much better than it used to be.
I don't see that as so much a replacement for them as it is a way for them to become moreefficient.
And then of course with tools like Adobe, you have more of that walled garden, so you haveless IP issues.
Of course, nowadays when you're utilizing Gemini,
or utilizing chat GPT, the image creation is increased exponentially.
(22:51):
But then if you're working within a regulated environment, are you really allowed toutilize those images?
Or is it more just you're trying to figure out what you can generate from ideas andutilizing it as far as storyboarding?
Because that is also an effective way to utilize gen AIs.
You might not be creating a product that's gonna go out to market, but you canconceptualize things which help to speed up not only the strategic process, but also the
(23:15):
creative process as well.
I think that was a really uh solid example of how you provide uh you may have even usedthe term a walled garden for people to go play in
Have you found that outside of specific tool sets, suites like that, that organizationsyou're consulting with like to set up or are interested in setting up their own walled
(23:38):
garden?
So, you know, do they deploy their own Claude models in an Azure or a bedrock stack orthings like that?
And then just say to folks, hey, here's this tool, go try it out.
Or are more organizations really focused on identifying that workflow first and gettingpeople adapted into
type of mindset.
(23:59):
What are you seeing?
Are you seeing both?
What do you see out there?
Yeah, that's a great question.
I I really see, I see both because it depends on the maturity of the organization, right?
So the organizations that are a little bit more mature and have had time to kind of workwith AI for a little bit longer are setting up more of those walled gardens and the
ability to kind of detail and look at the workflow as opposed to kind of just, you know,looking at one tool and the other and it not really being connected or integrated into
(24:24):
their workflow.
So I think it depends on the experience of the individual and also the maturity of theorganization.
But what I always
recommend is starting off and taking a look at first what's the strategy for the companyoverall on the marketing side where does marketing fit into that strategy and making sure
that you really map out your entire workflow and figuring out what tools you can utilizein order to make that workflow more efficient to support the overall company strategy.
(24:55):
Let's go back to, uh again, one of the things you touched on initially, uh working in ahighly regulated industry, because financial services, you've got compliance
considerations that a lot of other sectors simply don't.
And so I feel like based on my experience and what you're saying in our conversationtoday, there's a tension there.
(25:15):
And that tension that as I see it is that tension between the desire to move fast ah andthen the compliance needs.
So I'm interested, how did you
navigate that tension and how do you with your clients today navigate that tension betweenmoving fast and compliance needs?
Yeah, so for sure.
mean, AI can definitely amplify the risk for sure.
(25:39):
And so it's about first, if you're working in that regulated environment, know, small, midor large sized companies is as, as most uh smart marketers would do is to create alignment
between the compliance team, marketing team, and other teams that are involved in thatprocess.
Legal is sometimes within the legal side, the compliance is plugged in or sometimes
(26:03):
Sometimes they're separate departments, but work together.
So first off is to create alignment, right?
In that they're all working towards the same goal for the same strategy and not gettingtoo far ahead before you're checking in.
And what I mean by that is that obviously marketers are very curious.
They like to experiment, which is a good thing.
(26:25):
But at the same time, you need to be eyes wide open about that and make sure that you'repartnering up with the compliance, the risk teams, legal teams.
oh
and be able to create something where because you're in alignment, then you have theability to test things as well.
And so you're testing things obviously internally first, calling upon industry informationas well, and then trying to figure out what you can utilize in order to put out into the
(26:51):
market.
with still staying within uh the compliance that you need to be.
But you also need to be brave in the sense that you're looking at it from a marketingstandpoint to support the organization.
And so there are business decisions that need to be made.
And as long as you are following what you need to as far as guidelines are concerned forwhatever regulated industry you are, you also still need to be kind of at the forefront of
(27:20):
leading things as opposed to sitting
and being told what to do.
I absolutely love uh how you finished up your remarks on that question there with thisidea of being brave.
I can't tell you how many C-suites I've walked into where I find cowardice uh that ismasquerading as responsibility.
(27:43):
ah And so this idea of being brave and leaning into this change moment that we're in rightnow, I think that's that's really solid advice.
I'm really glad that you went there because that's
That's an underappreciated aspect of this is that so many times, especially in compliancebased industries, we have leaders that are managing the risk.
(28:03):
And I think they manage the risk so much that they don't see the boulder that's bearingdown on them, uh know, a la Indiana Jones in the Temple of Doom.
Right.
So there's uh there's a point where your your risk mitigation, I think, turns against youand bites you quite hard.
That's true and you're going to see that from competitors, right?
(28:24):
So if you're in the financial services area, you're going to have competitors that justnaturally based on the size of organization or perhaps the rules around it can move
quicker than you.
And so if they're in fintech, as an example, they're not necessarily quote unquote a bankor a credit union.
Obviously there's still rules and regulations around that.
(28:46):
But by nature, some of those smaller, more nimble companies are going to sometimes
able to beat you to the market.
So you have to look at that standpoint and also figure out, okay well then what can we doin order to help us as opposed to just reacting or you know like you were saying earlier
just being a laggard and if you're a laggard someone else is going to take your business.
(29:07):
It's just that's just the way it is.
I don't think that there's a widespread appreciation of how fast those those laggards aregoing to have their lunch eaten over the next two years.
oh It is it's it's absolutely happening and it's really clear to me that we've got theability now for two five and ten person firms to punch with the weight of 50 100 500
(29:31):
person firms if they're leveraging the proper tools.
And so many folks just don't see that yet.
And it's it's going to be interesting.
Now for sure, your point is well taken in that in the past you look at, let's just say, amid-sized bank as an example, which of course I've worked for a few.
Mid-sized bank they've got a good strategy.
(29:53):
They've got good products, you know, they're well accepted in the marketplace But who aretheir competitors?
You know, their competitors are the super banks, right?
They are a chase.
They are a B of a they are a city.
They are a US bank They're Wells Fargo who have marketing departments in the hundreds andin your mid-sized bank Your marketing department might be ten people might be fifteen
(30:15):
people at the most.
Maybe you're using an agency outside.
Maybe you're not This is really
to your point you have the ability to punch above your weight.
Again, always keeping in line with the regulations and making sure that you're coveringthe risk as much as makes sense, but also not shooting yourself in the foot or to your
(30:37):
point kind of relying on the responsibility cloak of why not to do something.
This is where you have a lot of opportunity.
And the same in the consulting side.
You might be a one, two, or three man or woman organization.
you have the ability now to compete because you're able to utilize tools better thanothers.
you
(30:57):
Absolutely.
All right.
So since you ended that remark on tools, let's shift uh back to tools and let's talk aboutgenerative AI.
uh And I'm less interested uh in specific tool endorsements, but I know that you and yourteam uh use ChatGBT, they used Adobe Firefly.
ah I know in your consultive practice, you're recommending a lot of different tools outthere for content creation and design.
(31:23):
What's your process either back then uh or now?
evolved.
I hope it's evolved, but what's your process for deciding which tools to adopt and just asimportantly, which ones to say, nep, see you later and take a pass on.
Yeah, no, it's a great question.
So first of all, I'm just going to state I am not an expert in any tool.
And I think anyone that tells you they're an expert in a tool with AI, you might want tosecond guess that.
(31:48):
But by saying that, you really have to test them.
And so you really have to get involved in figuring out, as an example, everybody talksabout prompts.
Prompts are important, but it's also the information that you're putting into the AI.
and figuring that out.
So you were talking about chat GPT.
So the paid version of chat GPT, that is a tremendous help and has come a long way.
(32:15):
But you also need to be able to add a certain amount of information into it in order forit to learn, understand, and be able to work with you, right, in that kind of assistant
capacity.
And so to begin with is really being able to share what the strategy is.
uh
information you don't necessarily want to put in, but you also need to obviously make sureyou have all of the settings set up properly so that you can be a little bit more
(32:43):
comfortable with it.
But really being able to put the time into the tools to figure out what they look like.
It's just a continual process as far as marketing is concerned, same as before, where youhave to test and learn.
And so it's as simple as that.
You're testing Claude for content creation, you're testing Chad GPT for content creation.
What works better?
(33:03):
Is one better than the other?
you're gonna have debate continually.
So I'm not gonna get into that debate on specific tools, but what I can say is that italso depends on how you work, the information you're putting into it and what are you
getting out of it, how quickly you can make adjustments, what you can do to quickly learnfrom it.
And again, not as a search engine, but as an actual tool that you can leverage across theboard to be able to create content, to be able to create images, to be able to develop and
(33:35):
test strategy, to be able to do research, to be able to do analysis.
All of those things and more can be done in these tools and they save you a tremendousamount of time.
Now I still see areas where there's issues, right?
Even with something as simple
as, hey, let's create, you know, the old standard, let's create a PowerPoint, right?
(33:57):
And how you're able to utilize that with AI or how you're able to utilize that even withinthe Microsoft suite, I think, as one example, is still lacking tremendously.
But then there's other tools that you're able to develop inside, as an example, ChatGPTthat you never thought possible before.
So just as an example, it's not necessarily specifically in the marketing realm, but itdoes have some applicability is the ability to
(34:22):
information with respect to, let's say, ah
an app that you want to create, but it's not necessarily going to be in the app store.
So it's more of a web-based app.
And then being able to give a framework for exactly what you're looking to create and thenbe able to actually generate and use that.
