August 6, 2024 • 50 mins
Welcome to another episode of How I Met Your Data! In this episode, hosts Sandy Estrada and Anjali Bansal are thrilled to welcome Malcolm Hawker, a seasoned expert in data and analytics with over 25 years of experience. Currently the Chief Data Officer at Prophecy, Malcolm shares his insights on master data management and data governance.
The episode kicks off with a discussion about the recent CDOIQ conference in Boston, where Anjali and Malcolm co-hosted a session. They delve into the topics covered during their session, emphasizing the need for transforming data culture within organizations. Malcolm highlights the importance of delivering value and shifting mindsets to foster a positive data culture.
Listeners will gain valuable insights on how to prioritize collaboration, continuous engagement, and leveraging product management principles in data leadership. Malcolm also shares his thoughts on the future of data management, including the potential of data fabrics and governance as a service.
Don't miss this engaging conversation filled with practical advice and forward-thinking perspectives on the evolving landscape of data and analytics.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Hi. Welcome to another episode of How I Met Your Data. Sandy Estrada here.
Myself and my co-host Anjali Bansal are thrilled to have Malcolm Hawker with us on this episode.
Malcolm has over 25 years of experience in data and analytics.
He is a recognized thought leader and advisor in all things data,
but particularly master data management and data governance.
(00:22):
He has played a pivotal role in helping some of the world's largest businesses
improve their enterprise information management strategies.
He is currently the chief data officer for Prophecy, where he leads a global
field of master data management and data governance practices at the firm.
He also hosts the CDO Matters podcast, where he interviews senior data leaders
from various industries and domains.
(00:43):
Now, if any of you have ever heard Malcolm Hawker speak, you would know he's
an extremely passionate data leader.
We dive into a lot lot of different topics. But first and foremost,
this episode is recorded right after Anjali and Malcolm returned from their
CDOIQ conference in Boston, where they co-hosted a session.
And we're going to talk about that session. We're also going to talk about a
(01:05):
number of other topics that I think were little pet peeves of Malcolm and mine and Anjali's.
So we get into that discussion as well.
He's got a lot of experience, so
it's great just having these types of conversations with someone like him.
So without further ado, let's get the show going.
Music.
(01:40):
We'll be right back. so welcome malcolm and anjali fresh from cdo iq so anjali,
How did you end up at CDOIQ with Malcolm?
I had met Malcolm last year through a number of different conferences that we
just happened to be at. At one point, we were on a closing panel.
(02:01):
I think we probably agreed and disagreed much at the same time.
We just struck up a friendship, tended to run into each other at other conferences
as well. And one of the things that Malcolm had kind of offered as advice to
me a number of times was get to CDOIQ.
You're going to meet a lot of great data leaders.
It's in Boston, easy to get to. And so kind of at the beginning of the year,
(02:23):
when CDOIQ put out their call for speakers,
I ended up putting in a paper that was actually inspired by one of the conversations
Malcolm and I had had in London about data culture. So I put this paper in.
I had initially envisioned it as a panel.
As it got approved, I went, you know, we've done panels before.
(02:43):
They're met with, you know, kind of different levels of success.
So I wanted to do something a little bit differently and kind of couldn't get
away from this good cop, bad cop idea. yet.
And so just kind of thinking about that, I was like, you know who would be a really good bad cop?
Because they are very, you know, very passionate about telling you what is wrong
(03:07):
with the world today, but also offering,
not just telling you what's wrong, but also offering you some tangible outcomes to strive towards.
So, you know, so I, you know, we ran into Malcolm at Data Universe.
I had a Canadian chocolate bar with me specifically for Malcolm.
And I made my ask. I said, hey, I had this talk that's been approved.
(03:32):
It was actually inspired by our conversation.
And would you like to partner with me on this and kind of do something a little bit different, but fun?
And he said yes, which was a little surprising knowing that he actually ended
I ended up presenting Monday through Thursday.
Our talk was on the last day. So I'm like, oh my God, by the time he got to
Thursday, how his voice wasn't completely hoarse and gone is beyond me.
(03:56):
Now, Malcolm, when she made the pitch to you, did she inform you that you had to play the bad cop?
Or is that a surprise? It was inferred. Okay.
She she told she told me about the format and
kind of what we were going for right you know i don't know
good cop bad cop but more like buddy movie maybe but
but either way i mean i'd kind of assume when it when
(04:18):
it came to some of these things that it would would it be inferred that
i would be the provocateur and that that angela would be would be the ambassador
because because she is so i just kind of naturally figured that out and it's
a role that i relish in that i i am I'm happy to embody at conferences or anywhere else because I'm,
(04:39):
as said, I am passionate.
Nobody's ever going to accuse me of not being passionate.
And I think we can do a lot better as data leaders.
And I may not have all the answers, but I do know that the things we've been
trying just aren't working that well.
So anytime I have an opportunity to kind of poke a little bit and to talk about
(05:00):
doing things differently, I pounce. So thank you for including me.
Oh, no, thank you for joining me. It was fun. It was a really fun talk.
Yeah. And I will say we were on the last day of the conference where you can
kind of expect a mixed set of results in terms of participation, in terms of audience.
(05:21):
We actually had a lot of people join us. We did. It was good.
Yeah. And they were engaged. One of the things that I really enjoyed about our
format was the fact that we were able to leverage the conference's app to poll the audience.
So not only poll the folks that were in the room, but there's a virtual component
to the conference as well.
(05:42):
And the virtual attendees could also kind of live enter in responses.
Oh, that's fantastic. Yeah, I mean, there were 3x the number of virtual attendees
that there were actual attendees, which I think is great.
Kind of, you know, it's broadened the tent. And that Hova, Hova,
Hova, whatever, W-H-O-V-A app is really good.
(06:05):
I think it's actually one of the best. Usually at these conferences,
when the conference organizer makes me download the app, I'm like, oh, God, another app.
I need another app. But that one actually has utility. It works and it's just easy.
So yeah, I was glad that we had the poll. I'm glad that we had really good turnout.
The topic was good. The discussion was good. It was awesome.
(06:29):
So let's cover the topic. What was the topic? So the title of the talk was Transforming
Your Organization to Embed DNA in Your DNA.
So, you know, as we kind of started shaping the content, it was really talking
about how data leadership has been kind of operating in the same mode for the last decade plus.
(06:53):
And things haven't really changed. We haven't really moved the needle.
We've got like great opportunities ahead of us, but we're not really capitalizing
on them in the way that we would have expected or hoped or, you know,
have the potential to do.
And so part of that or a big part of that is really rooted in the data leadership
and the culture that exists in the organization.
(07:13):
Yeah, I mean, a major takeaway of our presentation was that this idea that culture
is not a deliverable, right? Culture is actually an outcome.
If you deliver value, if you embody leadership characteristics,
and this is something that we stress in the presentation as well,
if you kind of embody these four attributes of a good leader,
(07:36):
of a good data leader, if you do those things, if you deliver value to your customer,
then culture will follow. it necessarily has to follow.
So the idea that culture is just some sort of deliverable, like a checkbox on
a requirements doc, right?
It's like, okay, well, I got to do this. I got to do this. I got to deliver data culture.
(07:57):
And then when I do those things, I can deliver value to my customers.
I used to hear this as a Gartner analyst all the time.
We see it in the actual data. We see it in Gartner's CDO survey where CDOs are
asked, what are your biggest impediments to delivering value to your organization?
What are your biggest roadblocks? And the answer consistently,
(08:18):
top three, one of the top three is always there's a lack of a data culture.
And to me, that perspective is completely upside down, right?
The idea that it's a deliverable, that's kind of upside down because you're
talking about a three, four, five, 10-year journey, an ongoing journey that
never really kind of ends.
As Anjali rightly said in our presentations, like who isn't using data? We're all using data.
(08:41):
Everybody uses data all day, every day. So even the idea that there isn't a
culture that wants to use data or wants to be fact driven is probably untrue.
You kind of put all these things together.
And the main takeaway, the main thing that I was trying to say is like,
hey, double down on value, find ways to deliver value, then the culture will follow you.
But any idea where culture is some sort of dependency is probably hindering
(09:06):
you more than it's helping you. Hmm.
That's a really good and thoughtful point out. It's so thoughtful because the
reality is it's not, you know, it's not something that you go create a team around and focus on.
And I think they're also, there's a lot of competing factors in terms of conflating
culture with data literacy as well.
I've seen that a lot where, you know, data, what they really mean is data literacy,
(09:30):
but they're saying it's culture.
Those are two very different things. And I do find that people People conflate
those two sometimes as well.
So you're absolutely right. You deliver value, people will become more literate
because you're delivering value that they want to leverage.
Therefore, one begets the other. Yeah, this idea of culture as a dependency,
I think, has some rather perverting effects in what we do as leaders.
(09:55):
A focus on data literacy, I think, is one of them.
To me a focus on literacy without focusing
on the delivery of value with without focusing on having a really easy to use
product without focusing on user-centric design principles without doing the
things that we know are necessary to build great products that people get value
(10:16):
from if you're just focused on literacy because you see it as as a dependent
action to drive the data culture,
and you're not doing those other things, well, you're going to alienate the
very customers that you're there to serve because you're basically telling them,
hey, you need to figure out all this stuff.
You need to up your game. You need to do all of these things.
And what do I need to do as a data leader?
(10:37):
Well, I need you to get you to see the value of data.
And again, I would see this all the time when I was an analyst.
I would ask my customers, okay, well, what are you doing around this idea of
data product management?
Or what are you doing to drive value? What are you doing to measure value?
Practically nothing is usually the answer in that case. You know,
(10:57):
what are you doing to understand your customer needs?
What sort of feedback mechanisms do you have with your customers around their
dating? And a lot of the answers were very, very, very light.
And then I would ask, okay, what else are you doing? Well, I'm focusing on data
literacy as a means to deliver on my data culture mandate.
Oh my goodness. I think you're approaching it backwards.
Yeah. Yeah, absolutely. So what was the response in the room or even online
(11:21):
related to that messaging?
You know, there was a lot of it. Like I said earlier, there's a lot of engagement, right?
There's a lot of head nodding. I think that the message very much did resonate
because we had a number of folks come up to us after the talk was completed
with comments, with questions, with their own reflections.
