Subscribe
Share
Share
Embed
Connor Jeffers, CEO of Aptitude 8, interviews marketing, sales, and customer success leaders about how they are using artificial intelligence to innovate, optimize, and scale their go-to-market operations.
Prompt engineering,
which is something that a lot of people
should think about too and a lot of people
are just dropping in prompts but if you
start looking into prompt engineering
and it's literally 15, 30 minutes,
you can really actually level yourself
up to do things like tree of thought.
What follows is a conversation
with Nate Roybal, a distinguished
B2B sales and partnerships
leader known for turning complex
challenges into profitable ventures.
His recent accomplishments
include exceeding sales quotas and
spearheading white label product
strategies at Syncari as the director
of partnerships and GTM strategy.
Nate and I talk about his background in
sales and his career journey into rev ops
and systems orchestration, how enterprise
versus SMB buyers buy and who holds
the influence in those buying cycles.
How to become and build custom GPTs and
why for your organization, his creation
of a Syncari specific custom GPT, AI
literacy is a standard skillset people
will need to have an AI literacy versus
the rising function of prompt engineering.
Let's listen in.
Nate.
What's up, man?
Welcome to the show.
Thanks for having me, man.
Good to see you again.
Always.
I'm actually, so this is always really
fun with people that I, I know because
I know a very like narrow slice of you.
And so we get to figure out
like how did you become to be
the Nate that we know and love?
And then you're doing really
cool stuff because you were a fan
of the show.
And obviously we're friends and you were
like, Hey, I'm doing so much AI stuff.
I really want to tell you about it.
And so I conned you into not just telling
me about it, but telling the masses
Exactly.
Yeah, I'm excited to be here, man.
Thanks for
before, before, we get into AI stuff I
like to help give people context on who
we're talking to and why, and I actually
don't know like anything about you
before, how, and what I know you now.
So maybe as a good starting point
is like, what is your background?
How did you come to be in the position
you are today and have a perspective
on GTM and AI and you can start back as
part, you can go all the way back to
your birth if you think it's relevant.
I'm going to tell this story a little
differently than I normally do.
And I don't usually like to call
this out, but I think that AI
is really good for people like
me that have dyslexia and ADHD.
it's been amazing for me.
In that world, and so is a lot
of the technology advances.
And so I think that is
a big part of my story,
right?
So, when I started out in sales,
I was maybe like 23 and I kind of
dropped out of college like the
third time or maybe fourth time.
I don't know.
At that point I was kind of messing
around, found this telecom reseller job,
started picking up the phones and smiling.
What
did you
sell
and
to who?
I was selling to anybody who
wanted cable jacks or racks, baby.
It was startups in the Bay Area.
It was a random mom and pop shop.
It was some, you know, implementers
or installers, stuff like that.
But yeah, it was pretty
much anyone and everyone.
And sometimes I would just find
shit on Amazon and just ship it in.
But you know, you gotta
do what you gotta do.
And did that for three or four years and
realized I needed to use my brain more.
It was more like, Hey, color,
how fast can you get it?
All that stuff.
I was like, this is not doing it for me.
So back to school nights, got my
degree and tried to get into SaaS.
They didn't want me as a salesperson,
despite I've been doing it for years.
And so I went in as an SDR, right.?
And you know, luckily like.
This was right around the time, like,
you know, things like outreach came out.
There was really good spellcheck,
all this stuff too, right?
So just when I had to start writing emails
all day, every day, a spellcheck was
there
like emails got really easy.
yeah.
And all these templates were there.
And so like, you know, I still
can't spell very well, right?
A lot of people don't know that.
Right.
I'm a
You're inhibited by, but I always
have something checking my spelling.
So I have no clue.
I just smash words
it works out.
Exactly.
Like that's, it's been great for
me like this they told me when I
was a kid, like, you're not gonna
have a dictionary in your pocket.
And I was like, guess
what?
do one better.
Yeah, it actually writes the thing for me.
But yeah, so then, you know, I got into
SaaS and I was kind of interested in what
I didn't know what we're kind of did,
right?
So I had to scale up very quickly.
And I was like, what is integration?
What is
all
this What is...
Workato was SaaS company one?
Yeah.
It was a SaaS
company in 2017.
So that was like six,
seven, seven years ago.
Yeah.
Seven years ago now.
Basically worked my
way up from the bottom.
Right.
So it became a, an AE and then I
was doing events and then I became
right when COVID hit no more events
and I became a partner manager,
right?
And that's kind of when I think
you and me got connected., Right?
Back when I was, yeah,
about four years ago now.
I need to do like a, we need to
have like a retro on the show.
The number of people that, that start in
SDR backgrounds, I think, I don't know.
I think that there's probably a
reason everyone like glorifies and
vilifies it at the exact same time.
I was like, you have to go through
the grind house that I went through.
And I think, but I do, I feel like
most people that I know that are
doing interesting, successful things
in GTM, like at one point did that.
And I think that it both creates, grit,
but to your point, it also creates like
a level of actual knowledge of how things
work that, that helps tremendously.
And I think it depends, right?
It depends because there's
just like any other motion.
Like there are Places you go that
everything's on rails, follow the
script, follow these templates
click, and that's your job, right?
But then like, you know, Workato I
was like the second SDR basically
hired at the same time as the
first SDR, like a week later.
And so we didn't have enablement,
we didn't have tech support, right?
So I had to just learn everything.
And that's how I got really good at what
I do now, which is understanding all
these integrations and things like that.