Those are things that are highly powerful, have applicability as far as marketing andreally anything else that you can dream up.
(34:48):
Those capabilities exist now so that you don't need to be able to write code.
You just need to be able to communicate effectively.
You need to be able to know exactly what your strategy is and you need to be able to thenbe able to test the tools that are being created.
So you you may start with something in ChatGPT and then it becomes a web-based app.
(35:10):
That's just one example of really leveraging a tool to a little bit more extent.
There's folks that are doing well beyond that as well, but just giving an example thatit's not just a tool as far as search engines, it's not just a tool to create some fun
images.
These are tools to be able to create actually things in business uh as well as yourpersonal life that can help you do more.
(35:32):
Mm-hmm.
You touched on a couple things that I want to go a little deeper on.
You talked about the settings.
You kind of referenced that.
And then you also referenced, I think it was earlier in our conversation, but perhaps nowagain about appending a knowledge base.
You didn't use the terms knowledge base, but basically giving information to the tool thatis going to make the output more relevant.
(35:56):
so I can get even more granular for this question here.
uh I find a lot of organizations don't understand how to append their
guidelines, their brand voice, ah or even develop their brand voice, and then tweak thosesettings to produce what is going to be a very unique, bespoke type of output for that
organization.
So here's my question, right?
(36:18):
I'm contemplating all of this.
I'm thinking about brand voice.
I'm thinking about our brand guidelines.
I'm thinking about all the cool stuff that I want to do.
How do I...
handle intellectual property concerns from my organization or from the output that I'mgenerating here, if I'm generating images and uh using logo files, things like that.
How do you counsel and coach your clients on uh that topic?
(36:40):
Yeah, so I mean it starts with really understanding the regulations within your industryand then also working with the working with working with the legal team, whether that's in
house or whether that is, you know, third party and understanding what the risks are andthen being able to, you know, make decisions.
And I think that's a way we go back to more of that walled garden approach where you havethose knowledge bases that being created by teams across the board within organizations
(37:07):
where people feel more comfortable because the information that's being pulled is
being pulled from the company's own information as opposed to being collected in variousways from across, you know, from across platforms.
So that's definitely one way to do it.
think the other thing too is that you have the ability, so let's say it's, you know, let'ssay it's style, tone of voice, et cetera, as far as messaging is concerned.
(37:31):
You know, I would approach it in the sense that you already should have brand guidelinesand then being able to leverage those, you need to be able to
You need to be comfortable.
if it makes sense in your organization to be able to build that into a knowledge base sothat you're getting information out of a tool that is more relevant.
(37:52):
You don't have to keep teaching it, right?
I think that's key.
But the tools that people are creating within-house and have those knowledge bases can betremendously effective because again, you're not having to search information across an
intranet or across a website.
It's pulling it for you.
It's relevant or it's doing it just in an automated fashion.
(38:13):
So it's not even something that you really have to pull up manually.
It's reading information within the system.
So let's say it's within the CRM.
So there's a question that's being asked by a customer.
It's reading that information.
It's already providing it back to them.
And then being able to look at that and figure out how you can utilize that for thingslike marketing.
Okay.
(38:34):
That opens the door for us to circle background of the human component, which I'm alwaysfascinated with when I'm having AI conversations.
It's this intersection of humanity and technology and getting a marketing team to actuallyadopt AI tools.
I think that's more of a change management challenge in as much as it's a technologicalone.
(38:56):
We've had a lot of guests on.
We've talked about different change management strategies, blah, blah.
All that's great.
How do you get your team on board and how do you get your clients teams on board and whatdo you do with the resistors?
Because I know they're there.
I see them when I give these workshops.
uh So how do you get teams on board and what do you do with the resistors?
Yeah, so more than anything, it's really identifying who within a specific team has aninterest to be a part of what it is you're doing on the AI side.
(39:25):
So the curiosity aspect.
So starting off with, again, the workflow and looking at it as helping to support whatthey're doing and not replace.
figuring out who those people are on your team that could benefit most early enough, andthen put together a plan, right?
So it's really a test and learn plan.
(39:46):
And then being able to then, just like you would with any project that you're doing that'snew, is be able to identify exactly what the goal is, how you're gonna measure it, and
then as you come to uh successes or failures, that you have a next step there as well.
So as far as the AI side, being able to kind of celebrate, so to speak, victories andidentify
(40:07):
identifying exactly what was done, how much time was saved, how much time, the time thatwas saved, what did that roll into an actual dollar saved, so that you can figure out,
this is really being efficient.
It didn't take away from the individual's role at all.
It actually helped them do more with, in less, in much, much less time, with potentially asuperior product and a superior outcome.
(40:31):
So,
It's not so much that there's a new template on how to do that.
It's really just making sure that you're digging into the tools that AI has to offer.
to help you do that because you'd be doing the same thing if you were testing uh outputfrom let's say, paid search on Google versus paid social, right?
(40:52):
You'd come up with a plan.
Hey, this is gonna fit into our strategy.
This is what we're looking at as far as success metrics and what we need to be theoutcome.
And then along the way, what did you do in order to get to that outcome?
So now it's like fitting in the AI tools to be able to figure that out and then figuringout which tools worked
or how you work with them.
(41:13):
Those are the things that not only I've done, but also I communicate with others is makingsure that you start with a plan.
It's really the basics.
It's always the basics.
Getting people on board is really being able to show them, let's say at the start, a goodway is just how much time you're gonna save, right?
So that you have time to do other things.
(41:36):
So start with a plan.
shouldn't just go wing it.
Radical, radical concept.
oh What do you find in terms of the the people piece?
Do you find that there's certain roles or functions that are more predisposed to AIadoption than others or is it across the board?
I think nowadays it's pretty much across the board, but there are definitely standoutswithin a marketing team.
(42:01):
I mean, obviously if you've got copywriters, if you've got designers, those are some thatstand out to begin with, as well as those that are dealing with data, right?
Those are kind of standouts.
But I believe that everyone in a traditional, so to speak, contemporary marketing team isgonna be able to benefit.
(42:22):
And a lot of it is also even just through ideation.
utilizing the tool as essentially being able to do research and again figure out what thesources is, asking for the sources, make sure these are not hallucinations, these are real
information, but it can speed up research a tremendous amount.
And so being able to do that and be able to ideate based on ideas that you have, researchthat you're doing, and then be able to figure out how you can put that in place is pretty
(42:55):
powerful because
you're being able to essentially have a sounding board without necessarily having topresent the information in a typical environment.
So you can be able to iterate in a lot of a in a lot faster manner to be able to get tothat point where you're then like hey this is a pretty good idea here's what it's based on
(43:16):
you know these are solid things I have sources right this is there's no IP issues and thensee where it can go from there.
That research piece is something that um gets overlooked an awful lot and I've been usingdeep research tools since they came out and I often times get looked at um as if I'm some
(43:38):
type of heretic uh when I can produce a deep research brief on a company.
couple that with some established workflows or prompts that I've developed and get a veryspecific targeted output for the client.
uh I sometimes I feel like it's burn the witch, right?
uh What did they just do?
So I'm interested how or since you had surfaced the issue of research and kind of broughtthat around here, I'm interested, how are you advising clients to use research and maybe
(44:06):
um what's a really good example for our listeners about how you have guided someone onusing
using research to get that type of superior output.
Sure.
You know, one way is let's say that you're looking at prospecting to begin with, right?
So kind of more on the lead generation side.
So being able to use tools, some of them are built within CRMs already, some are sittingoutside of CRMs, but being able to dig in and kind of creating a pretty quick model in
(44:36):
terms of identifying prospects, individuals as far as well as businesses, and then beingable to tie in the next steps of that.
So being able to set up sequencing and journeys as an example, as far as communication.
methods, the research that you can do in a lot quicker, more accurate fashion, it'schanged a lot.
(44:57):
And so that is a way to kind of step into that world as far as research is to say, we're abusiness, we sell things, right?
Who are we selling them to?
What is going to motivate them?
What are the triggers?
And how can we more quickly identify decision makers that are at those companies that weneed to make contact with?
(45:20):
What can we learn about those decision makers that is available through research toolsthat we can also then confirm is accurate following the right steps?
Because that's where you can do a lot of customization.
Things aren't as cookie cutter, and then you really develop
into that more model that I was talking about earlier where you may be working yourself orwith business intelligence teams and collaborating with them and being able to create
(45:47):
segmented lists of existing customers as well as potentially prospects at a much higherlevel and then be able to have a much higher return because of it.
Wonderful.
Alright.
Let's I was going to say let's jump in the way back machine, but we don't have to jump inthe way back machine.
It could be more current.
You remember that one?
(46:08):
Yeah.
I'm interested in the biggest lesson that you've learned from a mistake.
in your AI journey, either when you were in that seat or as a consultant, something thatyou'd tell another CEO or CMO to avoid because we've all made some mistakes and I'm sure
there's something that jumps out at you that would help our audience learn from yourexperiences.
(46:32):
So what's the biggest mistake you've made in your AI journey?
Yeah, so I'd say there's probably two things.
So one is, don't believe the hype, right?
Don't get too excited about a tool that says it can do X, Y, and Z until you actually testit out.
That's one.
vendor?
What?
