GTM 165: Vibe prompting, AI fluency, and the new rules of go-to-market | Kieran Flanagan - The GTMnow Podcast

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Kieran Flanagan (00:00):
Everyone really has a PhD level intern that
they can work with for free.
Prompton is a think or swimskill.
If I had to tell you thatpromptin is a skill to learn, I
don't think you're the rightperson to be in the company.
I'll give you three great tips.

Sophie Buonassisi (00:38):
Why do you think prompt engineering, like
the ability to ask the rightquestions and guide these
models, will be such a definingskill for the next few years?

There is like a 100x difference in your output
if you really knew how to promptthe engines in the correct way.

Two years ago, Jiren Flanagan made himself
a commitment.

There's just no one that will know more than AI
about me if you're ago-to-market practitioner
because no one is going to workas much as me in AI.

Since then, he's treated prompting like a
craft, spending days perfectingsingle prompt.
That obsession led him to aneven bigger insight.
Prompt engineering, how youact, and context engineering,
what you feed the model, arequickly becoming four skills
that every single person willneed.
In this episode, we break downhis personal playbook for up
leveling your own AI skills,from vibe prompting to simple

(01:26):
habits.
And stay to the end for anunexpected creative talent he's
been quietly sharpening with andwithout AI talk.
His first reaction when it cameup.

I can't believe Brian said it was of all the
things Brian could have saidabout me.
All right, let's get into it.

Kieran, welcome to the podcast.

Yeah, thanks for having me on.
I'm excited to uh to be on thepodcast.

Yeah, we're excited to have you here.
And there are so many differentareas that I want to dig in
today, so we'll jump right in.
But a big one is something thatif anyone, which probably
everyone here, does follow youon LinkedIn, they've heard you
talk about.
And it's really that everyoneis talking about using AI.
But you said that, you know,the unlock is really how we work

(02:05):
with it.
Why do you think promptengineering, like the ability to
ask the right questions andguide these models, will be such
a defining skill for companiesin the next few years?

Yeah, I think prompt engineering is still like
the skill to learn.
I remember when I first startedto really obsess about AI, uh,
really when ChatGBD came outaround three years ago, and I
talked to this really well-knownCPO of the Fortune 500 company
who was building all of their AItechnology.
And he was talking to me abouthow he had hired engineers who'd
become better prompt engineersthan the ones they had in Open

(02:36):
AI.
Whether that was true or notwas like beside the point
because what he was reallytelling me was there is like a
hundred X difference in youroutput if you really knew how to
prompt the engines in thecorrect way.
And that was three years ago.
We have always kind of thoughtabout prompt engineering, which
is the ability to kind of likeask the AIs in the right way for

(02:57):
the task you want to completeand the outcome you want to get.
And we've always, I've alwaysthought like eventually do you
not need that skill, right?
Because the LLMs become sosmart that they can just do the
prompts themselves.
And we can get into that.
I think that is not true today,but they are certainly great
assistance.
And but for me, I I that'sstuck in my brain, which is wow,
like if I really learn thisskill, I'm gonna be so much,

(03:19):
much better than everyone else.
And I still think that's how Ifeel about prompt engineering
today, which it is a likecritical skill if you are a
technology worker to be able tounderstand how to work with
these models.
Now, the thing I would layer ontop of that is there's prompt
engineering and there's alsocontext.
And I think context is a reallyimportant thing, which is how
do I provide the right contextto the model so it knows enough

(03:42):
about what I'm trying to do togive me a really great output.
So I think that context, whichis becoming its own discipline,
context engineering, contextplus prompts really give you the
skill set you need to be anincredible modern day knowledge
worker.

Super interesting.
I haven't heard a lot of peopletalk at length about the
context side.
How would you say those aredifferent skills from each
other?
Like if any go-to-marketoperator, founder, investor is
looking to upscale in both ofthose areas ideally, how do they
differ from each other and howshould people be approaching
them?

Yeah, I think prompt engineering is how I
craft the ask to the LLM.
Now we could say, well, part ofthe prompt is to give the model
context.
So giving the model context isgiving it the right amount of
information to be able tocomplete the task at the level I
need.
And I think all of these modelsget much, much better when you
can give it the right, rightcontext.

(04:37):
And that goes for all of AI itwe're deploying across across
our go-to market.
Like if you can actually ingestthe right context for the model
so it really understands whatyou are trying to do and it
really understands what goodlooks like, it can actually
produce much, much betterresults.
But understanding how to giveit the context is a skill to

(04:58):
learn.
Like a really simplistic onefor kind of most marketers is,
you know, you you when I spenttime with marketers about a year
ago, I was really into is AI agood writer?
And can AI replicate someone'swriting style?
And can AI produce somethingthat you could just copy and
paste and no one would know?
And I spent time with a lot ofmarketers because I built a tool
at the time to kind of like seeif that was true and look to

(05:20):
see how they prompted the AI tocreate content because they were
all telling me it's rubbish,it's bad.
And you would go to them andyou would look to see what they
were doing, and they would say,write me a LinkedIn post about
how to do lead generation.
That's it, that's their entireprompt.
And there's some fundamentalthings wrong with that prompt.
So, first of all, you've givenit, you haven't given it context
on what a good post looks likeand why that good post is a good

(05:42):
post.
Like you do some analysis, yousay, here's the context, here's
a post that did really well,here's the context of why it did
well, and this is the kind ofthing I wanted to replicate to.
The other interesting thinghere, which is like a really
interesting little tidbit aboutprompting, is in that prompt,
because I say create me a quoteunquote LinkedIn post, because
the LLM has a training set withlots of LinkedIn content, and

(06:04):
LinkedIn content is generallynot good, it's gonna create
something not that good.
And so you should not sayLinkedIn, you should just say
like a great post, right?
Don't give it the platformbecause it will try to skew
towards that platform.
But basically, if they hadcrafted a prompt to give it a
context of like a post thatperformed really well, some
context on why that postperformed really well, and then
the other thing, just kind ofmake that prompt a little longer

(06:27):
to say, here are some writersthat I admire and give the names
and are really good, mimictheir style and it would go off
and like replicate that style.
So there's like ways that youcan make it much, much better.
But a lot of people do lazyprompt them.
They just say, like, create mea post about this thing and 10
points, right?
And that's why you you putgarbage into it, you get like

(06:48):
really bad results back, andthat's what's happening.