(11:43):
So, you know, I think the message really landed and
gave people give people food for thought yeah i
i agree that that that tends to
be the response to a lot of the things that i share
because often i'm sharing perspectives that
i i don't want this to sound egocentric but but
i'm sharing perspectives often that i believe they probably
(12:04):
haven't heard before right the like the idea that
data literacy is is problematic right the
idea that you're you're you're forcing people to get
trained on things they may not like using
right right or have difficulty using or
lack confidence to use the idea that you're forcing them to get training on
(12:25):
that could be doing more harm than good like just that that as as a as an assertion
for a lot of people would be like what do you mean data literacy is good everybody's
telling me data literacy is good it has to be good because everybody's telling
me it's good gardener's is telling me it's good.
My consultants are telling me it's good. Everybody's saying it's good.
What do you mean it's bad?
So in a lot of the presentations that I give, there's a time needed for people
(12:50):
to kind of absorb the message.
But as Angeli said, there were many that had said, hey, listen,
this is great. I really appreciate it. So I think all in the response was actually quite good.
Because on the flip side, I didn't hear anybody tell me, and maybe they're just
avoiding conflict, I don't know. I didn't hear anybody tell me,
hey, you're crazy, right?
Like, you know, go back to the drawing board because you completely missed the
(13:12):
mark. I didn't hear any of that. So that's good.
I want to dig into what you said because I haven't heard anybody say that before.
And I have, over the last few years, I've kind of jumped on that bandwagon as well.
You mentioned the whole fact that you're trying to train somebody on something
that may not be ready to use or they may not like to use or they're just never
(13:36):
going to use is a problem in itself.
That message is, I feel, is new.
I feel a lot of people still have the blinders on in terms of,
I need to do this fancy dashboard and that's how I'm going to deliver the information
versus understanding who the user is.
(13:57):
What would you say is kind of a tactic there in terms of helping individuals hear that message, one?
And two, what should they do with that message? Well, to me,
the overarching tactic that we as data leaders need to be thinking more about,
I will broadly encompass in the in the notion of integrating product management
(14:18):
as a discipline into data management, right? Like that, to me,
that's the broader wrapper.
And if we did that, if we hired product managers,
if we were rabidly focused on understanding how our customers and I use that
word purposefully, customers, not users, not stakeholders, not business.
Business people, customers.
(14:39):
If we hired product managers who all they did all day, every day was understand
customer needs, understand how customers were using our products,
literally standing over people's shoulders, watching them use a dashboard.
What are they clicking on? What are they not clicking on? Where are they getting,
where are they stopping?
Those types of those types of activities if
we're doing that if we were polling our customers asking them
(15:00):
are you getting value from this even even
going so far as is having conversations maybe even around pricing what would
you be willing to pay for this if you had to pay for it right like think of
this as a real product right if we're serious about data products products have
price tags the idea that you can walk into a retail store and have everything
on the shelf have no price tag on it, that means it's not a product.
(15:22):
It's kind of like saying I play poker and then asking, okay,
well, what do you usually wager? Well, we don't wager anything. Well, it's not poker.
You got to be betting to play poker. And if you're talking about products,
there's got to be a price tag.
So it's these types of things, Sandy, that we need to be focused on.
And if we do all of those things, we understand what customers want,
what they need, what their challenges are, how data is used to overcome those challenges.
(15:48):
Well, that's a very different enterprise than saying, okay, I've got some training
that you're going to need to get on our data products in order for you to get value out of it.
I would argue if you do all those other things and take a product management
approach, you're going to build products that people want to use.
Will there be a training aspect to it? All products have some go-to-market function. They all do.
(16:10):
Training, marketing, there's all sorts of things that we need to do as product
people to make sure that the products are successful, including lifecycle management,
archiving, right, sunsetting products.
But if we do all those things, maybe there's some training required,
maybe, but it won't be mandatory necessarily, or it will be received very positively
(16:31):
because the product will be seen as something that is driving value to the organization.
So it's a totally different mindset. It's a different model.
That's great. That's great. Love it.
So how long was this session? Was it your standard bare 20 minutes or was it
like a 45 minutes? 45 minutes. Yeah.
So what other ahas did you have on this 45 minute journey?
(16:53):
Let's see. I mean, we dug into just a couple of the areas of failure, right?
And one of them that we continue to hear, especially from a governance perspective,
is our data is garbage. It's, you know, of such poor quality.
And, you know, that really is something that we, you know, we kind of anchored
in and said, that's not the right message.
(17:15):
You know, like, if we really want to drive a culture of data within your organization,
we need to assume positive intent, you know, in terms of your data,
people's behavior. Nobody's trying to make your life's difficult.
So if you're going around saying that your data is garbage, what message are
you communicating to your people?
Like look at that as an opportunity to either think about your data differently,
(17:38):
maybe look at the context of your data so you can allow for two things to be true at the same time.
But like really take on that model of behavior around positive intent and collaborate
with your people to understand what's most important to them.
I like that. Flip the script. Yeah.
Yeah, I love it. I mean, this kind of traces back to this idea of seeing culture
(18:02):
as a deliverable, having perverting effects on what we do as data leaders.
Positioning data quality as a burden and not an opportunity is a classic example of that.
When I was at Gartner and I had conversations around culture day in and day
out, What I came to see often was that when people said there's a lack of a data culture,
(18:25):
what they really meant is my customers aren't doing what I want them to do.
Literally. They're not going to governance committees. They're not owning data,
whatever the heck that means.
They don't care about data quality. All they care about is their business processes
and they don't care about my data. So I'm going to make them care about my data.
(18:47):
And when they don't, I'm going to express my frustration using words like garbage in, garbage out.
You can lead a horse to water. You can't make them drink.
They just don't get it. They just don't get it, Sandy. They don't see the value of data.
So when we approach everything like, okay, I'm going to make them see the value of data.
(19:12):
I'm going to make them understand why it's important to understand systems and
processes sitting underneath the data.
When I'm going to tell them over and over again that they're making my job harder
and harder and harder, out of a frustration because I can't get them to get it.
I can't get them to see the value. I can't get them to go to governance committee
(19:32):
meetings. And so I'm frustrated.
And I'm going to use words like, you need to own.
And you need to stop throwing garbage over the fence.
I just see it just add all this stuff up. And it's like, wow,
we're really disempowering ourselves here.
We're really disempowering ourselves. If we think that the pithy metaphors or
(19:53):
analogs that I use here is like,
just imagine if the chief operating officer of Frito-Lay said their corn was
garbage all the time, or if the
COO of Anheuser-Busch said their water was garbage all the time, right?
That's one end of the spectrum. The other end of the spectrum is we're sitting
on this mountain of unrefined gold.
We're sitting on a mountain of unrefined gold that everybody, everybody,
(20:16):
it doesn't matter, everybody is saying is the new gold and everybody is saying
is going to be the thing that helps my company deliver transformative value
and I'm responsible for that?
Holy cow, what an opportunity. This is incredible.
Instead, it's garbage in, it's garbage out. What can you do?
(20:36):
What can you do? Yeah, they dropped the responsibility and handed it off to somebody else.
You need to fix this. This is your doing and I can't run bad data.
The trope that they don't understand blows my mind because the reality of the
situation is every single operational team on the planet is knee deep in Excel,
(21:00):
punching data every day. That is what they're doing.
They all have analysts. They're all looking at data all the time.
They understand the value of it.
So it's not a matter of understanding the value. It's a matter of helping them
understand the process that is required to refine and ensure that value is there.
But it's even in the words, it's the perspectives, but it's also the words that we choose to use.
(21:21):
And it's the way we choose to express things. I'll give you an example.
There's soundbites, plenty of soundbites floating around saying that 80% of
data scientists' time is lost due to data quality issues.
Issues like what a ridiculous statement but we hold on to those things and we
use them we we use them to validate the fact that our jobs are so hard so hard
(21:45):
and we're and we're and we're inefficient,
and we're not as efficient as we could be and we're not driving as much
value as we could be because you make it hard because look at
this stat 80 of a data scientist time just gets
thrown out the window do you know how much we're paying for those data scientists and you're
making them them wrangled all of this data while the fact remains
the data is of different structure different
(22:06):
formats different definitions by definition by
design crm systems function differently than erp systems by design and if you
want to blame somebody blame henry ford don't blame the people that actually
coded those those applications right because they're they're i mean it's how
it is by design these things are different because our business processes are different.
(22:30):
The applications mirror the business process.
The data mirrors the application. So that's it. And it's not some sort of flaw
in the system. It's actually by design.
Sorry. See, I'm passionate. I told you I was passionate. No, I love that.
All right. So we have deliver value, change your language. Is there a third one?
(22:53):
So, I mean, we went with, you know, create the right mindset,
prioritize collaboration.
It was continuous engagement. So, you know, instead of sitting in the ivory
tower saying you must do things this way, work with your people.
Like get out there, get in the muck and actually understand what people are
doing. What are truly their pain points?
(23:14):
What do they need in order to be successful and thereby make the organization
successful, deliver that value.
And then we can start to build the right products for that. Malcolm had a great
example around a baker baking bread.
And they're the only bakery in town and nobody's buying the bread.
And the baker's standing there going, well, you just don't understand the value
(23:36):
of my bread. You don't understand how hard it is to make the bread.
Well, maybe they're all celiac and they can't eat the gluten.
Did you ask? That's a really good metaphor.
Yeah, yeah. To me, the shifts here are actually kind of subtle, right?
Like the really good news here is that, you know, we're not necessarily talking
(23:59):
about fundamental shifts in technology.
It's not like we need to like pull out all the technology and put new technology in, right?
I think a lot, we can make a lot of positive differences here with just subtle
shifts in how we manage our team. subtle shifts in how we approach problems.
There are already kind of roles that kind of align to this idea of product management,
(24:21):
although I would hire a product manager because it is a unique skill set.
The idea that a business analyst could be a product manager,
but the roles are pretty close, right?
The idea of I need to understand and document requirements, I mean, that's fairly close.
I would argue from the perspective of what do you need to do?
And Angela just touched on some some of the characteristics that you need to
(24:42):
embody. When it comes to brass tacks and like, what does the organization look
like? Who do I need to hire?