So that's one.
uh And then probably the other is is really taking that deeper dive in terms ofcollaborating with the other teams across the organization so that you really can take a
(47:03):
look into this in a collaborative fashion.
It's not always going to be perfect.
Some of the people you're going to have to kind of pull along so to speak but reallylooking at it from the standpoint of we're in alignment because we're all supporting a
specific strategy.
Everybody's on
board with that strategy.
So getting that conversation going early enough, it's not so much what marketing wants todo this or IT wants to do this, you know, or customer service wants to do this, is what's
(47:32):
the strategy?
Everyone's bought into the strategy, they should have bought into the strategy if they'redoing their job and have had input into that and then have that alignment so that you can
collaborate with them.
towards that end goal.
It's not always going to be perfect.
It doesn't mean that there's not going to be rough steps along the way.
try to create that, try to keep that common goal in mind and continue to surface it sothat you can stay aligned and be able to get to that next point in a much easier fashion
(48:03):
as opposed to fighting it along the way.
You've referenced strategy a number of times in our conversation, and appropriately so.
So this isn't a trick question here.
ah You've referenced strategy appropriately a number of times in our conversation.
And when I go in and I work with organizations on strategy,
Most of the time I am finding that when it comes to AI, they're treating that as atactical component as part of the strategy as opposed to something that's going to reshape
(48:33):
portions of the strategy.
And I firmly believe that we're past the point where that is not possible.
I think it's not only possible, but it's probable for a lot of the AI tools and theofferings to begin to reshape the overall strategy of an organization.
So I'm interested
Have you seen what I have seen in terms of tactical adjustment based on uh AI tools in theevolving landscape there or have you actually seen it begin to tweak and adjust
(49:04):
organizational strategy?
Yeah, I I've seen both.
I mean, I think that there's always this comparison.
It's like AI and everything that AI can do is so sweeping, similar to the start of theweb, right?
Now maybe more so.
But it's interesting to your point because if you kind of think back to kind of the earlydays of people utilizing the web on a regular basis, let's say like the late 90s, right?
(49:32):
A lot of them were using it very tactical, right?
They weren't necessarily using it strategic step.
And I think the same thing is with AI, is that you have people who are looking at it froma utilization of a strategy, even just from a business model.
The folks that are gonna win is you really have to look at it from a clean slate and say,okay, utilizing AI, how can we even look at our business strategy overall and figure out
(49:56):
is it going to work continually in the next 12 months, in the next...
36 months and so forth.
So you have the opportunity now to really start doing research and be able to start doingmodeling to figure out as far as your business is concerned, do you have the right product
set?
Is it the same product set you've had for 20 years?
(50:16):
Are there other products that based on what you do well as an organization, you shouldlook to test and get into that are going to give you the type of return that you need that
are maybe is not as competitive and not as commoditized as other products?
That's a key thing.
to look at that potentially change your strategy as well as you know the overall businessitself.
(50:40):
Getting into new businesses that you haven't been in before.
Yeah.
Well, and I think that's uh
That's a really interesting point.
I've talked about that with an individual too on this uh show recently.
And I think trying to balance the hype against the reality that we're not too far off fromcompanies being able to A-B test company initiatives.
(51:06):
uh Let me rephrase that.
Being able to A-B test actual companies uh versus product offerings uh in a way that we'veA-B tested web pages and product
offerings before.
So yes, I like where you went with that conversation about the strategy because I thinkthere's a whole new vista of opportunity about to open up for organizations here in the
(51:29):
coming couple years.
Yeah, I think there's other things that are very interesting that you talked about aswell, since we're talking about here and now and in the future, that are still challenging
to get your head around, which is companies that are completely running themselvesutilizing AI, where the amount of hands-on work by humans
(51:51):
is so low that it's totally different than what you see today.
That's something to really think about and figure out how you can leverage that.
And again, I'm not saying it from the standpoint of let's replace the human side, butdifferent types of businesses that you can even test out and run without the same level of
(52:11):
investment to be able to figure out if they work.
Because sometimes when you're starting off new businesses, new products, et cetera, acrossthe board, especially in financial services and highly regulated
industries, that can cost a lot of money and it can cost a lot of time.
If you can do something within, so to speak, a walled garden and be able to run it andtest it out, you can save a tremendous amount of time and to your point that can have
(52:34):
potentially a dramatic positive effect over time on the business.
Mm-hmm.
One of the other things that pops up and I've been noodling around recently and I've seenthis out there so I can't claim that this is a wholly original thought um is how we
measure certain types of investment and the types of investment that we're willing to makein organizations.
(52:57):
You made the point earlier in our conversation that the IT department should not be incharge of a lot of this stuff, which I completely agree.
um They have a role to play, but they should not be owning AI deployment, especiallywithin individual uh business units.
And I bring that up because I think right now organizations are very comfortable with theidea of their standard overhead, right?
(53:20):
And I'm just going to pick on payroll.
Payroll is a standard overhead.
I think that this is going to shift to our conversation here from where we're solelyworried about overhead and payroll being a large component of it towards perhaps its
compute.
right, or different types of resources that we're deploying there because we no longerneed to throw people at the problems in order to scale and to achieve our results.
(53:44):
And the strategic implications of that, think, are vast and they're, I wouldn't saythey're above my pay grade, but there's a lot yet to be determined there.
I totally agree with that.
mean, we were talking about the research and being able to put together ideas and put theminto play.
Yeah.
(54:05):
Utilizing the right tools can save a tremendous amount of time with that but then alsoJust the processes that are behind certain things that are running so, know one of the
things that you know, I've seen in the in the past was utilizing essentially like ML toolsalong with other tools to be able to do things that you would think need to be Manual, but
(54:28):
you're seeing them done right before your eyes.
I this is you know, this is 18 months ago So some of it may be already well known
But you know being able to being able to run checks on data and be able to run checks onvarious amounts of information all in an automated fashion across systems all happening at
one time You can drive down, you know the amount of uh
(54:53):
human involvement by tremendous amount and also decrease the cost, increase the efficiencyand allow people to really spend time on the areas that they should be doing, as opposed
to kind of more of the monotonous work.
Mm-hmm.
So one of the amazing things about where we are now several years into the journey um andwe'll be pushing toward our hundredth episode this year on this podcast alone, um which
(55:22):
gives you a sense.
thank you.
Yeah, it'll be exciting.
um So I think one of the exciting things, um but also the terrifying thing about where weare is we've got and we referenced this before.
We've got people on on such wide ends of the spectrum.
So right now we were just talking about machine learning and leveraging a lot of this and
I referenced compute versus payroll, yet I know we've got listeners, maybe it's theirfirst time, if so, welcome uh tuning in right now.
(55:48):
And they've just been dabbling, right?
So if we were going to bring this back around to the dabbler, to the person, to the leaderthat's just getting in right now, what type of minimum investments should they be thinking
about to begin exploring this?
And what should they be doing to begin kicking this off in their organization?
Because the best time to start was yesterday, candidly.
(56:09):
ah The second best time is this Monday morning.
Right.
Yeah, absolutely.
What I would share is that one is taking a look at your business and then being able toidentify what tools you want to test with it.
If you've got a circle within your industry that you can tap into in terms of what theyhave used, the learnings from that, both bad and good, can help speed up the process.
(56:38):
And then once you're looking at, what can I use as, so to speak, like a proof of concept,I think is an important place to start with.
So making sure that you're identifying, let's say like three, you know, high leverage usecases, tying them of course to performance, you know, from a revenue standpoint, from a
cost standpoint, making sure, you know, who is actually owning that, you're measuring it,and then making sure that based on what you're measuring, you're able to proceed in, you
(57:09):
know, one or many directions.
And at the same time, you're still building for yourself and your organization, yourliteracy as far as AI is concerned.
And that comes by doing.
It does not come by watching.
It's one of these things that from the CEO level, you know, all the way across anorganization that people have to, people have to really dig in and learn.
(57:33):
And there's a lot of different resources for people to learn how the systems work.
know, LinkedIn learning is a great source.
There's also Google as well as many others that can give you perhaps a little bit moreobjective feel than by starting off with some of the tools directly.
And so you're also learning from people who are more knowledgeable.
(57:56):
I kind of hesitate with the expert side, but are more knowledgeable than maybe where yousit is a good place to start so that you really can dig in.
But you got to dig in now because again, so many people are utilizing AI really just as asearch engine.
and it's so exponentially beyond that, it's really difficult to even talk about those twothings in the same breath.
(58:20):
I think that that is a fantastic place for us to conclude our conversation today.
An hour goes fast.
And uh if I was going to pull three things out from our conversation today, it would be bebrave, right?
Dig in now and find the air quote experts that can help you with your journey because theyare out there even if we are loathe to call them and ourselves experts from time to time.
(58:46):
So.
If people want to take advantage of your expertise, Andar, where can they find you?
What's the best way to reach you and get in touch?
Absolutely, thank you for that.
So on LinkedIn, so Ondartarlo, very easy to find, unique name, all over LinkedIn.
uh In addition to LinkedIn, also you can go to my website, Ondartarlo.org, and you canfind me there, you can send me a request, and I'll be able to get back in touch with you.
(59:15):
Wonderful.
I appreciate it.
I really enjoyed our conversation today and I appreciate you coming on and sharing yourknowledge and experience with our audience.
Thank you very much.