Yeah, the the context almost matters the most
now.
And I've heard you talk aboutAI fluency at length and how
it's gonna be the skill that Imean is is absolutely central
for every single go-to-marketperson and founder.
So when we think aboutprompting, are there frameworks
or mental models that you knowyou coach people on to use when

(07:13):
designing these prompts?
Like you just gave a really,really great tactical one.
We even take a step back.
Like, how do people begin?

Yeah, so I I've had a lot of good conversation
with founders who are trying tomake their team much more AI
fluent.
And there's kind of I I alwaystell them do not over-engineer
this.
But plus, I'm definitely on themore cutthroat side.
So I want to like the the letmaybe the less empathetic side.
So I basically tell them look,there's three things that are
gonna matter to make your teammuch more fluent in AI.

(07:42):
First one, prompting is a skillthey need to learn.
Second one, inspiration is abig part of how people
accelerate and expand theirusage, which means if you have a
shared Slack channel, you havepeople who put the things that
they're doing that are workingon that Slack channel, and
there's just a stream of thingsthat people are doing.
That's actually one of thenumber one ways that people
learn within companies how touse this for their own

(08:03):
discipline because they can seehow other people are using it.
And so that's that inspirationpart is really, really big.
And then the third one is likemini hackathons.
So hackathons used to be adeveloper thing we did, right?
When we were for specificallyfor developers and engineers, I
was an ex-engineer anddeveloper, was never very good
at hackathons because it was areally bad coder, but can I have
a vibe code, which is reallygood.
But hackathons, now we can alldo them.

(08:24):
Like go to market teams can dothem.
Marketing can build AIworkflows, sales can build AI
workflows.
And so these hackathons wherepeople get together and build
things in AI is another greatlearning technique.
But to come all the way back tothe first thing in terms of
prompting, I think prompting isa sync or swim skill, which
means if I am a person, if I'myour manager, your founder, if I

(08:46):
have to tell you that promptingis a skill to learn, I don't
think you're the right person tobe in the company because there
is enough information to tellyou that AI is important.
Prompton is how you work withthese machines.
And every there is so muchinformation out there about
them.
So if you go to OpenAI, theyhave these cookbooks that tell
you how to do prompt and permodel.
If you go to Anthropic, theyhave prompting guides.

(09:07):
If you go to Gemini, they haveprompting guides.
There's no excuses, but I cangive some tips for people to
really accelerate theirprompting skills in a quote
unquote vibe way, right?
Like, which is basically theLLMs are going to do a lot of
this for you.
And that's the shortcut.
I don't think people realizethat the LLMs, if you work with
them correctly, can actuallyaccelerate your ability to

(09:28):
prompt pretty, pretty rapidly.

I don't know if it's out there already, but I
think you just coined vibeprompting, Kieran.

I know.
And it's actually what'sinteresting is right, we have
ViMarketing, Vibe, VibeMarketing, Vibe Coding.
We have anyone who hangs arounda lot on LinkedIn, I'm on
LinkedIn, LinkedIn a lot.
We have Vibe Cometon, which isbasically people are putting
their LinkedIn comments andautopilot and having AI do them.
The worst use of AI ever.
I don't know why people thinkthat's a good use of AI.
But VibeProm is a good idea.

I think you put a post out on that
yesterday.

Yeah.
Like Vibe Prompton is like aspowerful as Vibecodin.
I think it's like a reallyincredible way to use LLMs.

Mm-hmm.
It's really cool because youhave an engineering background
and you know, now you'reessentially doing the same, same
kind of hackathon setups and soforth, but in a go-to-market
context.
And that's a really valuableskill that lends itself to every
go-to-market team.
I know we ourselves we had alittle hackathon where we
outlined essentially everysingle workflow that AI would be

(10:27):
monumental or incremental in.
And then we rated each, rankedeach, and now we go through
workflows where we build out AIand then for each of those and
then progressively, you know,progress them on a maturity
curve.
But the importance is doing itregularly.
So Yeah.
How do you think that's a greattime?
How do you think about ROI?

(10:50):
Obviously, we all know it'sincredibly valuable, but do you
think about it as time saved,lead quality, conversion lift,
something else?

Well, I think if you if you integrate AI
across your go-to market,there's places where it will
actually improve yourperformance.
And also there's places whereit improves your efficiency.
And so I'll give you a coupleof examples.
If you integrate AI across allof your kind of email workflows,
so we automate a lot of theprospect and that our sales reps
do, and we can, because wegather enough data and data,

(11:23):
data layer, the data layerreally matters to the
performance of your AI usecases.
The better the data quality,the better the AI outcomes.
And when we integrate AI acrossall of our workflows and we
personalize those emails to theindividual, because AI does a
really great job of that, we seecontinual improvements in
conversion rates.
So in that way, we can measuremore and more meetings being

(11:44):
booked for the sales reps andless amount of sales reps having
to spend their time to do that.
Now there's other ones wherewe're building AI functionality
for sales reps to actually usewhen they're selling, right?
That actually helps themincrease deal velocity, it helps
them to increase their connectrates, it helps them to do all
of these different things.
So imagine you're in our CRMand you have an agent that
basically can summarize all ofyour deal details, they can

(12:07):
summarize all of your previousinteractions, they can provide
you great talking points whenyou're on the call to say, hey,
these are the talking points youshould hit.
In that case, what we look atis if the rep used that AI
functionality, are we seeing anincremental increase in close
one rates?
So we can actually correlateusage to incremental deals once.
So again, we can actually havea good number around that.