What I would recommend is one, a product manager, two is something called the value engineer.
I would argue that at a certain size of company, well, this is all sizes,
but it depends whether it's an FTE or not.
But at a certain size of company in a data analytics function,
(25:02):
you could easily justify an FTE to be a value engineer consigliere,
to be the right-hand person of a VP of data analytics to help do things like
budgeting and forecasting and planning.
You're probably already doing that already, but take that role one step further
and actually get it into the realm of starting to understand and build models
(25:23):
of how does improving data quality actually drive business outcomes?
How do our key business drivers, how are they influenced by better data?
Start measuring that thing out.
With a north star of getting to the point where your store of data products,
actually each product has a price tag on it.
Now, we could get into interesting conversations about, well,
(25:45):
data is not actually a balance sheet asset and fall down that rabbit hole if
we wanted to. I don't think that that's a useful conversation.
I'm talking really about a conceptual framework here, some sort of thought exercise
maybe even to say, if you had to put price tags on everything, could you?
That's different than saying, well, my CFO doesn't think it's a balance sheet
item, so that means I'm just going to throw up my hands and not care.
(26:06):
Yeah. No, and you only improve what you track, right? So if you're tracking
the value of it, then you could improve on it.
So yeah, that makes absolute sense.
I think I think it's challenging when it's considered a cost group.
It becomes a very challenging issue for a VP of data, for example, or a CDO.
Speaking of which, I heard that you got a new role at Prophecy. Congratulations.
(26:31):
Thank you. Yeah. CDO of Prophecy. Are you working to implement some of these ideas there?
My role is an externally facing one.
So I would best be called a field CDO. There are a number of companies that
do what we do, including Microsoft as a CDO who's more externally facing.
(26:51):
So my role, my daily responsibilities largely have not changed.
I support our clients, I support our prospects, I support our field,
and I evangelize in the market, right?
And I do things like this to help other CDOs understand how to best optimize
their organizations, how to structure their organizations, how to approach governance,
management, you name it, from having done this for 30 years.
(27:13):
So new title, largely same job.
Now that said, from the perspective of implementing what I suggest and following
my recommendations, I do come at my knowledge honestly, right?
I have managed data and analytics groups.
I've managed governance efforts. I've implemented MD. I have been the person
(27:34):
making decisions about budgets, about org structures, about all of these things.
I feel like I could live on either side. I know I could live on either side
of the equation, whether I was internally facing only or externally facing.
But given my passion, given my knowledge, given what appears to be rather kind
of decent skills for sharing knowledge, if this dog is learning some old,
(27:56):
this old dog is learning some new tricks.
I actually think I'm best suited.
I think I'm, from a value extraction perspective, doing what I'm doing and the
role that I've got, I think is a perfect word for me. So am I hearing like a
2025 Studio IQ, a year as a studio reflection piece?
Oh, that sounds great. These stories write themselves. I mean,
(28:17):
they write themselves. I think that would be great.
Maybe put my money where my mouth is and measure my own value.
We can frame it up and create the right mindset, prioritize collaboration.
Continuous engagement.
And what was the last one? Drive change.
So yes. Yes. I'm all about driving change. we desperately need change.
(28:38):
So I find it interesting, just a fun little aside, that in my presentation,
my individual presentation, not the one that I did with my learned colleague
here, I included a graphic of the Albert Einstein doing the same thing over
and over again, expecting different results. I included a graphic like that.
And the very next presentation I went to was my friend Juan Cicada from Data.World,
(29:00):
and he had the exact same graphic in his presentation.
So I was like, Oh, I wonder. Yeah. And we had, and we had Angela,
we had gone to dinner the night before with Juan having this conversation.
And he said, I'm going to alter my presentation because of the things that we just talked about.
I didn't actually get confirmation from him that, that he added that graphic
because of our conversation, but who knows?
I had an early, I had an early preview of his, of his deck that was already in there.
(29:25):
So he was already thinking about that. Sure.
The clock track around it was altered based on some of the things that we were
talking about. At some point.
Do we, as data advisors, at the end of the day, we're all advisors.
We're helping executives.
We are. At one point, it feels like an uphill battle, doesn't it?
(29:47):
With those messages, like, hey, you guys keep doing the same things,
expecting different results. We keep telling you the same things.
It feels daunting, doesn't it? It does, and it doesn't.
There's a consulting answer for you. Define daunting. I mean,
from a top-down perspective, if you look at this from a global kind of macro
(30:09):
perspective, yeah, it's daunting.
But if you take more of a bottoms-up, iterative, kind of agile approach and
just take some baby steps, and if a lot of people started to do them,
I think it would be less daunting.
But to your point, Sandy, I mean, yes, we live in a very complex ecosystem of
reinforcing behaviors that are reinforcing that where a lot of the players in
(30:33):
the market are reinforcing negative behaviors.
They're not enforcing positive behaviors.
I actually came up with a little graphic and a little theory to explain all
of this. I called it, what did I call it?
The semantic pedanticism feedback loop.
The semantic pedanticism feedback loop. It was a riff on why we keep inventing
(30:56):
new words for things, because we do.
However, it actually explains, when I look at this model that I built,
and now I'm going into crazy town, this complex interaction between software vendors, consultants,
and analysts, and customers, and users, people signing the SOWs.
(31:18):
There's a complex web that exists there where once an idea is birthed,
once something is created. I'll pick on data literacy.
Data literacy starts to bubble up, I would argue, six, seven years ago.
It's bubbling a little bit.
(31:39):
Then in 2019, Jordan Morrow, while he was working for Click,
commissioned a study. It's a vendor commissioned study.
Take that with a grain of salt. All vendors do studies to say,
hey, generally those studies already have the answers and they're just looking
for the study to validate the answers.
But that aside, that aside, Jordan commissions a study for Click that finds,
(32:02):
in essence, that finds that people aren't getting value from data because they
don't understand the data.
The underlying premise is, is what Jordan would say, is there's a skills gap,
a skills gap on users when it comes to data and the way that we deliver value is to close the gap.
This was the assertion coming out of this. So this report gets a little bit of steam.
(32:26):
People at Gartner start to hold onto this and start to smell something here.
Well, interestingly, analyst firms, I won't pick on Gartner,
I will just say analyst firms.
These are annual subscription businesses that need people to re-up every year.
And if everything was evergreen, if everything we did didn't really change that
(32:48):
much, governance isn't changing that much, MDM isn't changing that much,
much. Data quality isn't changing that much.
If things didn't change that much, those annual subscriptions would be,
I would argue, less compelling because you don't need to re-up your annual subscription.
How do you get people to re-up an annual subscription? Well, you create new things.
You make new things. Hey, flash, flash, flash.
(33:08):
Hey, there's this new thing. Have you heard of it? It's the data mesh,
or it's the data fabric, or it's data literacy.
And there's all these things that you need to learn about. out.
And they start pushing and pushing and pushing and pushing. And then all of
a sudden, people start calling asking, hey, what's this data literacy thing? Tell me more.
And then Gartner has data to show, and analyst firms have data to show that
(33:31):
people really care about data literacy.
It becomes this self-fulfilling prophecy that all of a sudden,
vendors jump into the fray and say, oh, wow, this thing must be big.
People are asking about data literacy.
Consultants jump into the fray because their customers are saying,
hey, tell Hey, Sandy, tell me about data literacy. I'm hearing this data literacy thing.
And all of a sudden, this gets spun up. It's the hype cycle, right?
(33:54):
It's the hype cycle. In the case of data literacy, I would argue it came from
one study from one vendor where all it took was a few analysts to jump on that pile.
And lo and behold, a movement is born.
I mean, that's usually the case, either a study or a white paper from somebody.
I mean, data mesh blew up overnight, it felt like. And people just took the
(34:17):
highlights of it, right?
It just, yeah, it's one thing after another like that, for sure.
But your question was, hey, you're looking at things a whole different way.
You're suggesting that we need to dispense of some sacred cows.
If true, that seems daunting. And I would say, yes, it is daunting because there's
these other forces that continue to reinforce things that, in the case of data mesh, my goodness.
(34:44):
Did anybody stop to ask, wow, okay, conceptually, this is interesting.
Full decentralization, that's interesting. Having domain autonomy,
that seems interesting.
Federated computational governance, I don't know what that means,
but it sounds interesting.
All of these things, okay, this data products, ooh, that really sounds interesting
because nobody's using my products today.
(35:04):
And it seems like this thing can maybe help with that. So all these things are interesting.
But did anybody ask the question, it was like, okay, wow, wow,
when you go from a hub and spoke to a, to a spaghetti bowl, what is the cost?
Right. And did anybody ask, wow, well, hub and spokes as inefficient as they are.
And sometimes, you know, maybe not the best for our end users and maybe not
(35:26):
the best. And maybe, maybe we have to sacrifice some, some things at the domain
level, you know, airlines are a hub and spoke for a reason.
Network typologies are hub and spoke for a reason, but did anybody stop to ask,
okay, well, what would the cost be to go from that to something?
It is this basically the spaghetti bowl, like a spider web of connection,
because that's what it advocates.
It advocates domain to domain, this organic domain to domain sharing,
(35:49):
where it's quite okay to have the same report repeated 15 times over,
as long as you're sharing between domains. Did anybody ask?
No, nobody asked. But the hype gets created, everybody pumps it up.
Three years later, it's the biggest thing and the greatest thing since sliced bread.
And I'm talking to companies that are like, yeah, we just turned off our data
(36:10):
warehouse. I was like, what do you mean?
Yeah, we're following the mesh. We're implementing a data mesh.
We turned off our data warehouse.
Do you want, okay. How are you going to support cross-functional use cases?
Well, we're figuring that out.
How are you going to give the CFO the one report that says how many customers we have?
Anyway, sorry, I'm ranting. But we live in a complex ecosystem where contrarian
(36:34):
ideas don't tend to get inflated.
The ideas that reinforce the status quo tend to get inflated or hypey things that are unproven.
Hypey things. Hypey things. What's new?
And I fall prey to that. I was going to say, one of the other things that we
touched on a little bit during the presentation was the fact that just as consultants,
a lot of times our clients are coming to us saying, how do we benchmark against
(36:57):
our competitors? competitors, right?
Like how are we performing against everybody else?