Thank you, Chen.
It's a pleasure.
All right.
And thank you to our listeners for once again tuning into AI for the C-suite.
If this episode was useful, subscribe wherever you get your podcasts, follow us onLinkedIn and all the socials, just not Twitter, X, whatever they call it these days, I
(59:39):
don't know.
And check out AI for the C-suite dot com.
Until next time, keep your algorithms running, your leadership evolving and your AI incheck.
Take care, everybody.
I'm Chad Harvey and this is AI for the C-suite, the show for senior leaders who know AImatters and need to figure out what to do about it.
Each episode, we dig into what it all actually means for middle market organizations, howit's changing decisions, strategy, leadership, and the nature of work.
Let's get into it.
Today, I'm joined by Andar Tarlow, a marketing exec with more than 20 years of experiencebuilding brands and financial services, motorsports, and lifestyle industries.
(00:33):
Andar served as chief marketing officer at Connecta Federal Credit Union and PacificPremier Bank, where he led campaigns that generated over $1.7 billion in product volume
and earned him Diamond Awards and a Top 100 Marketer Recognition.
Andar's perspective is especially relevant for our listeners because he didn't just watchAI happen from the sidelines.
(00:56):
He rolled up his sleeves and put it to work.
His team used propensity modeling and machine learning to transform how they targetedcustomers and then layered in generative AI tools to overhaul their creative workflows.
We're talking five times improvement in campaign performance and 75 % reductions inproduction time.
(01:17):
Andar, welcome to AI for the C-suite.
Hey, Chad, how are you?
I am doing really well and I gotta say I am super excited to have you on today because thenatural approach point for a lot of organizations is through marketing and then I also get
a lot of questions from folks about highly regulated industries and you were in that aswell.
(01:38):
So I'm really looking forward to our conversation today.
Good good.
Well every superhero has an origin story so let's talk about your AI origin story maybe.
Take me back to the moment that AI stopped being abstract for you and started beingsomething that you needed to act on.
I'm sure that you saw something in the world of marketing that made you take notice andsay, we really need to lean into this.
(02:05):
So what was that?
What did that look like?
How did your experience unfold there?
Yeah, well, first of all, AI has moved so quickly over the last few years that you'retalking about something a month ago and it could have already advanced tremendously beyond
that and the way you're using it today is totally different.
But going back to the origin question was really working with some very, very smart folksin our business intelligence team who were utilizing AI and ML machine learning for some
(02:38):
time and identifying, how can we take a look
our book of business, existing customers, existing members on the credit union side andfigure out how we can do better for them.
How can we identify what types of products they might need on the financial side, whetherthat be a credit card, whether that be a personal loan, whether that be something on the
(03:01):
wealth management side is to figure out exactly what made the most sense.
And so by being able to dig in and be able to do things like A-B testing between
utilizing the preventively modeling in comparison to general filters and then figuring outyou know what actually was working and why it wasn't working.
So once we started seeing pickup in terms of simple campaigns like email marketing andsimple and simple factors in terms of moving forward like email open rate increasing
(03:33):
dramatically and then conversion conversion rates driving up as well.
We knew there was something to that and so
that we continue to kind of push on the accelerator to figure out exactly where we couldgo from there.
So that was kind of like the aha moment, but obviously everything that was happeningaround us at the time, and obviously continues exponentially now as far as AI, right?
(03:57):
Both on kind of the AI learning and modeling as far as data is concerned, and then also onthe gen AI side, which a lot of teams in marketing started to use a couple of years ago or
even earlier.
ah very immature versions of GEN.AI, whether it's for image creation, content creation, orother.
(04:18):
And now, taking a look at where it is today.
Also being able to figure out exactly when you say AI, what does that mean?
And then figuring out not so much, hey, we're going to go use AI for X, Y, and Z, but morealong the lines of starting off of what is the strategy overall as far as the company is
concerned?
is, how is marketing supporting that strategy?
(04:41):
And then how can we integrate AI into the entire workflow that we have?
think that's really key.
And you had a lot of marketing
including ourselves that started off on the content side, right?
So to help become a little bit more efficient.
So starting off with ideation and then working into things like social media and otherelements that content creation comes into.
(05:04):
But then at the same time, alongside that, and you always have to keep this in mind as faras regulated industry, so I'm glad you brought that up, is the fear factor, right?
So there's the fear factor in terms of, okay, what does AI mean?
What are the issues as far as IP rights are concerned in terms of content?
as well as in terms of images that are used and so we were basically doing a lot ofresearch at the same time and figuring out exactly how we could use it effectively and
(05:34):
then kind of fast forward to today where I'm doing a lot of consulting and utilizing AI ina manner of which I can become more and more efficient for clients and asking the right
questions and not just looking at AI as a search engine which a lot of people
still do, right?
It's like, yeah, no, I just use chat GPT, you know, versus Google.
(05:56):
Well, I mean, you're really not even scratching the surface as far as the power of how youcan use it as really an assistant.
And then of course, keeping in check.
whether you're in a regulated industry or not, always keeping in check with what are thesources of the information.
(06:16):
And so I constantly hammer on that because you'd still see it.
It happens, I would say, less from where I am than it did a year or two ago.
But you still see those hallucinations from time to time as far as GenAI is concerned.
And you need to make sure that you are double checking all of the work.
You're not just copying and pasting.
(06:37):
They really understand the sources.
and then you're utilizing it as a way, as I mentioned, to become more efficient and to beable to do more in a shorter period of time that it has more impact and then be able to
utilize and test it along the way.
I'm very interested in what the conversation looked and sounded like when you initiallyhad it.
(07:00):
And let me just give you some context for where I'm asking this question from.
ah You've.
been there, you've done that.
You have been as Dean Acheson famously wrote in his book title, President Creation, right?
You worked for an organization that leaned into this technology early on.
I can tell you, because I also do a lot of consulting, I've gotten called into six, maybeeight different banks and credit unions over the last couple of years.
(07:25):
And with the exception of some very broad stuff, almost none of them want to move forward,which is something I see a lot of in the middle market right now.
An awful lot of foot dragging.
think 2026 is going to be
wake up year for a lot of people in many ways positive and negative.
Be that as it may.
The highly regulated industries the financial services sector in particular seems veryreluctant to move forward.
(07:49):
So did you have to sell this?
Did you have to pitch it to the leadership team or was it pretty much.
OK, hey, the guys over in B.I.
or the gals did what they needed to do.
They've kind of cracked open the door.
This is aligned with our strategy.
Away we go.
What did the conversation look like?
I'm fascinated by that just because of all the roadblocks I see out there withorganizations.
(08:09):
Yeah, you're absolutely correct.
mean, in the industry, whether it's financial services or others, there's such a varyingdegree in terms of experience.
You have people who are getting close to being advanced.
I'd say it's difficult to really call anyone in the field an expert nowadays, consideringhow quickly it's moving and how you're utilizing it in your day-to-day or strategic work.
(08:33):
But basically, the conversations were to your point, which is you had the BI team that hadopened it
in terms of the analytics end of things, which was great.
But again, that's different than the Gen.ai side.
And what I mean by that is, now we're going to see if we can utilize it for imagery to beable to become more efficient and faster in terms of creation of advertising, etc., as
(08:54):
well as content creation.
So the conversations were continual, uh but also there was leadership in the C-suite andin other areas of uh the operation that could see that we needed to lean into it to really
be able
figure it out but also be able to put guardrails around it so that we didn't step too farto one side or the other where there would be issues later on.
(09:17):
And so I believe that the organization did the right thing in terms of understanding thatthey needed to not necessarily be on the bleeding edge of it but they definitely needed to
be more on the leading side than to your point kind of dragging along the way.
So the conversations dealt everywhere from more the legal implications from an IP side, uhall the way to the data side.
(09:39):
And so there were committees that were set up, right?
Typically always having some things that a committee that are set up in order to identifyexactly what this is technology wise and how we potentially use it and benefit from it.
And so that approach is an important one.
(10:01):
than not only at the C-suite level.
but at the levels across the organization, they really are forming teams to basically say,eh there are definitely opportunities for us to become more efficient as an organization.
There are opportunities for us to be able to leverage these tools across all the softwarewe're using, whether it's CRM, the Microsoft Suite, whatever it is, for us to get better
(10:24):
at what we're doing.
But we need to also be smart at the same time and understand what we need to do as far asthe guardrails are concerned.
So as we're testing things out, we're not launching something without understandingexactly what the implications may be.
Mm-hmm.
If.
You were going to if you were going to counsel or coach somebody that's in a comparablerole to where you used to be as a CMO for an organization, maybe in a regulated industry
(10:53):
or not.
What's one thing that you would tell them to be prepared to address head on?
And what's maybe one tip that you could give them if they're looking to try and bringgenerative AI tools into the organization?
Then we'll start moving forward with the conversation here.
But I just I feel like we've got such a wide range of experience sets out there with
organizations in terms of maturity about where they are on their AI journey.
(11:17):
I know a lot of people are still having conversations that you had three years ago.
So what can you share with us there?
Yeah, so first of all, should not be compartmentalized.
And what I mean by that is should not be within one specific department, which typicallythings that seem technical in nature are typically in the IT world to begin with, which is
a mistake.
(11:37):
The reason it's a mistake is because AI integrates with everything across the board.