(12:27):
Now there are other teams wherewhen you integrate AI, it's
hard, it's much harder tomeasure.
So our product marketing teamat Hustlot is an incredible team
for adopting AI.
And they use AI in a lot of uha lot of interesting ways.
It's just a harder thing tolike say, well, because you're
using AI, what exactly ishappening?
And they're not the onlyexample.

(12:47):
For the most part, people areusing it to be more productive,
but how do you measureproductivity?
What's the metric forproductivity?
I was before I was in HubSpot,left HubSpot, went to be the CMO
GM for self-serve at Zapier.
And Zapier are a company thatsaved people lots and lots of
time.
And I saw that all the time,which is trying to say, well,
you know, how do I actually, howdo we actually showcase the

(13:08):
amount of time we're saving youbecause we're automating all of
your work?
It's a much harder thing to do.
One of the ways you can do islook at usage.
Are you actually using these AItools and just measure adoption
and usage and believe that it'sgetting better?
But productivity is harder thanthat kind of binary metric
where you can see something goodhappening in your go-to-market.

Mm-hmm.
Yeah, unless you've got timetracking across every single
activity and every single FTE.
It's it's a bit more talkingabout it.

That's impossible.

Yeah, and it's so hard to do that.
Quick pause.
Are you a B2B marketer runningcampaigns for target accounts?
And you know the struggle.
Tedious and manual processes,endless delays to get things
live, and sales feeling likeyou're not doing enough.
That's where Mutiny comes in.
Mutiny is the fastest place tolaunch breakthrough campaigns
for your target accounts.
AI agents, research youraccounts, build personalized

(13:58):
landing pages, and scaleeverything from LinkedIn ads to
sales handouts in one seamlessworkflow.
No more stitching toolstogether, just smarter, more
impactful campaigns powered byreal data.
So you can launch in days, notweeks.
See why teams who use Mutinygenerate 3x more account
engagement.
Book a demo at mutinyhq.com.
That's mutinq.com.

(14:20):
Also in the show notes.
All right, back to it.
And you mentioned just beforethis that, you know, AI
prompting, if people aren'tinterested in learning about it,
it's not a good fit.
Now, you yourself, I knowyou've written about the best
way to actually learn promptingor learn AI in general, is to
actually just do get your handsdirty, be building.

(14:41):
What about for folks, you know,that are looking to learn
overall and want to get better?
Like what would you recommendfor people to learn more, to do
more, to just overall upscaletheir AI adoption and prompting
skills?

Yeah, I'll give you three great tips.
And I have I have this comingout in a Substack post around Vi
Prompton.
And so I can't give you theactual prompts because they are
quite long.
And so I can't just like readthem out because people will be
bored to tears.

We'll pop it in the show notes for everyone.
Yeah.

I can give you generally what I'm doing.
And so one of the things I'mobsessed about is memory.
And Chat GPT, now Claude hasmemory.
I suspect Gemini will havememory if it does not yet.
And so memory is incredible.
So when we have thought aboutthese LLMs, we were like, hey,
how do they have lock-in costs?
Like, how do they haveswitching costs?
And why can't you just go fromone AI assistant to the other AI

(15:30):
assistant?
And memory is the lock-in,right?
ChatGPT knows so much about me,my results get better over
time.
But what's really phenomenalabout that is ChatGPT can be an
incredible coach.
And if you actually promptChatGPT to basically say, from
memory, from what you know aboutme, look at my role, look at
how I generally use you for thekind of workflows and use cases.

(15:52):
And based upon all of that,show me how I could prompt
better.
So basically show me some ofthe use cases that I have used
you for prompting, and then takethat use case and show me a
before and after prompt on howyou will improve that prompt.
So ChatGPT itself is going tostart to coach me on how I can
prompt via ChatGPT by looking athow I prompt in ChatGPT.

(16:15):
It's kind of like inception,right?
And it is incredible.
Now that works for anything.
If you ask ChatGPT how I canjust use you better based upon
memory, it can give you reallygreat guidance just how I can
use ChatGPT.
It can give you new use cases,it can tell you how you can use
it much much more deeply.
So that's number one is becauseif you are using ChatGPT and
that cloud has memory, you canactually ask them, if you know

(16:38):
how to prompt, you can actuallyask them to basically guide you
on how to improve your promptthen with them.
And it does a really great job.
So that's like tip number oneis ChatGPT as a prompting coach.
The second is actually the mostvaluable.
I started doing this a longtime ago, is every model for the
most part has a prompt andguide.

(16:58):
And so GPT5 has a prompt andguide that basically came out
pretty recently.
Every model has its own promptand guide.
And so what you would do is youwould take the prompt guide and
then you would take a customGPT who is built on that prompt
guide to be your promptengineer.
So I have a prompt engineerthat's a custom GPT for every

(17:20):
single model that basicallylooks at the documentation and
then provides me with the rightprompt based upon whatever
output I'm trying to get.
So I would say to it, hey, I'mtrying to do this task.
I want to do it in GPT 5.
Can you give me the rightprompt?
And it will give me a perfectprompt because it's trained on
that documentation.
Now, the other cool tip there,if you want to go one step

(17:41):
further, is I love perplexity.
I use Perplexity Labs a lot.
You can ask Perplexity Lab tocreate a quote unquote
onboarding doc for prompt in foreach model based upon elite
practitioners.
Remember that word, elitepractitioners.
It will go look for people whohave real domain authority in
the space, create an onboard anddoc, and you can use that to

(18:02):
create a custom GPT trained onthat onboarding doc for that
model.
So again, you have a promptengineer in your back pocket.
So there are two.
The third one is you cancreate, if you really know how,
you can create prompts to doself-evaluation.
And so you can basically havethe different models evaluate a
prompt and give you edits andtell you how to get better.
So that's why it's true vibeprompting, right?