And I just sit there going, why do you want to know?
Like, let's deal with your problems and figure out like what we need to do to
get you moving forward as opposed to comparing yourself to somebody else.
Because guess what? They have the same issues, right? Like we're all solving it together.
They're no further ahead than you are. They're not doing anything differently.
(37:21):
So why do you want to know? you know and is it
just that safety in numbers is it just therapy to know that maybe others are
suffering in the same way it just it's kind of an interesting interesting quandary
in terms of like why do you want to know and like like take a risk like let's
just do something differently to move forward if i had a dollar for every time
(37:42):
i was asked when i was an analyst.
What does good look like or where do i stack up in the industry i'd be a rich
man even if they were Canadian dollars, I would be a rich man.
So completely concur, Adelaide. I mean, it's, it's, but I see that again,
getting back to this idea of mindset and how we approach what we do,
(38:02):
I see that being a fairly defensive posture.
Meaning if I'm trying to define a strategy, or if I'm trying to execute on something,
and I take the perspective of, I need to do just slightly better than the other guy or gal, right?
And if a good enough is good enough, and if I can check that box and say that
Gartner has approved this plan, or that I know that I'm doing things the same
(38:27):
as everybody else, that's a defensive posture.
Because it's not innovative. It's heard. H-E-R-D.
Right? Instead, to your point, well, okay, do you want to be in the herd?
I think some people want to go down proven pastures, right?
They want to walk the path others have walked.
They want to know that if they take those steps, it's going to work for them,
(38:51):
which isn't always the case.
Nike's culture is completely different than an older company,
an older manufacturing firm.
So it's hard for, I do not like those conversations. organizations and we have the data, right?
I mean, we work for one of the largest consulting firms.
We have the data, but it hurts. It hurts to present that because no matter what
(39:13):
we say, there are so many other dimensions to the success of that factor that
may not be present at that organization.
And they don't have control over some of those things, whether it's budget,
you know, how the organization's outlined, like structured, et cetera.
There's things there you you don't have control over that are outside of the
(39:33):
scope of your domain. So let's stay focused on what you can control.
I mean, those are hard messages to deliver because people are not ready to listen to those.
And maybe just kind of following the pack is good enough, right?
Maybe that's okay, but that's not the mandate most CDOs have been given over
the last three years, right? I think the last data that I saw was nearly 30%
(39:57):
have some sort of change mandate.
Right? So if you've been given a change mandate, if you're in charge of a digital
transformation or a digital acceleration, then you necessarily can't be following the herd.
I mean, Nike, what a great example, right? And there are others out there,
Procter & Gamble, McDonald's, Nike, it goes from 15% direct-to-consumer sales
(40:18):
to over 50% direct-to-consumer sales in five years.
That's crazy, right? And in the middle of a pandemic, Right.
So there are examples out there of companies doing this stuff and getting and getting it right.
But I want to, at least for me, how I approach this is I do want to deliver change.
(40:41):
I do want to deliver transformation because there's untapped value in our pile
of unmined gold or unrefined oil. It's there.
It's absolutely there. And I don't ascribe to the herd mentality,
but maybe for your organization, it's okay.
Maybe it's good enough. And I think for some, that may be okay.
But I think for many others, it's not. Depends where they are, for sure.
(41:05):
The question I need to ask you, Malcolm, since we have you for another few minutes,
Gen AI, obviously, forefront of everybody's minds.
Let's skip that for a second, because that was the next new thing. It's been here.
We're all trying to deal with it, good or bad or ugly. What do you think is the thing after?
I think I know your answer, but I am very curious where you're going to take
(41:26):
that. Oh my gosh. Well, what's after?
Depends on how after you mean by after, because there's an after out there that frankly perturbs me.
I read a book recently titled Our Last Invention by a guy named James Barrett, B-A-R-R-A-T.
(41:47):
His after is after AGI, generalized intelligence, where the machines have become
smarter than us, where they start solving novel problems in ways that we couldn't even have imagined.
That after starts to think, I think starts to look a little dystopian and we
don't need to talk about that after.
But the after between AGI and...
(42:09):
Where we are today, I think is really exciting, right?
There may be some sort of, you know, Skynet future out there. Who knows?
Hopefully it's a long time in the future and hopefully we get our you-know-what
together societally and together as a people.
But between now and then, wow, exciting. And in the data and analytics space,
(42:32):
I am a huge believer in what I would loosely call the data fabric.
I don't see this necessarily as being hypey, although if you said that I was
hyping it, I think you wouldn't be incorrect.
But I see a data fabric being far more than what it is today.
So today, V1 of a fabric is this hyper-virtualization layer that allow you to
(42:56):
connect across multiple sources,
do a SQL query against a graph database or a Cassandra database it didn't it
doesn't matter it's all this kind of virtualized access layer which is really
really cool because to me what that has done is is basically functionally eliminated,
the differences in many ways between a lake and a warehouse right and and if
(43:21):
you and if you can do that you're you know some of the vendors that are focusing
on fabric like microsoft or to me are kind of pulling the rug out from the data
bricks and snowflakes of the world but that's a separate issue.
V1 of the fabric, I think pretty exciting.
V2 of the fabric will be when we start to what Gartner would loosely call activate metadata.
Where we start to kind of farm and deeply analyze the metadata of our organizations
(43:48):
to get to a point where the data can start to classify and govern itself, okay?
And if you look at kind of artificial intelligence, it's a spectrum, right?
Over on here is completely manual and over here is completely automated.
Somewhere over here is augmentation. And that's kind of where we're starting.
We're starting on this road to augmentation where the machines help us make
(44:10):
decisions. The machines can help us to better model our data.
They can help us to better govern our data or define quality rules.
Because there's data out there to tell us when data is high quality and when it's low quality.
We won't need humans to tell us when data is fit for purpose because the metadata
and the transactional data is going to tell us that.
(44:31):
We will know when transactions were successful. We will know when they were
quick. We will know when they were slow. low.
We will know when there is an error in a transaction, quote to cash,
procure to pay, pick any business process you want.
There's data that is going to tell us when things are working efficiently and
when they're not working efficiently, what data is needed to fuel those processes.
That's what I view as metadata activation. You layer in artificial intelligence
(44:55):
on top of that, where there is some sort of recommendation engine that is starting
to make recommendations on how we manage and govern data.
This includes some idea of what could loosely be called a semantic layer to
allow us to start querying that metadata in very natural language processes.
(45:15):
So a semantic layer also today is what we've got is v1, v2, v3s of semantic
layers start to change how we interact with data. We won't do it through dashboards anymore.
We'll all do it through whatever application we're using at that given moment.
And we won't need in the future, things like a common language will be mitigated
(45:36):
through AI because AI seems to be pretty good at language, seems to be pretty good at it.
So this future state where these things start to come together,
incredible, like just game changing how we interact with data,
how we manage data, how we govern data.
And companies like yourself, I think are uniquely positioned to help their clients.
(45:57):
Map out that roadmap, right? And slowly start to figure out what does my strategy need to look like?
What does my operating model need to look like? How do I need to adapt these
things and get to where we want to get to?
Because the value here is absolutely incredible. The companies that figure it
out will outdistance themselves in their competition. I have no doubt about it.
Absolutely. I love that perspective.
(46:18):
And I do agree with you. I think that the fabric, Rick, I'm seeing hints everywhere.
Snowflake and Salesforce are joining forces. Microsoft's getting included and
they're sharing data across their platforms so that the consumer or the customer
can activate all of it, regardless of where it's sitting, for example.
So they're virtualizing themselves, which is going to help with that V2 model that is coming.
(46:41):
So I'm very excited about that. I remember when IBM came out with,
I think it was IBM came out with something, must have been 16 years ago,
where they started to virtualize access to tables, for example,
across your organization.
I hated that thing. I remember my company bought it. They were like, here, Sandy, enjoy.
And I'm just like, I'm not taking responsibility for this. It's not going to function.
(47:03):
It's not going to work. But I think we're getting further and further into where
it should be. I'm excited to see the innovation continue.
And semantic layers, true semantic layers is everyone's hope and prayer.
Yeah, I agree. I think, you know, vendors are, some vendors,
I think, are very uniquely positioned
here to start looking at their businesses in very different ways.
(47:26):
I remember a time, this was 2008, Salesforce 2008 release, where they released
capabilities that they called Salesforce to Salesforce, which was basically
inter-instance sharing of data.
It blew up. They actually released the capability, but their model for it,
(47:46):
their vision for it was to allow people to share contact data across Salesforce
instances, because everybody complains about contact data.
People data is just notoriously high velocity and low quality.
And can we start to get to a point where people, basically it's commodity, right?
It's reference data if we can figure it out. Can we start to actually pool our
(48:07):
resources and pool our data and pool our management of that data,
aka stewardship, aka governance, to create a shared data set that everybody benefits from.
And if we do that, we'll lower your Salesforce annual fees. That was the original
model of what they wanted to do.
What they ended up doing was just basically to enable cross-instance sharing
(48:28):
for complex value chains and complex supply chains or partner networks.
And if you want to expose your leads to your partners for or lead sharing, that kind of thing.
But they never carried through with their, kind of what I saw as their vision
at the time. And nobody has.
2024, nobody has. What I'm talking about here is governance as a service,
(48:52):
stewardship as a service, data quality as a service.
Because if you're Salesforce, Oracle, IBM, doesn't matter, you look horizontally
across your customers, all of your customers, assuming your user agreements
allow you to do this, lawyers, cover your ears,
assuming you have the right to do that,
you've got a record for Acme, you've got a record for Acme, you've got a record
(49:16):
for Acme, and you've got a record for Acme.
And it's all managed largely the same.
It's all managed largely the same. And can you expose using AI,
using semantic layers, using data fabrics, can you expose data management as a service.
We'll manage your quality for you. And by the way, we know what good looks like
because we can look across all marketing organizations for your industry.
(49:38):
Because it's not company by company, it's division by division.
The marketing division of company A looks more like the marketing division of
company B than the finance division of company A.
So it's horizontally and can vendors start to figure this stuff out?
I was talking about this at Gartner four years ago with MDM vendors.
I said, hey, could you maybe start enabling data quality as a service for your
(50:03):
customers by looking across your customer sets?