And so if you house it within one specific area because you feel comfortable with it,you're going to lose out in terms of what opportunities there are across your
organization.
And so even if it seems a little concerning, right, and it does seem technical in nature,you need to dig in.
(12:00):
So what I would say to anyone that I'm consulting with is they have to dig in, they haveto learn it.
They can't sit on the sidelines and look at it and say, oh, you know what, let's have ourCIO lead this or let's have our CMO lead this.
And they really need to dig in and understand themselves.
They need to be able to.
really understand by utilizing it and not just utilizing a chat GPT as an example as asearch engine, but looking at what they're doing day to day and figure out by testing,
(12:32):
figure out how they themselves can become more efficient because it's across the board.
Obviously it's with customer service.
It's with customer experience.
It really is soup to nuts nowadays.
And so really the, you know, the genies out of the bottle.
And so
if you believe that it can be contained within one area or the next, you're going to bemissing out tremendously.
(12:54):
So you have to look at it kind of from the standpoint of...
what tools should we use, right?
It's chat, it's Claude, or it's more developed tools that are very specific forindustries, uh legal, et cetera, and really look at it from a standpoint as a starting
point of efficiency.
So where are we leaking time, right?
(13:16):
Where do we spend a lot of time?
Let's look at our entire workflow, department to department.
Where do we spend a lot of time and how can AI as our assistant, if you will, help withreducing time?
to market for various things, whether it's products and product development, whether it'stechnology, whether it's marketing across the board.
(13:39):
Love it.
If this interview stopped right there, think you just gave our listeners some solid gold.
So I appreciate that very much.
Thank you, Andar.
It's all by trial and error, trust me.
Continually.
Continually.
Failing forward, that's right.
All right, I want to go back to one of the things you touched on before and dig a littlebit deeper here.
(14:00):
I want to talk about the propensity modeling and the email marketing segmentation that youreferred to.
I'm interested in maybe having you walk us through a little bit more how that worked andwhat were you trying to predict versus what the data actually told you, because I've had
some conversations with folks where what they thought they were going to end up with, theydidn't actually end up with.
(14:20):
And maybe that was your experience.
it wasn't.
So walk us through what worked and what were you trying to predict and what did the datatell you?
Sure.
Yeah, so what I can share is basically what we were looking at is kind of, if you will,kind of the next best product, so to speak, right?
So if you're dealing with a client of financial services, let's say in banking, we werebasically looking at and saying, okay, here are the set of products that they have.
(14:47):
And then based on the accounts that they have, based on their behavior, everything we cancollect about them, plus additional data that we had available.
What would make the most sense in terms
of their need.
So as I was speaking about earlier, is it looking as though they may need a personal loanbecause they're drawing down a lot from their deposits or other activities are occurring
(15:11):
or is it something where a credit card might be helpful for them?
digging into that and being able to run models was really what we were working with on theBI side.
And so on the BI team, working with a lot of experts as far as data analytics andidentifying based on what we were looking to do.
we basically, again, we had to start at the strategy is, where did we need to generate notonly business, but that business then could support our customer set, right?
(15:43):
So it's really all in one.
Starting off with that point to then take a look and say, okay, what do we can what can wemodel and then what can we test?
And so as an example, let's say that we were we were looking at it and saying, okay wewant to make sure that our customers are Healthy as far as finances are concerned, right?
And what does that look like?
(16:04):
And so being able to then work with the BI team to be able to model that come back andthen look at the entire customer set and say, okay here we have identified these
tranches of customers that need different types of products and then being able to testthat out.
And so again, it's trial by error or trial and error uh in terms of trying to figure out,what is going to work?
(16:28):
So an easy uh method for that, as far as marketing medium is concerned, or easier, shouldsay, was email marketing.
So first off, putting out opportunities for our existing customers and figuring out whatresonated.
So based on looking at it on one side of kind of that A-B testing, here's the standardfilters that we would select.
(16:51):
And then let's look
at that against a list that was essentially created by utilizing those propensity modelsand then comparing the two.
And what we saw then was that uh much larger increase in open rate, which means thatthere's higher interest and then all the way down the funnel to conversion as well.
So how long did it take from say the initial model build uh to getting meaningful results?
(17:16):
Cause I think you walked through a couple of different phases there.
I'm just interested in maybe it's months, maybe it was a year.
What, how long did that take?
Yeah, so just to start off with it was probably more than six months because this is alsoearlier in kind of the AI days.
So things weren't as advanced.
Things didn't move as quickly.
Even now I can see out in the marketplace is taking much less time as long as you as longas you have access to the data.
(17:42):
The data is still very key and then being able to also merge other data sources as well.
So it definitely took a fairly significant amount of time.
But that amount of time has definitely been
decreased over the past year or so just with the advancement in tools and the speed ofutilization of AI overall.
(18:04):
And the improvement that you saw with the click to open rates and and some of the other uhpositive indicators that you saw Were you able to uh to put that into real dollar terms
for the business or as a percentage of uh net profit or how were you?
uh Measuring success and what kind of success did you see there in terms of bottom lineimpact?
(18:27):
Yeah, so what I can share is that we were taking a look at it overall in terms of productvolume across specific products and then also products per customer or products per
member, as you would say, kind of in the CU world, identifying where we're at at one pointin time to the next point in time and then what did that look like as far as the bottom
line.
So we looked at it in multiple ways to be able to figure out is it actually beingeffective or not.
(18:52):
And what we saw is that it was being effective and so we were utilizing it more and more.
Then you kind of fast forward also in terms of being able to utilize AI within other toolsas well, which also help with the overall, not only awareness, but also overall conversion
as far as those products are concerned.
So utilization for members of the customer service team, right?
(19:17):
Whether utilizing a CRM or utilizing other tools to help be able to push that forward,right?
So not just utilizing email to make that effective, but
really across the board with every touch point that you can.
Got it.
Well, and I think that's a perfect uh segue into something I wanted to ask you about.
uh When you and I chatted before our call, you talked about the key not being uh specificAI tools or identifying specific AI tools, but figuring out where they fit in existing
(19:48):
workflows.
And so I'm wondering if we can dig into that a little bit since you touched on that with acouple of different departments there.
What does a workflow first AI adoption look like in practice?
Yeah, so first off, again, you need to take a look at what is your workflow.
As you're working on, let's say, a campaign on the marketing side, how are you startingoff?
(20:12):
Is it starting off with you are looking at generating content to begin with as a startingpoint after you've got the strategy and the business case, et cetera?
so that if you're looking at beginning with content, what tools can you use in order tomake that process not only more efficient, but also be able to leverage it just from a
(20:35):
power perspective.
So what tools can you utilize to be able to fit in?
So it's really mapping out the workflow and then figuring out at this point, let's trythis.
At this next point, let's try this.
But not looking at it as a replacement to begin with, but looking as an assistant alongthe way.
It
you can empower your team to be able to utilize those tools to be able to do more, moreeffectively in a shorter period of time.
(21:02):
Did you find that people uh attempted or maybe desired to use those tools outside of adefined workflow since they had this shiny new toy?
For sure.
mean, uh any good marketer is going to be curious.
But you need to also, of course, have those, as we talked about, those guardrails.
But you need to be able to also be able to test and be able to figure out things on yourown as well.
(21:27):
I think the interesting thing is that maybe it was roughly around two years ago.
was in uh several different meetings at conferences.
And the big debate was people being so concerned about AI replacing their jobs.
which is still a concern today.
What I did see with various marketing teams that I would interact with plus my own is alot of curiosity in terms of how they can figure out to essentially help them to be more
(21:54):
empowered in their role.
Hmm.
And so while I believe that that fear and in certain instances, rightly so, still existsout there, I think that the people who are doing very, very well are also looking at it as
how they can basically empower themselves to do more.
So whether it's a designer that is utilizing tools within the creative suite for Adobe andutilizing AI within Photoshop, which has gotten much, much better, still not perfect,
(22:27):
but much better than it used to be.
I don't see that as so much a replacement for them as it is a way for them to become moreefficient.
And then of course with tools like Adobe, you have more of that walled garden, so you haveless IP issues.
Of course, nowadays when you're utilizing Gemini,
or utilizing chat GPT, the image creation is increased exponentially.
(22:51):
But then if you're working within a regulated environment, are you really allowed toutilize those images?
Or is it more just you're trying to figure out what you can generate from ideas andutilizing it as far as storyboarding?
Because that is also an effective way to utilize gen AIs.
You might not be creating a product that's gonna go out to market, but you canconceptualize things which help to speed up not only the strategic process, but also the
(23:15):
creative process as well.
I think that was a really uh solid example of how you provide uh you may have even usedthe term a walled garden for people to go play in
Have you found that outside of specific tool sets, suites like that, that organizationsyou're consulting with like to set up or are interested in setting up their own walled
(23:38):
garden?
So, you know, do they deploy their own Claude models in an Azure or a bedrock stack orthings like that?
And then just say to folks, hey, here's this tool, go try it out.
Or are more organizations really focused on identifying that workflow first and gettingpeople adapted into
type of mindset.
(23:59):
What are you seeing?
Are you seeing both?
What do you see out there?
Yeah, that's a great question.
I I really see, I see both because it depends on the maturity of the organization, right?