(18:24):
I'm using the LLM as anassistant to get better.
Now I'll just end with this.
I've been building a softwarewith a couple of good friends,
developers.
I I basically was an engineer,wanted to be a coder, was not a
very good coder, had got fell inlove with vibe coding because I
was like, I can build an appmyself, kind of ship 30,000
lines of code, and then realize,but I can't really ship

(18:46):
production ready softwaremyself.
So I've got two developers, andI've worked on all of the, I've
been working on all of theprompts.
And I would just like when youwhen you think of vibe, you need
to still do work.
And so the LLM can give you afirst draft, but you really need
to work hard and diligently onthe prompts to improve them,
right?
Don't ever just you can cut andpaste and it's a pretty good
prompt, but if you want great,you still need to actually edit

(19:09):
and improve that promptyourself.
I've worked on prompts, uhsingle prompts for days on end,
like days on end, just trying toperfect this prompt.
And so there's some tips thatyou can get much, much better
just by using Chat GPT, buildingthese custom GPTs, and using
the LLM to do self-evaluation.

Days to create a prompt.
That is incredible.

Yeah, just like iterate and iterate and
iterate.
And Darmesh had a co-founder ofHouseFot had this really great
line recently that uh reallysticks with me, which is the
quality of outcomes is basedupon the amount of rep
repetition you do for thatoutcome, right?
It's uh it's th those lines arein sync.
That is how I feel aboutprompt, and the quality of
prompt is is pretty much in syncwith the amount of repetition

(19:51):
you do.
And I re if you repeat, repeata little bit better, a little
bit better, you can get it youcan iterate your way to like
pretty great results.

Which is like most skills, even sports, as
you train, right?
The greater repetition, thebetter skill set that you come
out of it with.
Now, when you think aboutprompting and developing these
like these quite complex promptsthat you're working on, first
and foremost, I know for youknow more the beginner side at
least, and something that I leanon a lot is just asking ChatGVD

(20:20):
to create the prompt for me tobe a little hack rather than
doing the myself and then tryingto refine it and trying to
learn backwards.
But when people are progressingand becoming more mature to the
rate that you are runningprompting, what makes a prompt
system scalable?
Like, how do you actually scaleyour prompt system?
Because it sounds like refiningthese prompts like it's quite

(20:41):
time consuming.
How do you actually scale?

Yeah, I think I can I can only speak to some of
the things that I do.
There are like great tools nowthat do a prompt evaluation,
prompt version and control.
But I use Entropics Console andOpen OpenAI's playground.
And so what they are is you canbasically go in and create your
prompts.
They have really cool toolswhere like I can go in and say,

(21:05):
hey, here's my first version,and I can click a little button
and say optimize and say, Canyou make this better?
And it will provide againsuggestions to like make that
thing better.
And what I really love reallyis the version and control.
So version and control for aprompt is I have that first
version that I did.
Now I have a second versionthat Claude has made much
better, and I can run both ofthem, and I can see, I can give

(21:27):
it the kind of inputs, and I cansee based upon the output how
much better it's gotten.
And so those two systems haveworked really well for me.
Like again, just using theOpenAI Playground and I tropic
console.
But there are probably a lot oflike more sophisticated systems
that allow you to scale yourversion and control, your kind
of prompt evaluation and thingslike that.

(21:48):
I just I just have not likeused them.

Super interesting.
And people a lot of time referto AI now as your co-pilot,
right?
Or your co-worker or yourco-founder.
Do you feel like now with AI ateveryone's disposal, everyone
is a manager or working in ateam, even though they were or
are an individual contributor?
Has that changed the way thatteams are overall structured?

Yeah.
I I think you have what'sinteresting is everyone really
has a PhD level intern that theycan work with, right?
Because that's the that's wherethe open AI model is.
And so that's kind of bananas.
Like you, you know, you used tohire PhD level interns and you
they were awesome.
And now you kind of have onefor free.
And you don't just have a PhDlevel intern for free, you have

(22:40):
as many as you want because Ican run multiple prompts all at
the same time across multiple AIassistants.
I do think it's a new skill tolearn.
Like prompting is basicallyasking asking a smart person to
do something for you.
Now you have to ask it incertain ways, but there's a lot
of people that haven't had towork with anyone before, right?
And so, like just working withpeople and you know, giving

(23:03):
someone tasks is brand new.
And so everyone has someone,now everyone has someone that
they're managing, and thatmanage that person they're
managing is this AI assistant.
So that is like how you canintegrate that person into your
work and start to really thinkabout well, what is the things
that that person can take offme?
And it does take some amount ofthought to do that.
And because of that, whereshould I spend my time to get

(23:26):
more leverage?
Right?
Like, where should I kind of10x my skill set if AI is able
to do a bunch of the things thatI used to do?
Because I think this is goingto be much more important in the
future for me to be a masterat.
And then I think the other wayit changes it is like eventually
you will have a team of agentsinternally, and knowledge
workers will have teams ofagents.

(23:46):
And I think teams of agents,the skill you have to get really
good at is how to train thoseagents, right?
So when you deploy an agent,you give it a bunch of context,
basically onboarding.
Like you onboard it to the taskand you tell it how that how
what good is, but over time youhave to continue to like teach
it and tell it how to getbetter.

(24:06):
And so this notion of having anAI trainer, I think is going to
be a role in most companieswhere that person is really
going to help train those agentsto get better at their task
over time and have someonethat's gonna manage those
agents, deploy them, onboardthem, and improve them over
time, and maybe maybe eventuallygive them performance reviews
and do all that kind of weirdstuff.

Yeah, yeah, very true.
Everyone is a manager.
And I've read that you knowpeople can manage personally
about six to eight agents at atmaximum right now.
Whether that's true or not.
I mean, what's your take?
How many agents do you think ispossible for one individual to
manage themselves now?
And then what do you think itwill progress to in the future?