I talked about it with so many different vendors and it's like, oh, that sounds hard.
Our user agreements, user agreements, user agreements. Lawyers are never gonna let us do that.
But that's kind of where I see things going for vendors that are brave and wanna
take those steps. Love that.
Well, you have it here, folks. folks, the future, the next step,
(50:24):
what we are planning to see and hope to see organizations deliver.
Malcolm, thank you so much for your time today. And Anjali, for the recap,
I am just so happy to see your smiling faces.
Music.
Hi. Welcome to another episode of How I Met Your Data. Sandy Estrada here.
Myself and my co-host Anjali Bansal are thrilled to have Malcolm Hawker with us on this episode.
Malcolm has over 25 years of experience in data and analytics.
He is a recognized thought leader and advisor in all things data,
but particularly master data management and data governance.
(00:22):
He has played a pivotal role in helping some of the world's largest businesses
improve their enterprise information management strategies.
He is currently the chief data officer for Prophecy, where he leads a global
field of master data management and data governance practices at the firm.
He also hosts the CDO Matters podcast, where he interviews senior data leaders
from various industries and domains.
(00:43):
Now, if any of you have ever heard Malcolm Hawker speak, you would know he's
an extremely passionate data leader.
We dive into a lot lot of different topics. But first and foremost,
this episode is recorded right after Anjali and Malcolm returned from their
CDOIQ conference in Boston, where they co-hosted a session.
And we're going to talk about that session. We're also going to talk about a
(01:05):
number of other topics that I think were little pet peeves of Malcolm and mine and Anjali's.
So we get into that discussion as well.
He's got a lot of experience, so
it's great just having these types of conversations with someone like him.
So without further ado, let's get the show going.
Music.
(01:40):
We'll be right back. so welcome malcolm and anjali fresh from cdo iq so anjali,
How did you end up at CDOIQ with Malcolm?
I had met Malcolm last year through a number of different conferences that we
just happened to be at. At one point, we were on a closing panel.
(02:01):
I think we probably agreed and disagreed much at the same time.
We just struck up a friendship, tended to run into each other at other conferences
as well. And one of the things that Malcolm had kind of offered as advice to
me a number of times was get to CDOIQ.
You're going to meet a lot of great data leaders.
It's in Boston, easy to get to. And so kind of at the beginning of the year,
(02:23):
when CDOIQ put out their call for speakers,
I ended up putting in a paper that was actually inspired by one of the conversations
Malcolm and I had had in London about data culture. So I put this paper in.
I had initially envisioned it as a panel.
As it got approved, I went, you know, we've done panels before.
(02:43):
They're met with, you know, kind of different levels of success.
So I wanted to do something a little bit differently and kind of couldn't get
away from this good cop, bad cop idea. yet.
And so just kind of thinking about that, I was like, you know who would be a really good bad cop?
Because they are very, you know, very passionate about telling you what is wrong
(03:07):
with the world today, but also offering,
not just telling you what's wrong, but also offering you some tangible outcomes to strive towards.
So, you know, so I, you know, we ran into Malcolm at Data Universe.
I had a Canadian chocolate bar with me specifically for Malcolm.
And I made my ask. I said, hey, I had this talk that's been approved.
(03:32):
It was actually inspired by our conversation.
And would you like to partner with me on this and kind of do something a little bit different, but fun?
And he said yes, which was a little surprising knowing that he actually ended
I ended up presenting Monday through Thursday.
Our talk was on the last day. So I'm like, oh my God, by the time he got to
Thursday, how his voice wasn't completely hoarse and gone is beyond me.
(03:56):
Now, Malcolm, when she made the pitch to you, did she inform you that you had to play the bad cop?
Or is that a surprise? It was inferred. Okay.
She she told she told me about the format and
kind of what we were going for right you know i don't know
good cop bad cop but more like buddy movie maybe but
but either way i mean i'd kind of assume when it when
(04:18):
it came to some of these things that it would would it be inferred that
i would be the provocateur and that that angela would be would be the ambassador
because because she is so i just kind of naturally figured that out and it's
a role that i relish in that i i am I'm happy to embody at conferences or anywhere else because I'm,
(04:39):
as said, I am passionate.
Nobody's ever going to accuse me of not being passionate.
And I think we can do a lot better as data leaders.
And I may not have all the answers, but I do know that the things we've been
trying just aren't working that well.
So anytime I have an opportunity to kind of poke a little bit and to talk about
(05:00):
doing things differently, I pounce. So thank you for including me.
Oh, no, thank you for joining me. It was fun. It was a really fun talk.
Yeah. And I will say we were on the last day of the conference where you can
kind of expect a mixed set of results in terms of participation, in terms of audience.
(05:21):
We actually had a lot of people join us. We did. It was good.
Yeah. And they were engaged. One of the things that I really enjoyed about our
format was the fact that we were able to leverage the conference's app to poll the audience.
So not only poll the folks that were in the room, but there's a virtual component
to the conference as well.
(05:42):
And the virtual attendees could also kind of live enter in responses.
Oh, that's fantastic. Yeah, I mean, there were 3x the number of virtual attendees
that there were actual attendees, which I think is great.
Kind of, you know, it's broadened the tent. And that Hova, Hova,
Hova, whatever, W-H-O-V-A app is really good.
(06:05):
I think it's actually one of the best. Usually at these conferences,
when the conference organizer makes me download the app, I'm like, oh, God, another app.
I need another app. But that one actually has utility. It works and it's just easy.
So yeah, I was glad that we had the poll. I'm glad that we had really good turnout.
The topic was good. The discussion was good. It was awesome.
(06:29):
So let's cover the topic. What was the topic? So the title of the talk was Transforming
Your Organization to Embed DNA in Your DNA.
So, you know, as we kind of started shaping the content, it was really talking
about how data leadership has been kind of operating in the same mode for the last decade plus.
(06:53):
And things haven't really changed. We haven't really moved the needle.
We've got like great opportunities ahead of us, but we're not really capitalizing
on them in the way that we would have expected or hoped or, you know,
have the potential to do.
And so part of that or a big part of that is really rooted in the data leadership
and the culture that exists in the organization.
(07:13):
Yeah, I mean, a major takeaway of our presentation was that this idea that culture
is not a deliverable, right? Culture is actually an outcome.
If you deliver value, if you embody leadership characteristics,
and this is something that we stress in the presentation as well,
if you kind of embody these four attributes of a good leader,
(07:36):
of a good data leader, if you do those things, if you deliver value to your customer,
then culture will follow. it necessarily has to follow.
So the idea that culture is just some sort of deliverable, like a checkbox on
a requirements doc, right?
It's like, okay, well, I got to do this. I got to do this. I got to deliver data culture.
(07:57):
And then when I do those things, I can deliver value to my customers.
I used to hear this as a Gartner analyst all the time.
We see it in the actual data. We see it in Gartner's CDO survey where CDOs are
asked, what are your biggest impediments to delivering value to your organization?
What are your biggest roadblocks? And the answer consistently,
(08:18):
top three, one of the top three is always there's a lack of a data culture.
And to me, that perspective is completely upside down, right?
The idea that it's a deliverable, that's kind of upside down because you're
talking about a three, four, five, 10-year journey, an ongoing journey that
never really kind of ends.
As Anjali rightly said in our presentations, like who isn't using data? We're all using data.
(08:41):
Everybody uses data all day, every day. So even the idea that there isn't a
culture that wants to use data or wants to be fact driven is probably untrue.
You kind of put all these things together.
And the main takeaway, the main thing that I was trying to say is like,
hey, double down on value, find ways to deliver value, then the culture will follow you.
But any idea where culture is some sort of dependency is probably hindering
(09:06):
you more than it's helping you. Hmm.
That's a really good and thoughtful point out. It's so thoughtful because the
reality is it's not, you know, it's not something that you go create a team around and focus on.
And I think they're also, there's a lot of competing factors in terms of conflating
culture with data literacy as well.
I've seen that a lot where, you know, data, what they really mean is data literacy,
(09:30):
but they're saying it's culture.
Those are two very different things. And I do find that people People conflate
those two sometimes as well.
So you're absolutely right. You deliver value, people will become more literate
because you're delivering value that they want to leverage.
Therefore, one begets the other. Yeah, this idea of culture as a dependency,
I think, has some rather perverting effects in what we do as leaders.
(09:55):
A focus on data literacy, I think, is one of them.
To me a focus on literacy without focusing
on the delivery of value with without focusing on having a really easy to use
product without focusing on user-centric design principles without doing the
things that we know are necessary to build great products that people get value
(10:16):
from if you're just focused on literacy because you see it as as a dependent
action to drive the data culture,
and you're not doing those other things, well, you're going to alienate the
very customers that you're there to serve because you're basically telling them,
hey, you need to figure out all this stuff.
You need to up your game. You need to do all of these things.
And what do I need to do as a data leader?
(10:37):
Well, I need you to get you to see the value of data.
And again, I would see this all the time when I was an analyst.
I would ask my customers, okay, well, what are you doing around this idea of
data product management?
Or what are you doing to drive value? What are you doing to measure value?
Practically nothing is usually the answer in that case. You know,
(10:57):
what are you doing to understand your customer needs?
What sort of feedback mechanisms do you have with your customers around their
dating? And a lot of the answers were very, very, very light.
And then I would ask, okay, what else are you doing? Well, I'm focusing on data
literacy as a means to deliver on my data culture mandate.
Oh my goodness. I think you're approaching it backwards.
Yeah. Yeah, absolutely. So what was the response in the room or even online
(11:21):
related to that messaging?
You know, there was a lot of it. Like I said earlier, there's a lot of engagement, right?
There's a lot of head nodding. I think that the message very much did resonate
because we had a number of folks come up to us after the talk was completed
with comments, with questions, with their own reflections.
(11:43):
So, you know, I think the message really landed and
gave people give people food for thought yeah i
i agree that that that tends to
be the response to a lot of the things that i share
because often i'm sharing perspectives that
i i don't want this to sound egocentric but but
i'm sharing perspectives often that i believe they probably
(12:04):
haven't heard before right the like the idea that
data literacy is is problematic right the
idea that you're you're you're forcing people to get
trained on things they may not like using
right right or have difficulty using or
lack confidence to use the idea that you're forcing them to get training on
(12:25):
that could be doing more harm than good like just that that as as a as an assertion
for a lot of people would be like what do you mean data literacy is good everybody's
telling me data literacy is good it has to be good because everybody's telling
me it's good gardener's is telling me it's good.