So the organizations that are a little bit more mature and have had time to kind of workwith AI for a little bit longer are setting up more of those walled gardens and the
ability to kind of detail and look at the workflow as opposed to kind of just, you know,looking at one tool and the other and it not really being connected or integrated into
(24:24):
their workflow.
So I think it depends on the experience of the individual and also the maturity of theorganization.
But what I always
recommend is starting off and taking a look at first what's the strategy for the companyoverall on the marketing side where does marketing fit into that strategy and making sure
that you really map out your entire workflow and figuring out what tools you can utilizein order to make that workflow more efficient to support the overall company strategy.
(24:55):
Let's go back to, uh again, one of the things you touched on initially, uh working in ahighly regulated industry, because financial services, you've got compliance
considerations that a lot of other sectors simply don't.
And so I feel like based on my experience and what you're saying in our conversationtoday, there's a tension there.
(25:15):
And that tension that as I see it is that tension between the desire to move fast ah andthen the compliance needs.
So I'm interested, how did you
navigate that tension and how do you with your clients today navigate that tension betweenmoving fast and compliance needs?
Yeah, so for sure.
mean, AI can definitely amplify the risk for sure.
(25:39):
And so it's about first, if you're working in that regulated environment, know, small, midor large sized companies is as, as most uh smart marketers would do is to create alignment
between the compliance team, marketing team, and other teams that are involved in thatprocess.
Legal is sometimes within the legal side, the compliance is plugged in or sometimes
(26:03):
Sometimes they're separate departments, but work together.
So first off is to create alignment, right?
In that they're all working towards the same goal for the same strategy and not gettingtoo far ahead before you're checking in.
And what I mean by that is that obviously marketers are very curious.
They like to experiment, which is a good thing.
(26:25):
But at the same time, you need to be eyes wide open about that and make sure that you'repartnering up with the compliance, the risk teams, legal teams.
oh
and be able to create something where because you're in alignment, then you have theability to test things as well.
And so you're testing things obviously internally first, calling upon industry informationas well, and then trying to figure out what you can utilize in order to put out into the
(26:51):
market.
with still staying within uh the compliance that you need to be.
But you also need to be brave in the sense that you're looking at it from a marketingstandpoint to support the organization.
And so there are business decisions that need to be made.
And as long as you are following what you need to as far as guidelines are concerned forwhatever regulated industry you are, you also still need to be kind of at the forefront of
(27:20):
leading things as opposed to sitting
and being told what to do.
I absolutely love uh how you finished up your remarks on that question there with thisidea of being brave.
I can't tell you how many C-suites I've walked into where I find cowardice uh that ismasquerading as responsibility.
(27:43):
ah And so this idea of being brave and leaning into this change moment that we're in rightnow, I think that's that's really solid advice.
I'm really glad that you went there because that's
That's an underappreciated aspect of this is that so many times, especially in compliancebased industries, we have leaders that are managing the risk.
(28:03):
And I think they manage the risk so much that they don't see the boulder that's bearingdown on them, uh know, a la Indiana Jones in the Temple of Doom.
Right.
So there's uh there's a point where your your risk mitigation, I think, turns against youand bites you quite hard.
That's true and you're going to see that from competitors, right?
(28:24):
So if you're in the financial services area, you're going to have competitors that justnaturally based on the size of organization or perhaps the rules around it can move
quicker than you.
And so if they're in fintech, as an example, they're not necessarily quote unquote a bankor a credit union.
Obviously there's still rules and regulations around that.
(28:46):
But by nature, some of those smaller, more nimble companies are going to sometimes
able to beat you to the market.
So you have to look at that standpoint and also figure out, okay well then what can we doin order to help us as opposed to just reacting or you know like you were saying earlier
just being a laggard and if you're a laggard someone else is going to take your business.
(29:07):
It's just that's just the way it is.
I don't think that there's a widespread appreciation of how fast those those laggards aregoing to have their lunch eaten over the next two years.
oh It is it's it's absolutely happening and it's really clear to me that we've got theability now for two five and ten person firms to punch with the weight of 50 100 500
(29:31):
person firms if they're leveraging the proper tools.
And so many folks just don't see that yet.
And it's it's going to be interesting.
Now for sure, your point is well taken in that in the past you look at, let's just say, amid-sized bank as an example, which of course I've worked for a few.
Mid-sized bank they've got a good strategy.
(29:53):
They've got good products, you know, they're well accepted in the marketplace But who aretheir competitors?
You know, their competitors are the super banks, right?
They are a chase.
They are a B of a they are a city.
They are a US bank They're Wells Fargo who have marketing departments in the hundreds andin your mid-sized bank Your marketing department might be ten people might be fifteen
(30:15):
people at the most.
Maybe you're using an agency outside.
Maybe you're not This is really
to your point you have the ability to punch above your weight.
Again, always keeping in line with the regulations and making sure that you're coveringthe risk as much as makes sense, but also not shooting yourself in the foot or to your
(30:37):
point kind of relying on the responsibility cloak of why not to do something.
This is where you have a lot of opportunity.
And the same in the consulting side.
You might be a one, two, or three man or woman organization.
you have the ability now to compete because you're able to utilize tools better thanothers.
you
(30:57):
Absolutely.
All right.
So since you ended that remark on tools, let's shift uh back to tools and let's talk aboutgenerative AI.
uh And I'm less interested uh in specific tool endorsements, but I know that you and yourteam uh use ChatGBT, they used Adobe Firefly.
ah I know in your consultive practice, you're recommending a lot of different tools outthere for content creation and design.
(31:23):
What's your process either back then uh or now?
evolved.
I hope it's evolved, but what's your process for deciding which tools to adopt and just asimportantly, which ones to say, nep, see you later and take a pass on.
Yeah, no, it's a great question.
So first of all, I'm just going to state I am not an expert in any tool.
And I think anyone that tells you they're an expert in a tool with AI, you might want tosecond guess that.
(31:48):
But by saying that, you really have to test them.
And so you really have to get involved in figuring out, as an example, everybody talksabout prompts.
Prompts are important, but it's also the information that you're putting into the AI.
and figuring that out.
So you were talking about chat GPT.
So the paid version of chat GPT, that is a tremendous help and has come a long way.
(32:15):
But you also need to be able to add a certain amount of information into it in order forit to learn, understand, and be able to work with you, right, in that kind of assistant
capacity.
And so to begin with is really being able to share what the strategy is.
uh
information you don't necessarily want to put in, but you also need to obviously make sureyou have all of the settings set up properly so that you can be a little bit more
(32:43):
comfortable with it.
But really being able to put the time into the tools to figure out what they look like.
It's just a continual process as far as marketing is concerned, same as before, where youhave to test and learn.
And so it's as simple as that.
You're testing Claude for content creation, you're testing Chad GPT for content creation.
What works better?
(33:03):
Is one better than the other?
you're gonna have debate continually.
So I'm not gonna get into that debate on specific tools, but what I can say is that italso depends on how you work, the information you're putting into it and what are you
getting out of it, how quickly you can make adjustments, what you can do to quickly learnfrom it.
And again, not as a search engine, but as an actual tool that you can leverage across theboard to be able to create content, to be able to create images, to be able to develop and
(33:35):
test strategy, to be able to do research, to be able to do analysis.
All of those things and more can be done in these tools and they save you a tremendousamount of time.
Now I still see areas where there's issues, right?
Even with something as simple
as, hey, let's create, you know, the old standard, let's create a PowerPoint, right?
(33:57):
And how you're able to utilize that with AI or how you're able to utilize that even withinthe Microsoft suite, I think, as one example, is still lacking tremendously.
But then there's other tools that you're able to develop inside, as an example, ChatGPTthat you never thought possible before.
So just as an example, it's not necessarily specifically in the marketing realm, but itdoes have some applicability is the ability to
(34:22):
information with respect to, let's say, ah
an app that you want to create, but it's not necessarily going to be in the app store.
So it's more of a web-based app.
And then being able to give a framework for exactly what you're looking to create and thenbe able to actually generate and use that.
Those are things that are highly powerful, have applicability as far as marketing andreally anything else that you can dream up.
(34:48):
Those capabilities exist now so that you don't need to be able to write code.
You just need to be able to communicate effectively.
You need to be able to know exactly what your strategy is and you need to be able to thenbe able to test the tools that are being created.
So you you may start with something in ChatGPT and then it becomes a web-based app.
(35:10):
That's just one example of really leveraging a tool to a little bit more extent.
There's folks that are doing well beyond that as well, but just giving an example thatit's not just a tool as far as search engines, it's not just a tool to create some fun
images.
These are tools to be able to create actually things in business uh as well as yourpersonal life that can help you do more.
(35:32):
Mm-hmm.
You touched on a couple things that I want to go a little deeper on.
You talked about the settings.
You kind of referenced that.
And then you also referenced, I think it was earlier in our conversation, but perhaps nowagain about appending a knowledge base.
You didn't use the terms knowledge base, but basically giving information to the tool thatis going to make the output more relevant.
(35:56):
so I can get even more granular for this question here.
uh I find a lot of organizations don't understand how to append their
guidelines, their brand voice, ah or even develop their brand voice, and then tweak thosesettings to produce what is going to be a very unique, bespoke type of output for that
organization.
So here's my question, right?
(36:18):
I'm contemplating all of this.
I'm thinking about brand voice.
I'm thinking about our brand guidelines.