I I think the I think it's as good as the back
end this so I've it's actuallyinteresting.
I was talking to someone aboutwhat I think sticky use cases
are in AI.
And I think the managementplatforms for AI assistance is a
really sticky use case.
And so you can imagine you havea platform where you can see
all of the work that agents aredoing, you can you can train
them within that app, you canonboard them to new tasks.

(25:11):
So you're like a realmanagement platform for for AI,
like a version of workday forAI, right?
So dependent upon how good thatis, it will increase the amount
of agents that you can actuallymanage.
Those platforms are far and fewbetween.
I don't think a lot of themactually actually exist.
So yeah, I think it I thinkit's going to be predicated.
But the other thing is I wasmess, you know, I I always mess

(25:32):
around with building things.
You can have an AI manager whomanages agents.
I have an app, the app that I'mworking on has a manager, and
that manager is the one givingthe other agents tasks, not me,
right?
And so eventually it's like,well, how many are managed by
the person and how many aremanaged by the the actual AI
managers themselves?

Super, super interesting.
And it it will be interestingto see if we see more workday
platforms for AI agentsemerging.
You know, we're seeing morelike hey Manning Medina's
company for monetizing agentsand so forth popping up.
So there's a whole realm ofcompanies that are either
emerging or or going to be thatare coming out.

Yeah, I I I I'm, you know, I try to be a
pretty active pre-seeded C stageinvestor.
I'm an investor in one that Ihave like a lot of faith in.
So that's like you know, I Iinvest in things that I have a
ton of faith in, and that's ause case I have a ton of faith
in.

Very cool.
And you use you touched aroundyour personal use case for AI in
a few different aspects now.
You use it also as a chief ofstaff.
What are the use cases you'drecommend to anyone for the most
impact when somebody's gettingstarted with building AI agents?
And of course, that's socontext-dependent to each

(26:48):
individual, but are thereoverall like synonymous use
cases across the board of go tomarket or founders that you
found you recommend to anyonebuilding and wanting to up level
their AI use split room?

Yeah, I have a couple of interesting, like
really quick hacks here.
Again, I wrote it by the an AIgrowth operating model that most
people could roll out.
And there's a couple of thingsin there.
So one of the easy uses I loveit for is if you are running a
team, one of the looks I like tohave is it's called a moment a
momentum deck.
And so it's basically just oneslide.
I all of my gro all of myoperating models are like

(27:24):
admin-like, because I wantpeople to work, not have to do
admin.
But there's a slide that saysbasically, what did I ship in
the past two months and what amI shipping in the next two
months?
And it's the the thing isstructured so every two weeks
the same deck is uploaded,updated, and it's structured in
a way where it's easy, whereit's easy for an AI to pull out
information.
So I can basically upload it tomy AI assistant and say, okay,

(27:46):
what have we missed?
What didn't we do that we saidwe were going to do?
What are the areas of overlap?
One of the things I look at isevery team fills out a blocker
and if they have a mic uhmitigation plan or not.
And I say, well, what are theblockers that don't have
mitigation plans?
So AI is the an ability to likehelp me keep on top of those
things is really good.
I do a similar one for KPI.
So that's a momentum.

(28:07):
That's a that's a momentumlook, which is basically how
quick are we going, are we doingthe things we said we were
doing.
The other one is theaccountability part, which is
like every month did we hit thedeliverable we said we would
hit?
And again, it's a single deck.
Every month has a new slide, soit's all in a singular deck.
And the reason they're insingular decks, just so people
know, is one of the frustratingthings if you're using ChatGPT

(28:29):
or Claude is you have tocontinue to like upload the
document every time there's anew update, right?
So if I'm if I've got my Augustupdate and then I get my
September update, I have tore-upload the document because
it has the September update if Iwant to query it.
Now, if you were queryingmultiple decks, like for each
month, they have a certainamount of files you can upload.
I think it's 20 in ChatGPT,it's certain similar in Claude.

(28:52):
So you'll just run out of theability to continue to upload,
like you upload the June one,you upload the July one.
So if you have them in singulardecks, it makes it much easier
because you can just upload thatone deck.
Um, and so that accountabilityone basically was again, it
shows did we do what we said wewere gonna do?
And it's really good because Ican just upload the doc each and
every month, and then I can runa bunch of prompts to say,

(29:13):
well, what areas are we missingon?
What areas are weoverperforming on?
What what are the bestopportunities the team has seen
that we should take advantageon?
They can even query it and itcan be really your chief of
staff in that way.
So they're they're two of thebest like it is a kind of like
chief of staff slash projectmanagement.
If you s if you structure yourupdates in in in the right

(29:34):
formats, that's that's one of mybest use cases.
I love that use case.
Um, the other one is if I havea really hard problem to solve,
it's a great thought partner.
Now it's certain model like Ihave you know all of the models,
so I have GPT Pro.
The GPD5 Pro, I can't evenremember what it is, like $200 a
month one.
I don't even know what they'recalled anymore.

(29:54):
GPD5, really powerful.
I don't know what it's called,but but basically, if I have a
hard problem to solve, I give itall.
All of the context about thatproblem, all of the historical
decks, everything that I thinkis important.
And it's a great thoughtpartner.
And one of the things I ask itto do that's really useful is
red team stuff.
Show me all of the ways thatthis is wrong.
My thought process is wrong.
It is a counterpoint to youwhere you think is the best use

(30:16):
of it.
Because what LLMs want to do islike reinforce your great,
you're great.
Yes, you're right.
Because that's who they'rethey're kind of like tuned to do
that.
So I force it to tell me I'mnot right, be critical.
And I that's what I love itfor.
Because even in the worksetting, we don't really like
being critical to each other,right?
Like I know we have likeradical candor and all these
things, but people aren't thatgood at it.
AI is really good at it if youtell it to be.

Yeah.
A humbling experience, that'sfor sure.