My consultants are telling me it's good. Everybody's saying it's good.
What do you mean it's bad?
So in a lot of the presentations that I give, there's a time needed for people
(12:50):
to kind of absorb the message.
But as Angeli said, there were many that had said, hey, listen,
this is great. I really appreciate it. So I think all in the response was actually quite good.
Because on the flip side, I didn't hear anybody tell me, and maybe they're just
avoiding conflict, I don't know. I didn't hear anybody tell me,
hey, you're crazy, right?
Like, you know, go back to the drawing board because you completely missed the
(13:12):
mark. I didn't hear any of that. So that's good.
I want to dig into what you said because I haven't heard anybody say that before.
And I have, over the last few years, I've kind of jumped on that bandwagon as well.
You mentioned the whole fact that you're trying to train somebody on something
that may not be ready to use or they may not like to use or they're just never
(13:36):
going to use is a problem in itself.
That message is, I feel, is new.
I feel a lot of people still have the blinders on in terms of,
I need to do this fancy dashboard and that's how I'm going to deliver the information
versus understanding who the user is.
(13:57):
What would you say is kind of a tactic there in terms of helping individuals hear that message, one?
And two, what should they do with that message? Well, to me,
the overarching tactic that we as data leaders need to be thinking more about,
I will broadly encompass in the in the notion of integrating product management
(14:18):
as a discipline into data management, right? Like that, to me,
that's the broader wrapper.
And if we did that, if we hired product managers,
if we were rabidly focused on understanding how our customers and I use that
word purposefully, customers, not users, not stakeholders, not business.
Business people, customers.
(14:39):
If we hired product managers who all they did all day, every day was understand
customer needs, understand how customers were using our products,
literally standing over people's shoulders, watching them use a dashboard.
What are they clicking on? What are they not clicking on? Where are they getting,
where are they stopping?
Those types of those types of activities if
we're doing that if we were polling our customers asking them
(15:00):
are you getting value from this even even
going so far as is having conversations maybe even around pricing what would
you be willing to pay for this if you had to pay for it right like think of
this as a real product right if we're serious about data products products have
price tags the idea that you can walk into a retail store and have everything
on the shelf have no price tag on it, that means it's not a product.
(15:22):
It's kind of like saying I play poker and then asking, okay,
well, what do you usually wager? Well, we don't wager anything. Well, it's not poker.
You got to be betting to play poker. And if you're talking about products,
there's got to be a price tag.
So it's these types of things, Sandy, that we need to be focused on.
And if we do all of those things, we understand what customers want,
what they need, what their challenges are, how data is used to overcome those challenges.
(15:48):
Well, that's a very different enterprise than saying, okay, I've got some training
that you're going to need to get on our data products in order for you to get value out of it.
I would argue if you do all those other things and take a product management
approach, you're going to build products that people want to use.
Will there be a training aspect to it? All products have some go-to-market function. They all do.
(16:10):
Training, marketing, there's all sorts of things that we need to do as product
people to make sure that the products are successful, including lifecycle management,
archiving, right, sunsetting products.
But if we do all those things, maybe there's some training required,
maybe, but it won't be mandatory necessarily, or it will be received very positively
(16:31):
because the product will be seen as something that is driving value to the organization.
So it's a totally different mindset. It's a different model.
That's great. That's great. Love it.
So how long was this session? Was it your standard bare 20 minutes or was it
like a 45 minutes? 45 minutes. Yeah.
So what other ahas did you have on this 45 minute journey?
(16:53):
Let's see. I mean, we dug into just a couple of the areas of failure, right?
And one of them that we continue to hear, especially from a governance perspective,
is our data is garbage. It's, you know, of such poor quality.
And, you know, that really is something that we, you know, we kind of anchored
in and said, that's not the right message.
(17:15):
You know, like, if we really want to drive a culture of data within your organization,
we need to assume positive intent, you know, in terms of your data,
people's behavior. Nobody's trying to make your life's difficult.
So if you're going around saying that your data is garbage, what message are
you communicating to your people?
Like look at that as an opportunity to either think about your data differently,
(17:38):
maybe look at the context of your data so you can allow for two things to be true at the same time.
But like really take on that model of behavior around positive intent and collaborate
with your people to understand what's most important to them.
I like that. Flip the script. Yeah.
Yeah, I love it. I mean, this kind of traces back to this idea of seeing culture
(18:02):
as a deliverable, having perverting effects on what we do as data leaders.
Positioning data quality as a burden and not an opportunity is a classic example of that.
When I was at Gartner and I had conversations around culture day in and day
out, What I came to see often was that when people said there's a lack of a data culture,
(18:25):
what they really meant is my customers aren't doing what I want them to do.
Literally. They're not going to governance committees. They're not owning data,
whatever the heck that means.
They don't care about data quality. All they care about is their business processes
and they don't care about my data. So I'm going to make them care about my data.
(18:47):
And when they don't, I'm going to express my frustration using words like garbage in, garbage out.
You can lead a horse to water. You can't make them drink.
They just don't get it. They just don't get it, Sandy. They don't see the value of data.
So when we approach everything like, okay, I'm going to make them see the value of data.
(19:12):
I'm going to make them understand why it's important to understand systems and
processes sitting underneath the data.
When I'm going to tell them over and over again that they're making my job harder
and harder and harder, out of a frustration because I can't get them to get it.
I can't get them to see the value. I can't get them to go to governance committee
(19:32):
meetings. And so I'm frustrated.
And I'm going to use words like, you need to own.
And you need to stop throwing garbage over the fence.
I just see it just add all this stuff up. And it's like, wow,
we're really disempowering ourselves here.
We're really disempowering ourselves. If we think that the pithy metaphors or
(19:53):
analogs that I use here is like,
just imagine if the chief operating officer of Frito-Lay said their corn was
garbage all the time, or if the
COO of Anheuser-Busch said their water was garbage all the time, right?
That's one end of the spectrum. The other end of the spectrum is we're sitting
on this mountain of unrefined gold.
We're sitting on a mountain of unrefined gold that everybody, everybody,
(20:16):
it doesn't matter, everybody is saying is the new gold and everybody is saying
is going to be the thing that helps my company deliver transformative value
and I'm responsible for that?
Holy cow, what an opportunity. This is incredible.
Instead, it's garbage in, it's garbage out. What can you do?
(20:36):
What can you do? Yeah, they dropped the responsibility and handed it off to somebody else.
You need to fix this. This is your doing and I can't run bad data.
The trope that they don't understand blows my mind because the reality of the
situation is every single operational team on the planet is knee deep in Excel,
(21:00):
punching data every day. That is what they're doing.
They all have analysts. They're all looking at data all the time.
They understand the value of it.
So it's not a matter of understanding the value. It's a matter of helping them
understand the process that is required to refine and ensure that value is there.
But it's even in the words, it's the perspectives, but it's also the words that we choose to use.
(21:21):
And it's the way we choose to express things. I'll give you an example.
There's soundbites, plenty of soundbites floating around saying that 80% of
data scientists' time is lost due to data quality issues.
Issues like what a ridiculous statement but we hold on to those things and we
use them we we use them to validate the fact that our jobs are so hard so hard
(21:45):
and we're and we're and we're inefficient,
and we're not as efficient as we could be and we're not driving as much
value as we could be because you make it hard because look at
this stat 80 of a data scientist time just gets
thrown out the window do you know how much we're paying for those data scientists and you're
making them them wrangled all of this data while the fact remains
the data is of different structure different
(22:06):
formats different definitions by definition by
design crm systems function differently than erp systems by design and if you
want to blame somebody blame henry ford don't blame the people that actually
coded those those applications right because they're they're i mean it's how
it is by design these things are different because our business processes are different.
(22:30):
The applications mirror the business process.
The data mirrors the application. So that's it. And it's not some sort of flaw
in the system. It's actually by design.
Sorry. See, I'm passionate. I told you I was passionate. No, I love that.
All right. So we have deliver value, change your language. Is there a third one?
(22:53):
So, I mean, we went with, you know, create the right mindset,
prioritize collaboration.
It was continuous engagement. So, you know, instead of sitting in the ivory
tower saying you must do things this way, work with your people.
Like get out there, get in the muck and actually understand what people are
doing. What are truly their pain points?
(23:14):
What do they need in order to be successful and thereby make the organization
successful, deliver that value.
And then we can start to build the right products for that. Malcolm had a great
example around a baker baking bread.
And they're the only bakery in town and nobody's buying the bread.
And the baker's standing there going, well, you just don't understand the value
(23:36):
of my bread. You don't understand how hard it is to make the bread.
Well, maybe they're all celiac and they can't eat the gluten.
Did you ask? That's a really good metaphor.
Yeah, yeah. To me, the shifts here are actually kind of subtle, right?
Like the really good news here is that, you know, we're not necessarily talking
(23:59):
about fundamental shifts in technology.
It's not like we need to like pull out all the technology and put new technology in, right?
I think a lot, we can make a lot of positive differences here with just subtle
shifts in how we manage our team. subtle shifts in how we approach problems.
There are already kind of roles that kind of align to this idea of product management,
(24:21):
although I would hire a product manager because it is a unique skill set.
The idea that a business analyst could be a product manager,
but the roles are pretty close, right?
The idea of I need to understand and document requirements, I mean, that's fairly close.
I would argue from the perspective of what do you need to do?
And Angela just touched on some some of the characteristics that you need to
(24:42):
embody. When it comes to brass tacks and like, what does the organization look
like? Who do I need to hire?
What I would recommend is one, a product manager, two is something called the value engineer.
I would argue that at a certain size of company, well, this is all sizes,
but it depends whether it's an FTE or not.
But at a certain size of company in a data analytics function,
(25:02):
you could easily justify an FTE to be a value engineer consigliere,
to be the right-hand person of a VP of data analytics to help do things like
budgeting and forecasting and planning.
You're probably already doing that already, but take that role one step further
and actually get it into the realm of starting to understand and build models
(25:23):
of how does improving data quality actually drive business outcomes?