I'm thinking about all the cool stuff that I want to do.
How do I...
handle intellectual property concerns from my organization or from the output that I'mgenerating here, if I'm generating images and uh using logo files, things like that.
How do you counsel and coach your clients on uh that topic?
(36:40):
Yeah, so I mean it starts with really understanding the regulations within your industryand then also working with the working with working with the legal team, whether that's in
house or whether that is, you know, third party and understanding what the risks are andthen being able to, you know, make decisions.
And I think that's a way we go back to more of that walled garden approach where you havethose knowledge bases that being created by teams across the board within organizations
(37:07):
where people feel more comfortable because the information that's being pulled is
being pulled from the company's own information as opposed to being collected in variousways from across, you know, from across platforms.
So that's definitely one way to do it.
think the other thing too is that you have the ability, so let's say it's, you know, let'ssay it's style, tone of voice, et cetera, as far as messaging is concerned.
(37:31):
You know, I would approach it in the sense that you already should have brand guidelinesand then being able to leverage those, you need to be able to
You need to be comfortable.
if it makes sense in your organization to be able to build that into a knowledge base sothat you're getting information out of a tool that is more relevant.
(37:52):
You don't have to keep teaching it, right?
I think that's key.
But the tools that people are creating within-house and have those knowledge bases can betremendously effective because again, you're not having to search information across an
intranet or across a website.
It's pulling it for you.
It's relevant or it's doing it just in an automated fashion.
(38:13):
So it's not even something that you really have to pull up manually.
It's reading information within the system.
So let's say it's within the CRM.
So there's a question that's being asked by a customer.
It's reading that information.
It's already providing it back to them.
And then being able to look at that and figure out how you can utilize that for thingslike marketing.
Okay.
(38:34):
That opens the door for us to circle background of the human component, which I'm alwaysfascinated with when I'm having AI conversations.
It's this intersection of humanity and technology and getting a marketing team to actuallyadopt AI tools.
I think that's more of a change management challenge in as much as it's a technologicalone.
(38:56):
We've had a lot of guests on.
We've talked about different change management strategies, blah, blah.
All that's great.
How do you get your team on board and how do you get your clients teams on board and whatdo you do with the resistors?
Because I know they're there.
I see them when I give these workshops.
uh So how do you get teams on board and what do you do with the resistors?
Yeah, so more than anything, it's really identifying who within a specific team has aninterest to be a part of what it is you're doing on the AI side.
(39:25):
So the curiosity aspect.
So starting off with, again, the workflow and looking at it as helping to support whatthey're doing and not replace.
figuring out who those people are on your team that could benefit most early enough, andthen put together a plan, right?
So it's really a test and learn plan.
(39:46):
And then being able to then, just like you would with any project that you're doing that'snew, is be able to identify exactly what the goal is, how you're gonna measure it, and
then as you come to uh successes or failures, that you have a next step there as well.
So as far as the AI side, being able to kind of celebrate, so to speak, victories andidentify
(40:07):
identifying exactly what was done, how much time was saved, how much time, the time thatwas saved, what did that roll into an actual dollar saved, so that you can figure out,
this is really being efficient.
It didn't take away from the individual's role at all.
It actually helped them do more with, in less, in much, much less time, with potentially asuperior product and a superior outcome.
(40:31):
So,
It's not so much that there's a new template on how to do that.
It's really just making sure that you're digging into the tools that AI has to offer.
to help you do that because you'd be doing the same thing if you were testing uh outputfrom let's say, paid search on Google versus paid social, right?
(40:52):
You'd come up with a plan.
Hey, this is gonna fit into our strategy.
This is what we're looking at as far as success metrics and what we need to be theoutcome.
And then along the way, what did you do in order to get to that outcome?
So now it's like fitting in the AI tools to be able to figure that out and then figuringout which tools worked
or how you work with them.
(41:13):
Those are the things that not only I've done, but also I communicate with others is makingsure that you start with a plan.
It's really the basics.
It's always the basics.
Getting people on board is really being able to show them, let's say at the start, a goodway is just how much time you're gonna save, right?
So that you have time to do other things.
(41:36):
So start with a plan.
shouldn't just go wing it.
Radical, radical concept.
oh What do you find in terms of the the people piece?
Do you find that there's certain roles or functions that are more predisposed to AIadoption than others or is it across the board?
I think nowadays it's pretty much across the board, but there are definitely standoutswithin a marketing team.
(42:01):
I mean, obviously if you've got copywriters, if you've got designers, those are some thatstand out to begin with, as well as those that are dealing with data, right?
Those are kind of standouts.
But I believe that everyone in a traditional, so to speak, contemporary marketing team isgonna be able to benefit.
(42:22):
And a lot of it is also even just through ideation.
utilizing the tool as essentially being able to do research and again figure out what thesources is, asking for the sources, make sure these are not hallucinations, these are real
information, but it can speed up research a tremendous amount.
And so being able to do that and be able to ideate based on ideas that you have, researchthat you're doing, and then be able to figure out how you can put that in place is pretty
(42:55):
powerful because
you're being able to essentially have a sounding board without necessarily having topresent the information in a typical environment.
So you can be able to iterate in a lot of a in a lot faster manner to be able to get tothat point where you're then like hey this is a pretty good idea here's what it's based on
(43:16):
you know these are solid things I have sources right this is there's no IP issues and thensee where it can go from there.
That research piece is something that um gets overlooked an awful lot and I've been usingdeep research tools since they came out and I often times get looked at um as if I'm some
(43:38):
type of heretic uh when I can produce a deep research brief on a company.
couple that with some established workflows or prompts that I've developed and get a veryspecific targeted output for the client.
uh I sometimes I feel like it's burn the witch, right?
uh What did they just do?
So I'm interested how or since you had surfaced the issue of research and kind of broughtthat around here, I'm interested, how are you advising clients to use research and maybe
(44:06):
um what's a really good example for our listeners about how you have guided someone onusing
using research to get that type of superior output.
Sure.
You know, one way is let's say that you're looking at prospecting to begin with, right?
So kind of more on the lead generation side.
So being able to use tools, some of them are built within CRMs already, some are sittingoutside of CRMs, but being able to dig in and kind of creating a pretty quick model in
(44:36):
terms of identifying prospects, individuals as far as well as businesses, and then beingable to tie in the next steps of that.
So being able to set up sequencing and journeys as an example, as far as communication.
methods, the research that you can do in a lot quicker, more accurate fashion, it'schanged a lot.
(44:57):
And so that is a way to kind of step into that world as far as research is to say, we're abusiness, we sell things, right?
Who are we selling them to?
What is going to motivate them?
What are the triggers?
And how can we more quickly identify decision makers that are at those companies that weneed to make contact with?
(45:20):
What can we learn about those decision makers that is available through research toolsthat we can also then confirm is accurate following the right steps?
Because that's where you can do a lot of customization.
Things aren't as cookie cutter, and then you really develop
into that more model that I was talking about earlier where you may be working yourself orwith business intelligence teams and collaborating with them and being able to create
(45:47):
segmented lists of existing customers as well as potentially prospects at a much higherlevel and then be able to have a much higher return because of it.
Wonderful.
Alright.
Let's I was going to say let's jump in the way back machine, but we don't have to jump inthe way back machine.
It could be more current.
You remember that one?
(46:08):
Yeah.
I'm interested in the biggest lesson that you've learned from a mistake.
in your AI journey, either when you were in that seat or as a consultant, something thatyou'd tell another CEO or CMO to avoid because we've all made some mistakes and I'm sure
there's something that jumps out at you that would help our audience learn from yourexperiences.
(46:32):
So what's the biggest mistake you've made in your AI journey?
Yeah, so I'd say there's probably two things.
So one is, don't believe the hype, right?
Don't get too excited about a tool that says it can do X, Y, and Z until you actually testit out.
That's one.
vendor?
What?
So that's one.
uh And then probably the other is is really taking that deeper dive in terms ofcollaborating with the other teams across the organization so that you really can take a
(47:03):
look into this in a collaborative fashion.
It's not always going to be perfect.
Some of the people you're going to have to kind of pull along so to speak but reallylooking at it from the standpoint of we're in alignment because we're all supporting a
specific strategy.
Everybody's on
board with that strategy.
So getting that conversation going early enough, it's not so much what marketing wants todo this or IT wants to do this, you know, or customer service wants to do this, is what's
(47:32):
the strategy?
Everyone's bought into the strategy, they should have bought into the strategy if they'redoing their job and have had input into that and then have that alignment so that you can
collaborate with them.
towards that end goal.
It's not always going to be perfect.
It doesn't mean that there's not going to be rough steps along the way.
try to create that, try to keep that common goal in mind and continue to surface it sothat you can stay aligned and be able to get to that next point in a much easier fashion
(48:03):
as opposed to fighting it along the way.
You've referenced strategy a number of times in our conversation, and appropriately so.
So this isn't a trick question here.
ah You've referenced strategy appropriately a number of times in our conversation.
And when I go in and I work with organizations on strategy,
Most of the time I am finding that when it comes to AI, they're treating that as atactical component as part of the strategy as opposed to something that's going to reshape
(48:33):
portions of the strategy.
And I firmly believe that we're past the point where that is not possible.
I think it's not only possible, but it's probable for a lot of the AI tools and theofferings to begin to reshape the overall strategy of an organization.