Yeah.
Yeah.
AI as a coach is a real greatuse case as well because it's
not biased.
It will just tell you likeyou're bad, get better.

Yeah, the the hard, honest truth, always.
Right.
Go-to-market.
Yeah, this is revolutionizingthe way that we operate
individually, like we've beentalking about.
But how does that thread itselfto the greater go-to-market
system, meaning how we'reactually building and selling
and scaling software companies?

Yeah, I think there's three trends that I
think a lot about that arehappening that I can give people
a quick synopsis of.
So I think we're gonna have tobuild influence, not clicks.
I think AI engine optimizationis the number one skill to
learn.
And I think multimodal, youryour entire website will
probably at some pointtransition to like more of a
multimodal experience, and I'llgo through each one.
So forever we've been trying tolike create content to acquire

(31:36):
clicks, and that's how B2B hasworked.
80% of all B2B buyer journeysstart in Google.
Like Google has been thehoneypot for how we've acquired
demand for our business.
It's estimated, I think in 2027or 2028, 95% of a buyer's
journey in B2B starts within anLLM.
Uh and and the problem is thatall of our clicks are

(31:56):
disappearing.
But you what you want to do isyou still want to influence your
buyer.
And I think the way youinfluence your buyer is not
through blogging, it's throughmediums like this.
It's through what I callpersonality-led growth, which is
like I think B2B will look verysimilar to B2C, where we
gravitate towards individuals,not brands.
And all of the channels thatare still growing and have great

(32:17):
momentum favor the individual,not brand.
Podcast, newsletter, YouTube, alot of the social channels,
they favor personality, notbrands.
It's why some of the bestfounders, if you look at Roy
from Cluy, what's he really goodat?
Personality-led growth,marketing.
A lot of the founders of AInative startups have real spicy

(32:38):
takes, have real thoughts, arelike really prevalent across
social and people gravitatetowards that.
So I think your content programlooks less like keyword
optimization and blogging andlooks much more like media and
creator-led programs.
So I think you'll have acollection of creators and
that's how you go to market.
The second one is that 80%started in Google, 95% will

(33:00):
start in LLMs.
AI engine optimization is howyou drive visibility in ChatGPT
and these different AIassistants because all of the
research is being done in there.
Now, when you look at the data,someone who's gone through an
AI assistant, let's say ChatGPT,converts four times higher than
if they came through Google'sblue links.
And people would say, wow,that's a good thing.

(33:21):
And that is a good thing,right?
They're much more qualified.
But why are they morequalified?
Because they've done all oftheir research in ChatGPT and
ignored all your marketingmaterial.
So by the time they come toyour website, they're qualified.
They have a couple ofquestions, they're ready to buy.
But you have to be visible inthose assistants.
And so you really have to learnthe kind of techniques for AI
optimization to increase yourvisibility, sure, voice in these

(33:44):
AI assistants.
And the third one ismultimodal.
Because they come to yourwebsite, they're ready for a
sales conversation.
But most people don't want todo a sales conversation.
But these multimodal agentsthat are able to do voice, do
see your screen, do audio, andeven sometimes like these kind
of digital avatars, I think asthey become much, much better,

(34:05):
we're going to see your websitetransition to a close-in
mechanism.
Right.
Today it's a lot of research.
Like we want to bring our brandto life and tell you why you
should buy.
We all, I've done all that inthe AI assistance.
I want to talk to someone whocan answer these final
questions, but I don't want itto be a human.
So I think these multimodalagents, your website is going to
shrink, and they're going tobe, you're going to have these

(34:26):
multimodal agents that can havereal conversations and answer
those questions.
And then you could book timewith a rapper decide to buy.
So those three trends, which iscreator-led marketing, AI
engine optimization, website asa closing mechanism with
multimodal agents kind of bakedin, is the biggest changes.
I some of the biggest changes Isee in the B2B go-to-market
playbook.

Those are drastic changes for go to
market.
So it'll be a very, veryinteresting time ahead.
But everything isn'tone-dimensional in the sense
like it is still feeding the LLNsometimes, depending on
authority and so forth.
So it's actually kind ofbifurcating the process where

(35:09):
it's supporting you top of thefunnel, but then it's actually
running the relationship sideand I'm a closing bottom of the
funnel.
That's fascinating.

Yeah, that's a really good point.
Yeah.
So like use so L how so what isone way that you can increase
your visibility in these AIengines?
And it's basically to createlots and lots of niche content
because the way we talk to an AIassistant is very different
than we were taught to likesearch in keywords in Google.
And so we're like it's like youand I having a conversation.
If we were having aconversation about a software

(35:38):
product, we're not doing what wewould do in Google, which is
like best 10, you know, best SMBsoftware product.
And so it means you need tocreate, instead of like one page
that optimized for threekeywords, you need to create a
thousand pages around a specificlike part of your product.
And so you do need a websitethat can cater to that.
But the interesting thing isthat's the first example, I
think.
Maybe not the first, but one ofthe key examples of where

(35:59):
you're building somethingspecifically for an agent, not
for the human.
Because the human is not goingto consume that content.
They don't care about thatbecause they've got their answer
from the AI assistant andprobably your multimodal agent.
And we're going to teachpeople, I guarantee this, right?
We are going to teach peoplenot to bother have to not have
to read for themselves.
People are just going to getlazier.
I like we see it all the time.

(36:20):
When something gets easier andfaster, the person consumer gets
lazier and they expect more.
So they're not going to go anddo like an hour's research.
They're going to just askChat2BT and then they're going
to talk to your agent.
But you do need all thatcontent for agents.
And it's like an interestingexample of where we start to do
go to market for the human andgo to market for the agents.
Because the thing I'minterested in, very interested
in for B2B, and Google releasedupdate this week, which allowed

(36:43):
agents to do payments, which Ithink is really huge.
Because no one, no one wakes upin the morning and says, you
know what I want to get reallygood at?
Buying B2B software.
I want to go and like reallyfigure out how to be great at BM
buying B2C BD software.
So why wouldn't we offload thatwhole procurement process to an
agent who can do payments?
And that way, then who am Ieven marketing to?