How do our key business drivers, how are they influenced by better data?
Start measuring that thing out.
With a north star of getting to the point where your store of data products,
actually each product has a price tag on it.
Now, we could get into interesting conversations about, well,
(25:45):
data is not actually a balance sheet asset and fall down that rabbit hole if
we wanted to. I don't think that that's a useful conversation.
I'm talking really about a conceptual framework here, some sort of thought exercise
maybe even to say, if you had to put price tags on everything, could you?
That's different than saying, well, my CFO doesn't think it's a balance sheet
item, so that means I'm just going to throw up my hands and not care.
(26:06):
Yeah. No, and you only improve what you track, right? So if you're tracking
the value of it, then you could improve on it.
So yeah, that makes absolute sense.
I think I think it's challenging when it's considered a cost group.
It becomes a very challenging issue for a VP of data, for example, or a CDO.
Speaking of which, I heard that you got a new role at Prophecy. Congratulations.
(26:31):
Thank you. Yeah. CDO of Prophecy. Are you working to implement some of these ideas there?
My role is an externally facing one.
So I would best be called a field CDO. There are a number of companies that
do what we do, including Microsoft as a CDO who's more externally facing.
(26:51):
So my role, my daily responsibilities largely have not changed.
I support our clients, I support our prospects, I support our field,
and I evangelize in the market, right?
And I do things like this to help other CDOs understand how to best optimize
their organizations, how to structure their organizations, how to approach governance,
management, you name it, from having done this for 30 years.
(27:13):
So new title, largely same job.
Now that said, from the perspective of implementing what I suggest and following
my recommendations, I do come at my knowledge honestly, right?
I have managed data and analytics groups.
I've managed governance efforts. I've implemented MD. I have been the person
(27:34):
making decisions about budgets, about org structures, about all of these things.
I feel like I could live on either side. I know I could live on either side
of the equation, whether I was internally facing only or externally facing.
But given my passion, given my knowledge, given what appears to be rather kind
of decent skills for sharing knowledge, if this dog is learning some old,
(27:56):
this old dog is learning some new tricks.
I actually think I'm best suited.
I think I'm, from a value extraction perspective, doing what I'm doing and the
role that I've got, I think is a perfect word for me. So am I hearing like a
2025 Studio IQ, a year as a studio reflection piece?
Oh, that sounds great. These stories write themselves. I mean,
(28:17):
they write themselves. I think that would be great.
Maybe put my money where my mouth is and measure my own value.
We can frame it up and create the right mindset, prioritize collaboration.
Continuous engagement.
And what was the last one? Drive change.
So yes. Yes. I'm all about driving change. we desperately need change.
(28:38):
So I find it interesting, just a fun little aside, that in my presentation,
my individual presentation, not the one that I did with my learned colleague
here, I included a graphic of the Albert Einstein doing the same thing over
and over again, expecting different results. I included a graphic like that.
And the very next presentation I went to was my friend Juan Cicada from Data.World,
(29:00):
and he had the exact same graphic in his presentation.
So I was like, Oh, I wonder. Yeah. And we had, and we had Angela,
we had gone to dinner the night before with Juan having this conversation.
And he said, I'm going to alter my presentation because of the things that we just talked about.
I didn't actually get confirmation from him that, that he added that graphic
because of our conversation, but who knows?
I had an early, I had an early preview of his, of his deck that was already in there.
(29:25):
So he was already thinking about that. Sure.
The clock track around it was altered based on some of the things that we were
talking about. At some point.
Do we, as data advisors, at the end of the day, we're all advisors.
We're helping executives.
We are. At one point, it feels like an uphill battle, doesn't it?
(29:47):
With those messages, like, hey, you guys keep doing the same things,
expecting different results. We keep telling you the same things.
It feels daunting, doesn't it? It does, and it doesn't.
There's a consulting answer for you. Define daunting. I mean,
from a top-down perspective, if you look at this from a global kind of macro
(30:09):
perspective, yeah, it's daunting.
But if you take more of a bottoms-up, iterative, kind of agile approach and
just take some baby steps, and if a lot of people started to do them,
I think it would be less daunting.
But to your point, Sandy, I mean, yes, we live in a very complex ecosystem of
reinforcing behaviors that are reinforcing that where a lot of the players in
(30:33):
the market are reinforcing negative behaviors.
They're not enforcing positive behaviors.
I actually came up with a little graphic and a little theory to explain all
of this. I called it, what did I call it?
The semantic pedanticism feedback loop.
The semantic pedanticism feedback loop. It was a riff on why we keep inventing
(30:56):
new words for things, because we do.
However, it actually explains, when I look at this model that I built,
and now I'm going into crazy town, this complex interaction between software vendors, consultants,
and analysts, and customers, and users, people signing the SOWs.
(31:18):
There's a complex web that exists there where once an idea is birthed,
once something is created. I'll pick on data literacy.
Data literacy starts to bubble up, I would argue, six, seven years ago.
It's bubbling a little bit.
(31:39):
Then in 2019, Jordan Morrow, while he was working for Click,
commissioned a study. It's a vendor commissioned study.
Take that with a grain of salt. All vendors do studies to say,
hey, generally those studies already have the answers and they're just looking
for the study to validate the answers.
But that aside, that aside, Jordan commissions a study for Click that finds,
(32:02):
in essence, that finds that people aren't getting value from data because they
don't understand the data.
The underlying premise is, is what Jordan would say, is there's a skills gap,
a skills gap on users when it comes to data and the way that we deliver value is to close the gap.
This was the assertion coming out of this. So this report gets a little bit of steam.
(32:26):
People at Gartner start to hold onto this and start to smell something here.
Well, interestingly, analyst firms, I won't pick on Gartner,
I will just say analyst firms.
These are annual subscription businesses that need people to re-up every year.
And if everything was evergreen, if everything we did didn't really change that
(32:48):
much, governance isn't changing that much, MDM isn't changing that much,
much. Data quality isn't changing that much.
If things didn't change that much, those annual subscriptions would be,
I would argue, less compelling because you don't need to re-up your annual subscription.
How do you get people to re-up an annual subscription? Well, you create new things.
You make new things. Hey, flash, flash, flash.
(33:08):
Hey, there's this new thing. Have you heard of it? It's the data mesh,
or it's the data fabric, or it's data literacy.
And there's all these things that you need to learn about. out.
And they start pushing and pushing and pushing and pushing. And then all of
a sudden, people start calling asking, hey, what's this data literacy thing? Tell me more.
And then Gartner has data to show, and analyst firms have data to show that
(33:31):
people really care about data literacy.
It becomes this self-fulfilling prophecy that all of a sudden,
vendors jump into the fray and say, oh, wow, this thing must be big.
People are asking about data literacy.
Consultants jump into the fray because their customers are saying,
hey, tell Hey, Sandy, tell me about data literacy. I'm hearing this data literacy thing.
And all of a sudden, this gets spun up. It's the hype cycle, right?
(33:54):
It's the hype cycle. In the case of data literacy, I would argue it came from
one study from one vendor where all it took was a few analysts to jump on that pile.
And lo and behold, a movement is born.
I mean, that's usually the case, either a study or a white paper from somebody.
I mean, data mesh blew up overnight, it felt like. And people just took the
(34:17):
highlights of it, right?
It just, yeah, it's one thing after another like that, for sure.
But your question was, hey, you're looking at things a whole different way.
You're suggesting that we need to dispense of some sacred cows.
If true, that seems daunting. And I would say, yes, it is daunting because there's
these other forces that continue to reinforce things that, in the case of data mesh, my goodness.
(34:44):
Did anybody stop to ask, wow, okay, conceptually, this is interesting.
Full decentralization, that's interesting. Having domain autonomy,
that seems interesting.
Federated computational governance, I don't know what that means,
but it sounds interesting.
All of these things, okay, this data products, ooh, that really sounds interesting
because nobody's using my products today.
(35:04):
And it seems like this thing can maybe help with that. So all these things are interesting.
But did anybody ask the question, it was like, okay, wow, wow,
when you go from a hub and spoke to a, to a spaghetti bowl, what is the cost?
Right. And did anybody ask, wow, well, hub and spokes as inefficient as they are.
And sometimes, you know, maybe not the best for our end users and maybe not
(35:26):
the best. And maybe, maybe we have to sacrifice some, some things at the domain
level, you know, airlines are a hub and spoke for a reason.
Network typologies are hub and spoke for a reason, but did anybody stop to ask,
okay, well, what would the cost be to go from that to something?
It is this basically the spaghetti bowl, like a spider web of connection,
because that's what it advocates.
It advocates domain to domain, this organic domain to domain sharing,
(35:49):
where it's quite okay to have the same report repeated 15 times over,
as long as you're sharing between domains. Did anybody ask?
No, nobody asked. But the hype gets created, everybody pumps it up.
Three years later, it's the biggest thing and the greatest thing since sliced bread.
And I'm talking to companies that are like, yeah, we just turned off our data
(36:10):
warehouse. I was like, what do you mean?
Yeah, we're following the mesh. We're implementing a data mesh.
We turned off our data warehouse.
Do you want, okay. How are you going to support cross-functional use cases?
Well, we're figuring that out.
How are you going to give the CFO the one report that says how many customers we have?
Anyway, sorry, I'm ranting. But we live in a complex ecosystem where contrarian
(36:34):
ideas don't tend to get inflated.
The ideas that reinforce the status quo tend to get inflated or hypey things that are unproven.
Hypey things. Hypey things. What's new?
And I fall prey to that. I was going to say, one of the other things that we
touched on a little bit during the presentation was the fact that just as consultants,
a lot of times our clients are coming to us saying, how do we benchmark against
(36:57):
our competitors? competitors, right?
Like how are we performing against everybody else?
And I just sit there going, why do you want to know?
Like, let's deal with your problems and figure out like what we need to do to
get you moving forward as opposed to comparing yourself to somebody else.
Because guess what? They have the same issues, right? Like we're all solving it together.
They're no further ahead than you are. They're not doing anything differently.
(37:21):
So why do you want to know? you know and is it
just that safety in numbers is it just therapy to know that maybe others are
suffering in the same way it just it's kind of an interesting interesting quandary
in terms of like why do you want to know and like like take a risk like let's
just do something differently to move forward if i had a dollar for every time
(37:42):
i was asked when i was an analyst.