So I'm interested
Have you seen what I have seen in terms of tactical adjustment based on uh AI tools in theevolving landscape there or have you actually seen it begin to tweak and adjust
(49:04):
organizational strategy?
Yeah, I I've seen both.
I mean, I think that there's always this comparison.
It's like AI and everything that AI can do is so sweeping, similar to the start of theweb, right?
Now maybe more so.
But it's interesting to your point because if you kind of think back to kind of the earlydays of people utilizing the web on a regular basis, let's say like the late 90s, right?
(49:32):
A lot of them were using it very tactical, right?
They weren't necessarily using it strategic step.
And I think the same thing is with AI, is that you have people who are looking at it froma utilization of a strategy, even just from a business model.
The folks that are gonna win is you really have to look at it from a clean slate and say,okay, utilizing AI, how can we even look at our business strategy overall and figure out
(49:56):
is it going to work continually in the next 12 months, in the next...
36 months and so forth.
So you have the opportunity now to really start doing research and be able to start doingmodeling to figure out as far as your business is concerned, do you have the right product
set?
Is it the same product set you've had for 20 years?
(50:16):
Are there other products that based on what you do well as an organization, you shouldlook to test and get into that are going to give you the type of return that you need that
are maybe is not as competitive and not as commoditized as other products?
That's a key thing.
to look at that potentially change your strategy as well as you know the overall businessitself.
(50:40):
Getting into new businesses that you haven't been in before.
Yeah.
Well, and I think that's uh
That's a really interesting point.
I've talked about that with an individual too on this uh show recently.
And I think trying to balance the hype against the reality that we're not too far off fromcompanies being able to A-B test company initiatives.
(51:06):
uh Let me rephrase that.
Being able to A-B test actual companies uh versus product offerings uh in a way that we'veA-B tested web pages and product
offerings before.
So yes, I like where you went with that conversation about the strategy because I thinkthere's a whole new vista of opportunity about to open up for organizations here in the
(51:29):
coming couple years.
Yeah, I think there's other things that are very interesting that you talked about aswell, since we're talking about here and now and in the future, that are still challenging
to get your head around, which is companies that are completely running themselvesutilizing AI, where the amount of hands-on work by humans
(51:51):
is so low that it's totally different than what you see today.
That's something to really think about and figure out how you can leverage that.
And again, I'm not saying it from the standpoint of let's replace the human side, butdifferent types of businesses that you can even test out and run without the same level of
(52:11):
investment to be able to figure out if they work.
Because sometimes when you're starting off new businesses, new products, et cetera, acrossthe board, especially in financial services and highly regulated
industries, that can cost a lot of money and it can cost a lot of time.
If you can do something within, so to speak, a walled garden and be able to run it andtest it out, you can save a tremendous amount of time and to your point that can have
(52:34):
potentially a dramatic positive effect over time on the business.
Mm-hmm.
One of the other things that pops up and I've been noodling around recently and I've seenthis out there so I can't claim that this is a wholly original thought um is how we
measure certain types of investment and the types of investment that we're willing to makein organizations.
(52:57):
You made the point earlier in our conversation that the IT department should not be incharge of a lot of this stuff, which I completely agree.
um They have a role to play, but they should not be owning AI deployment, especiallywithin individual uh business units.
And I bring that up because I think right now organizations are very comfortable with theidea of their standard overhead, right?
(53:20):
And I'm just going to pick on payroll.
Payroll is a standard overhead.
I think that this is going to shift to our conversation here from where we're solelyworried about overhead and payroll being a large component of it towards perhaps its
compute.
right, or different types of resources that we're deploying there because we no longerneed to throw people at the problems in order to scale and to achieve our results.
(53:44):
And the strategic implications of that, think, are vast and they're, I wouldn't saythey're above my pay grade, but there's a lot yet to be determined there.
I totally agree with that.
mean, we were talking about the research and being able to put together ideas and put theminto play.
Yeah.
(54:05):
Utilizing the right tools can save a tremendous amount of time with that but then alsoJust the processes that are behind certain things that are running so, know one of the
things that you know, I've seen in the in the past was utilizing essentially like ML toolsalong with other tools to be able to do things that you would think need to be Manual, but
(54:28):
you're seeing them done right before your eyes.
I this is you know, this is 18 months ago So some of it may be already well known
But you know being able to being able to run checks on data and be able to run checks onvarious amounts of information all in an automated fashion across systems all happening at
one time You can drive down, you know the amount of uh
(54:53):
human involvement by tremendous amount and also decrease the cost, increase the efficiencyand allow people to really spend time on the areas that they should be doing, as opposed
to kind of more of the monotonous work.
Mm-hmm.
So one of the amazing things about where we are now several years into the journey um andwe'll be pushing toward our hundredth episode this year on this podcast alone, um which
(55:22):
gives you a sense.
thank you.
Yeah, it'll be exciting.
um So I think one of the exciting things, um but also the terrifying thing about where weare is we've got and we referenced this before.
We've got people on on such wide ends of the spectrum.
So right now we were just talking about machine learning and leveraging a lot of this and
I referenced compute versus payroll, yet I know we've got listeners, maybe it's theirfirst time, if so, welcome uh tuning in right now.
(55:48):
And they've just been dabbling, right?
So if we were going to bring this back around to the dabbler, to the person, to the leaderthat's just getting in right now, what type of minimum investments should they be thinking
about to begin exploring this?
And what should they be doing to begin kicking this off in their organization?
Because the best time to start was yesterday, candidly.
(56:09):
ah The second best time is this Monday morning.
Right.
Yeah, absolutely.
What I would share is that one is taking a look at your business and then being able toidentify what tools you want to test with it.
If you've got a circle within your industry that you can tap into in terms of what theyhave used, the learnings from that, both bad and good, can help speed up the process.
(56:38):
And then once you're looking at, what can I use as, so to speak, like a proof of concept,I think is an important place to start with.
So making sure that you're identifying, let's say like three, you know, high leverage usecases, tying them of course to performance, you know, from a revenue standpoint, from a
cost standpoint, making sure, you know, who is actually owning that, you're measuring it,and then making sure that based on what you're measuring, you're able to proceed in, you
(57:09):
know, one or many directions.
And at the same time, you're still building for yourself and your organization, yourliteracy as far as AI is concerned.
And that comes by doing.
It does not come by watching.
It's one of these things that from the CEO level, you know, all the way across anorganization that people have to, people have to really dig in and learn.
(57:33):
And there's a lot of different resources for people to learn how the systems work.
know, LinkedIn learning is a great source.
There's also Google as well as many others that can give you perhaps a little bit moreobjective feel than by starting off with some of the tools directly.
And so you're also learning from people who are more knowledgeable.
(57:56):
I kind of hesitate with the expert side, but are more knowledgeable than maybe where yousit is a good place to start so that you really can dig in.
But you got to dig in now because again, so many people are utilizing AI really just as asearch engine.
and it's so exponentially beyond that, it's really difficult to even talk about those twothings in the same breath.
(58:20):
I think that that is a fantastic place for us to conclude our conversation today.
An hour goes fast.
And uh if I was going to pull three things out from our conversation today, it would be bebrave, right?
Dig in now and find the air quote experts that can help you with your journey because theyare out there even if we are loathe to call them and ourselves experts from time to time.
(58:46):
So.
If people want to take advantage of your expertise, Andar, where can they find you?
What's the best way to reach you and get in touch?
Absolutely, thank you for that.
So on LinkedIn, so Ondartarlo, very easy to find, unique name, all over LinkedIn.
uh In addition to LinkedIn, also you can go to my website, Ondartarlo.org, and you canfind me there, you can send me a request, and I'll be able to get back in touch with you.
(59:15):
Wonderful.
I appreciate it.
I really enjoyed our conversation today and I appreciate you coming on and sharing yourknowledge and experience with our audience.
Thank you very much.
Thank you, Chen.
It's a pleasure.
All right.
And thank you to our listeners for once again tuning into AI for the C-suite.
If this episode was useful, subscribe wherever you get your podcasts, follow us onLinkedIn and all the socials, just not Twitter, X, whatever they call it these days, I
(59:39):
don't know.
And check out AI for the C-suite dot com.
Until next time, keep your algorithms running, your leadership evolving and your AI incheck.
Take care, everybody.
Popular Podcasts
Betrayal Weekly
Betrayal Weekly is back for a new season. Every Thursday, Betrayal Weekly shares first-hand accounts of broken trust, shocking deceptions, and the trail of destruction they leave behind. Hosted by Andrea Gunning, this weekly ongoing series digs into real-life stories of betrayal and the aftermath. From stories of double lives to dark discoveries, these are cautionary tales and accounts of resilience against all odds. From the producers of the critically acclaimed Betrayal series, Betrayal Weekly drops new episodes every Thursday. If you would like to share your story, you can reach out to the Betrayal Team by emailing them at betrayalpod@gmail.com and follow us on Instagram at @betrayalpod and @glasspodcasts. Please join our Substack for additional exclusive content, curated book recommendations, and community discussions. Sign up FREE by clicking this link Beyond Betrayal Substack. Join our community dedicated to truth, resilience, and healing. Your voice matters! Be a part of our Betrayal journey on Substack.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by Audiochuck Media Company.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.