(37:04):
And how do I like for like howdo I get the agent to pick me?
Right.
I think that is a a tricky,like that's going to be a tricky
thing for software vendors tofigure out.

Definitely.
I love the point that you madeabout go-to-market for AI agents
and go-to market for humans.
It'll be interesting to see howthat actually happens and
whether it's full bifurcation orwhether it's integrated.
But one area, you know, on themarketing side, we talked about
websites and venian.
The other side is the creatorside.

(37:35):
Some argue that AI will erodecreative intuition.
What's your take?

I think creativity is I think AI makes
creativity more important thanever.
I think the way we stand outabove the noise is human and
creativity.
I think people will gravitatetowards the reason I think
creator led really works.
I've been talking about it fortwo years, but the reason I
think it really is going toaccelerate is because people
will gravitate towards peoplebecause that's who we are.
Like we're not going to we'regoing to trust people.

(38:03):
We want to hear people's pointof views.
We don't want to get all of ourcontent from AI.
And I think AI, is it a cre agood creative tool?
I think it's a great creativeassistant.
It's not as good as humans atcreating genuine creative
assets.
I think the human sk the humanskill to learn is still like
creativity that allows me tostand out above the noise.

(38:24):
So I actually think it makesthat skill set much, much more
important.

Yeah.
And what about reshapingmarketing overall?
It's crazy to say just two tothree years.
Beyond that is an even furthertime frame, but obviously
marketers should be learningprompting.
What other skills should theybe learning?
And what is a marketing teamfluent in prompts and AI
actually look like even one, twoyears down the line from now?

I think marketing is somewhat unique in
terms of a team in that it's acollection of like niche teams,
right?
Like if you're in a sales team,you kind of are a seller and
you have the same work and thesame career path.
If you're in the customersuccess team, the customer
support team, whatever the teamis, it's kind of the same work
and the career path is the same.
In marketing, you could be inthe product marketing team and

(39:20):
the brand team or the demandgeneration team or whatever
team, and they have a nicheskill set and their career path
may look a little different andtheir team size may look a
little different.
And so one thing I think AIdoes is probably force marketing
to be less specialized and moregeneralist because AI can do
this specialization.
Because why do we have suchbreakout of all these niche

(39:41):
skills?
It's because of the domainexpertise, right?
It's really hard to be a greatproduct marketer.
Uh, you can't, it's really hardto be a great brand marketer.
It's really hard to be a greatdemand generation marketer.
You need a lot of domainknowledge.
So it's hard to like do subsetlike multiple of those roles.
But if AI has a bunch of thatdomain expertise, the marketer
is like actually, I can be amuch more generalist and do more
work powered by AI, but I stillknow a lot about marketing.

(40:03):
I have like domain expertisewithin marketing.
So I think one of the shiftswill be we'll see more
generalist teams powered by AIable to do much more.
I also think marketing can takeon way more of the customer uh
journey because marketers arealways automation for the most
part starts with marketer.
And so we have these handoffpoints today that exist because

(40:24):
you know we have to hand theperson over to the sales or
whatever it may be.
But I think as AI becomes moreprevalent, it may just be that
marketing can do much more ofthe customer journey because
they can integrate AI, and AI isdoing all of the qualification
discovery.
AI is doing a bunch of thework, and marketers are like
managing the assistants andtraining the assistants and
training the agents to do thatwork.

(40:45):
Um, I don't know where we endup in two years.
I think the thing is it'schanging so fast.
So, what I tell people is themost important marketing skills
to have are be curious and bepersistent.
Curiosity, there's never been abetter time to be curious.
There's never there's neverbeen a more important time to be
curious.
I think it's the number oneskill set, the number one trait
to look for, because if you'reevery everything is getting

(41:07):
rewritten, and for me, that'sawesome.
I think I get really bored wheneverything is like optimization
stuff, twiddling the you know,twiddling the knobs, just
getting a little bit better.
It's way better like when youhave to rewrite everything.
And so people who are reallycurious will be able to do that.
And coming back to the Darmeshquote, your quality of outcome
will be dictated by your numberof reps you put in, which is

(41:28):
really the people hate to saythis because everyone likes to
say, well, like work-lifebalance, it's the grind, right?
Like the grind does matter.
As Darmesh says, more is more,which means like working hard
and grinding it out and learningis going to be a really
important skill set to have uhin this time.

Yeah, if you talk to anyone that's you know
been on the other side of themountain of climbing and
acquiring skills, they'dprobably say the same thing.
Of building company, just saythe same thing.
You know, you look peopleemulate and replicate the end
outcome, but we should reallyend and lead is the process.
And in that process arethousands and thousands and
thousands of reps.

(42:06):
And now, you know, it's nevermore important to learn AI, but
it's also never more fun.
It's never more fun to be acurious person.
Like this was like the world atyour fingertips of you can
build, you can create, you canlearn.
It's I think the best timemaybe a go-to-market
professional, the best timewould be in tech in general and
a burn career.

A hundred percent.
A hundred percent.
I I totally agree.
I think it's the best time.
It is the the it is the numberone time because the URL it's
like every start of every it'swhere like people make their
success, their careers, right?
Like a lot of my career wasmade by being one of the first
to adopt inbound and productlike growth.
And so people have like thesenew paradigm shifts, they they

(42:47):
reset everything and there's abunch of new winners.
And I think that is why it's soexciting because the new
winners are not based upon likeyour title or any of these
different things.
It's based upon your curiosity,your iteration, and your
ability to like really learnrapidly and and really work
hard.

Definitely.
And I'm also really interestedto see how the actual funnel
evolves to your point ofmarketing might take it longer.
It might not be a pass-off.
HubSpot actually spoke to thatrecently at inbound, right?
Introducing the loop instead ofthe funnel.
Yeah.
And just how it's not it's not alinear process anymore.
The buying process has evolved.