What does good look like or where do i stack up in the industry i'd be a rich
man even if they were Canadian dollars, I would be a rich man.
So completely concur, Adelaide. I mean, it's, it's, but I see that again,
getting back to this idea of mindset and how we approach what we do,
(38:02):
I see that being a fairly defensive posture.
Meaning if I'm trying to define a strategy, or if I'm trying to execute on something,
and I take the perspective of, I need to do just slightly better than the other guy or gal, right?
And if a good enough is good enough, and if I can check that box and say that
Gartner has approved this plan, or that I know that I'm doing things the same
(38:27):
as everybody else, that's a defensive posture.
Because it's not innovative. It's heard. H-E-R-D.
Right? Instead, to your point, well, okay, do you want to be in the herd?
I think some people want to go down proven pastures, right?
They want to walk the path others have walked.
They want to know that if they take those steps, it's going to work for them,
(38:51):
which isn't always the case.
Nike's culture is completely different than an older company,
an older manufacturing firm.
So it's hard for, I do not like those conversations. organizations and we have the data, right?
I mean, we work for one of the largest consulting firms.
We have the data, but it hurts. It hurts to present that because no matter what
(39:13):
we say, there are so many other dimensions to the success of that factor that
may not be present at that organization.
And they don't have control over some of those things, whether it's budget,
you know, how the organization's outlined, like structured, et cetera.
There's things there you you don't have control over that are outside of the
(39:33):
scope of your domain. So let's stay focused on what you can control.
I mean, those are hard messages to deliver because people are not ready to listen to those.
And maybe just kind of following the pack is good enough, right?
Maybe that's okay, but that's not the mandate most CDOs have been given over
the last three years, right? I think the last data that I saw was nearly 30%
(39:57):
have some sort of change mandate.
Right? So if you've been given a change mandate, if you're in charge of a digital
transformation or a digital acceleration, then you necessarily can't be following the herd.
I mean, Nike, what a great example, right? And there are others out there,
Procter & Gamble, McDonald's, Nike, it goes from 15% direct-to-consumer sales
(40:18):
to over 50% direct-to-consumer sales in five years.
That's crazy, right? And in the middle of a pandemic, Right.
So there are examples out there of companies doing this stuff and getting and getting it right.
But I want to, at least for me, how I approach this is I do want to deliver change.
(40:41):
I do want to deliver transformation because there's untapped value in our pile
of unmined gold or unrefined oil. It's there.
It's absolutely there. And I don't ascribe to the herd mentality,
but maybe for your organization, it's okay.
Maybe it's good enough. And I think for some, that may be okay.
But I think for many others, it's not. Depends where they are, for sure.
(41:05):
The question I need to ask you, Malcolm, since we have you for another few minutes,
Gen AI, obviously, forefront of everybody's minds.
Let's skip that for a second, because that was the next new thing. It's been here.
We're all trying to deal with it, good or bad or ugly. What do you think is the thing after?
I think I know your answer, but I am very curious where you're going to take
(41:26):
that. Oh my gosh. Well, what's after?
Depends on how after you mean by after, because there's an after out there that frankly perturbs me.
I read a book recently titled Our Last Invention by a guy named James Barrett, B-A-R-R-A-T.
(41:47):
His after is after AGI, generalized intelligence, where the machines have become
smarter than us, where they start solving novel problems in ways that we couldn't even have imagined.
That after starts to think, I think starts to look a little dystopian and we
don't need to talk about that after.
But the after between AGI and...
(42:09):
Where we are today, I think is really exciting, right?
There may be some sort of, you know, Skynet future out there. Who knows?
Hopefully it's a long time in the future and hopefully we get our you-know-what
together societally and together as a people.
But between now and then, wow, exciting. And in the data and analytics space,
(42:32):
I am a huge believer in what I would loosely call the data fabric.
I don't see this necessarily as being hypey, although if you said that I was
hyping it, I think you wouldn't be incorrect.
But I see a data fabric being far more than what it is today.
So today, V1 of a fabric is this hyper-virtualization layer that allow you to
(42:56):
connect across multiple sources,
do a SQL query against a graph database or a Cassandra database it didn't it
doesn't matter it's all this kind of virtualized access layer which is really
really cool because to me what that has done is is basically functionally eliminated,
the differences in many ways between a lake and a warehouse right and and if
(43:21):
you and if you can do that you're you know some of the vendors that are focusing
on fabric like microsoft or to me are kind of pulling the rug out from the data
bricks and snowflakes of the world but that's a separate issue.
V1 of the fabric, I think pretty exciting.
V2 of the fabric will be when we start to what Gartner would loosely call activate metadata.
Where we start to kind of farm and deeply analyze the metadata of our organizations
(43:48):
to get to a point where the data can start to classify and govern itself, okay?
And if you look at kind of artificial intelligence, it's a spectrum, right?
Over on here is completely manual and over here is completely automated.
Somewhere over here is augmentation. And that's kind of where we're starting.
We're starting on this road to augmentation where the machines help us make
(44:10):
decisions. The machines can help us to better model our data.
They can help us to better govern our data or define quality rules.
Because there's data out there to tell us when data is high quality and when it's low quality.
We won't need humans to tell us when data is fit for purpose because the metadata
and the transactional data is going to tell us that.
(44:31):
We will know when transactions were successful. We will know when they were
quick. We will know when they were slow. low.
We will know when there is an error in a transaction, quote to cash,
procure to pay, pick any business process you want.
There's data that is going to tell us when things are working efficiently and
when they're not working efficiently, what data is needed to fuel those processes.
That's what I view as metadata activation. You layer in artificial intelligence
(44:55):
on top of that, where there is some sort of recommendation engine that is starting
to make recommendations on how we manage and govern data.
This includes some idea of what could loosely be called a semantic layer to
allow us to start querying that metadata in very natural language processes.
(45:15):
So a semantic layer also today is what we've got is v1, v2, v3s of semantic
layers start to change how we interact with data. We won't do it through dashboards anymore.
We'll all do it through whatever application we're using at that given moment.
And we won't need in the future, things like a common language will be mitigated
(45:36):
through AI because AI seems to be pretty good at language, seems to be pretty good at it.
So this future state where these things start to come together,
incredible, like just game changing how we interact with data,
how we manage data, how we govern data.
And companies like yourself, I think are uniquely positioned to help their clients.
(45:57):
Map out that roadmap, right? And slowly start to figure out what does my strategy need to look like?
What does my operating model need to look like? How do I need to adapt these
things and get to where we want to get to?
Because the value here is absolutely incredible. The companies that figure it
out will outdistance themselves in their competition. I have no doubt about it.
Absolutely. I love that perspective.
(46:18):
And I do agree with you. I think that the fabric, Rick, I'm seeing hints everywhere.
Snowflake and Salesforce are joining forces. Microsoft's getting included and
they're sharing data across their platforms so that the consumer or the customer
can activate all of it, regardless of where it's sitting, for example.
So they're virtualizing themselves, which is going to help with that V2 model that is coming.
(46:41):
So I'm very excited about that. I remember when IBM came out with,
I think it was IBM came out with something, must have been 16 years ago,
where they started to virtualize access to tables, for example,
across your organization.
I hated that thing. I remember my company bought it. They were like, here, Sandy, enjoy.
And I'm just like, I'm not taking responsibility for this. It's not going to function.
(47:03):
It's not going to work. But I think we're getting further and further into where
it should be. I'm excited to see the innovation continue.
And semantic layers, true semantic layers is everyone's hope and prayer.
Yeah, I agree. I think, you know, vendors are, some vendors,
I think, are very uniquely positioned
here to start looking at their businesses in very different ways.
(47:26):
I remember a time, this was 2008, Salesforce 2008 release, where they released
capabilities that they called Salesforce to Salesforce, which was basically
inter-instance sharing of data.
It blew up. They actually released the capability, but their model for it,
(47:46):
their vision for it was to allow people to share contact data across Salesforce
instances, because everybody complains about contact data.
People data is just notoriously high velocity and low quality.
And can we start to get to a point where people, basically it's commodity, right?
It's reference data if we can figure it out. Can we start to actually pool our
(48:07):
resources and pool our data and pool our management of that data,
aka stewardship, aka governance, to create a shared data set that everybody benefits from.
And if we do that, we'll lower your Salesforce annual fees. That was the original
model of what they wanted to do.
What they ended up doing was just basically to enable cross-instance sharing
(48:28):
for complex value chains and complex supply chains or partner networks.
And if you want to expose your leads to your partners for or lead sharing, that kind of thing.
But they never carried through with their, kind of what I saw as their vision
at the time. And nobody has.
2024, nobody has. What I'm talking about here is governance as a service,
(48:52):
stewardship as a service, data quality as a service.
Because if you're Salesforce, Oracle, IBM, doesn't matter, you look horizontally
across your customers, all of your customers, assuming your user agreements
allow you to do this, lawyers, cover your ears,
assuming you have the right to do that,
you've got a record for Acme, you've got a record for Acme, you've got a record
(49:16):
for Acme, and you've got a record for Acme.
And it's all managed largely the same.
It's all managed largely the same. And can you expose using AI,
using semantic layers, using data fabrics, can you expose data management as a service.
We'll manage your quality for you. And by the way, we know what good looks like
because we can look across all marketing organizations for your industry.
(49:38):
Because it's not company by company, it's division by division.
The marketing division of company A looks more like the marketing division of
company B than the finance division of company A.
So it's horizontally and can vendors start to figure this stuff out?
I was talking about this at Gartner four years ago with MDM vendors.
I said, hey, could you maybe start enabling data quality as a service for your
(50:03):
customers by looking across your customer sets?
I talked about it with so many different vendors and it's like, oh, that sounds hard.
Our user agreements, user agreements, user agreements. Lawyers are never gonna let us do that.
But that's kind of where I see things going for vendors that are brave and wanna
take those steps. Love that.
Well, you have it here, folks. folks, the future, the next step,
(50:24):
what we are planning to see and hope to see organizations deliver.
Malcolm, thank you so much for your time today. And Anjali, for the recap,
I am just so happy to see your smiling faces.
Music.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
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.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com