(43:26):
So that'll be an interestingone to see how it it shapes up.

Yeah, exactly.

And Kieran, you know, we talked about how
you learn around AI andexperiment, but are there any
Shaver books that you have andreally shaped your career of the
years?

Oh, I should probably have a good answer for
this.
It's been like I've been soengrossed in AI for like two
plus years.
I've forgotten what I even readbefore.
I honestly don't have not readmuch at all other than uh uh
work.
I I like I consume a lot ofpodcasts, I consume a lot of
like content, but I have notread a lot of books, if I'm

(44:05):
being totally honest.

You have to learn from dynamic sources.

Yeah, yeah.
Like you have to, and I I thinklike there's a time to consume
and a time to work.
And I I've really kind ofleaned into the time to work.
Like I've learned prompting,like, how have I learned a lot
of things around AI?
I have a YouTube channel, AI, Ihave a Substack in AI.
I'm building product.
I have a product coming out inAI.
I work in AI every day withinHubSpot.

(44:37):
So like I made a commitment twoyears ago that there's just no
one that will know more than AIabout me if you're a
go-to-market practitioner,because no one is going to work
as much as me in AI.
And that was like my that wasthe only goal I had.
I didn't have like any, there'sno financial or anything.
That's that was it.
Like, and so I think there'slike times in your career where
there's like a good time forconsumption where you're really
trying to figure out how do Imaster a new skill, and there's

(44:58):
just time to act.
Um, and a lot of the content Iconsume is like in the moment
where I'm trying to figure outproblems.

It sounds like, I mean, first of all, I
love that.
It sounds like you spend a lotof time learning and upskilling
around AI.
Now I have a question, veryimportant question for you,
Karen.
There's a little birdie namedBrian Halligan, give me a
tip-pop, but you're actually apretty skilled rapper.
So, how do you have someone tolearn the rap skills while
learning in high?

Uh I can't believe Brian said it was of all
the things Brian could havesaid about me, me being a
skilled rapper, how do I so so Iam uh I I all I all I listen to
is hip-hop.
I used to hang out like inthere, used to be like a battle
rapping forum where you couldbattle rap people over text.
Now, I that doesn't existanymore.
If it did, it would be kind ofinteresting because ChatGPT is a
great battle rapper.

(45:47):
I'll tell you the funnest thingI've ever did on like in in
rapping.
So Fiverr is a really coolplatform, you can use it for a
lot of things.
And me and my bros, mybrothers, who all enjoy hip hop,
we used to battle rap eachother by paying this person who
did like Sesame Street puppets.
So we would we would create theback, the back end track, and

(46:07):
then we would have the puppetrap, and then we would send my
brother would send like me thevideo, and then I would respond
as a puppet and video.
So I've done a lot of weirdthings around rapping, but yeah,
I'm I don't I don't think I'muh a very good rapper, but I'm
an aspiring.
I have started to like useChatGPT to like relive a lot of
my youth fantasies, and like oneof them was to be a builder,

(46:28):
which is which is it's helpingme do that.
And then one of it is to writeraps, but at the moment I'm just
sending them to my brothers andthey're just offensively,
they're just like offensivethings.

But my brothers incredible, incredible.
Well, we'll see.
The worlds are intersecting, AIis supporting the rap dream,
and it's all coming together.
Who knows?

Yeah, I could have a I could have a I someone
was one of my team was showingme they had a person that they
love listening to and they werelooking for concerts of his and
they were like, hey, I justfound out there's no concerts
because this this guy is AI andhe's on Spotify, he's like
really popular, and so whoknows?
Like I could maybe create alittle AI rapper.
Uh it did inspire me that Icould create a hip-hop like

(47:07):
artist and like just put him putthat person out there.

I mean, I have heard you say actually that
voice is one of the underratedutilizations of AI.
And obviously, we're probablytalking about a different
context, and we're seeing a lotof technology come out
leveraging AI for voice usecases, but that's a great use
case.

Yeah, yeah, like the Hey Jen models and
these new models, they're justincredible.

Truly, truly.
I love it.
Well, you are, as you said,already one of the best people
to learn AI from.
You are committing to apply toactually be the best, most
knowledgeable person in AI.
And already so many people arefollowing your learnings, and
you're one of the people ofhelping to shape and disseminate

(47:51):
information around AI topeople.
So for anyone, if they don'talready follow you, where can
they get in touch and follow youacross all your platforms?
And these will all be in theshow notes for everyone.

Yeah, I think the number one thing I got asked
to do was start a Substackbecause everyone was like, hey,
you share so much, can you justdocument it in Substack?
Especially the prompts.
Everyone wanted the prompts.
And so I started a Substackcalled the AI Journalist.
I've actually been amazed howwell it's gone about three
months ago.
And so that's probably the bestplace to go because I suspect
people want where are the vibeprompt and vibe uh prompt and

(48:22):
prompts you talked about?
Where are all these things?
They're all in the substack.
So that's where you can go andyou can get it.
The other one is with myreally, really good friend Kip.
We do a podcast where we coverthis stuff as well.
It's called Marking Against theGreen.
So they're kind of the two coreplaces.

Love it.
Those will be in the shownotes.
I'm a huge fan of both.
Highly recommend.
Karen, this has been fantastic.
Really appreciate the time andyou sharing with everyone.

Yeah, thanks for having me on.

Absolutely.
Thank you.
Thanks to everyone for tuningin, and we'll see you next week.

All Episodes

GTM 165: Vibe prompting, AI fluency, and the new rules of go-to-market | Kieran Flanagan

Episode Transcript

Popular Podcasts

Stuff You Should Know

Dateline NBC

The Breakfast Club

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}GTM 165: Vibe prompting, AI fluency, and the new rules of go-to-market | Kieran Flanagan