And I was able to kind of propel me to
be successful today and kind of also get
interested in things like AI and kind of,
you know, I'm not a coder, but I
can do almost everything all the
way up to coding and maybe one day
I'll take the plunge, probably not.
But yeah that's kind of
how I got here, right?
and, yeah, AI has been really great.
So happy to talk more about that, but
And what the Syncari journey is
sort of getting to a place where
you were like, Hey, Syncari looks
really cool and interesting.
And I'm excited about
what these guys are doing.
And then, to now looking at how
AI is impacting your role and
then we'll jump into some of the
stuff you're actually working on.
Yeah.
Yeah.
so yeah, I mean, I was at Workato and
I kind of saw this gap in the market,
which, you know, was traditionally
solved by master data management.
More modern approaches.
Hey, let's get a snowflake
or a modern data stack.
And let's, you know, do this
big architecture move, right?
But there was no real answer to
customer 360 unless you could
build it all in HubSpot, right?
Which some of your customers can,
which they're very lucky, right?
But you know, people that have
much more complicated systems,
it's really hard to do that, right?
And it's not HubSpot dependent.
It's not any app dependent.
It's like ecosystem dependent.
And so, yeah, when I realized that's what
the problem they were solving, that's
a problem I wanted to come solve and
You know, we've been trying to figure
out how to solve it the right way for
the last many years, but it's been fun.
And
What do you own at Syncari right now?
So we'll fast forward a little bit
to like, what's your role today?
What are you spending your time on?
And then we can jump into some
of the actual cool AI stuff
you're working on, which is
I wanted to talk to
yeah.
I mean, it's,
I, so I, what I'm supposed to be
working on is, so what I'm supposed to
You're like, this is my job
title and this is what I do.
What I'm supposed to be working
on is partner recruitment, partner
enablement and getting partners
to source leads.
The reality is with partner world.
It's pushing a rope uphill.
You cannot force a partner
motion to move faster.
It will move it.
It's like an enterprise deal.
It's gonna move with however it's
gonna move and that's what it's
gonna, what's gonna take, right?
and so, you know, I've been spending
a lot of time coaching my sales team,
building up collateral, trying to figure
out how to talk to analysts correctly,
trying to figure out how to help
others position in our organization.
And then, you know, that, that's
everywhere from like tofu, like blog
stuff all the way down to like, Hey,
let's build this deck or this quote
and closing argument for a customer.
Right now I really like being
involved in the whole process.
Although, you know, I've got to be
careful that I don't spend too much
time there because it will draw
me
in.
everything else.
That's the fun.
That's the fun of early stage
companies though, right?
I think that one of the things that
I don't know, I spend it's, I have
such a weird perspective on this.
Cause I spend a lot of time operating
and being inside of really early stage
stuff selling to, and implementing things
for very large scale organizations.
Partnering with really on the hapily
side, like small HubSpot solutions,
partner folks, and then like managing and
networking and knowing all the things that
the HubSpot, the giant machine is doing.
And so I just, I see companies
across this huge stratosphere.
And I think what always has excited
me is what you're doing now, which
is like the small company do a
lot of things, jump into a lot of
places versus like the big company.
My role is like 10 words long, like
director of strategic alliances for like
mid market.
And it's like so narrow.
Yeah.
And I think for me, I've just always been
like, I don't know, maybe it's my own ADD.
It was like, I just like, well, what if
I did 500 other things at the same time?
Maybe like, we don't like that.
I
That's why you're a CEO, right?
You like being an all rounder.
Maybe or I'm a control freak.
One of the two actually goes,
Comme ci, comma ca, you know?
But before we get to the AI stuff,
what, something that you talked about?
And I'm just like interested, cause I
just had a trip with some friends and
one of them was telling me about analysts
and that being something that they
talk to, and in my world, analysts are
like, All public market investor folks.
And you had mentioned you think in, and
you're doing a lot of the analyst stuff.
How, what does analyst
mean to you and Syncari?
And what is your engagement with,
them
it's a tough question.
okay.
I apologize.
That's a tough question.
Yeah, no, analysts.
I mean, it could be anybody from like
somebody who's doing like a basic market
analysis for some, you know, small
website organization that does a market
analysis of some kind around a very
specific niche solution, potentially
even like, Hey, how does this work
for a leap to account matching?
Right?
Or this, or it could be all the way
up to Gartner and Forrester, right?
We were, you know, extremely expensive and
kind of, you know, pay to play a little
bit, but you know, you got to, you know,
play the game also, especially when you're
early stage, because you don't have that.
Yeah, you don't have that huge
customer base to draw from
and that huge, like backing to draw
from where you can say, Oh yeah,
we proved it out with, 500 million
in ARR that we are successful.
You'd really have to have
someone that would be like,
You
can
trust us..
Yeah, actually, this is a real
potential solution and they're not
just like selling you vaporware.
And
Do you think analysts play if you're
doing B2B and maybe it's enterprise,
maybe it's more mid markety, like at
what threshold do you think that's
even something worth investing in?
I think if you have a muddy, so like,
especially if you have a muddy market,
like, like there's a lot of players in the
market and you're having trouble defining
yourself in the market, and then you're
also like trying to punch up and punch
sideways and punch down against everybody
else around you and you have a really
hard
a big, Like a big, punching circle of
yeah, it's like a painful punching circle.
Which is sad.
imagining like, like a boss and
like an eight bit video game.
It's like a circle with all these just
punching like fists coming out of it.
I think like you can't even early,
like, but SMBs don't care, right?
SMBs really don't care at all, right?
I could be like, I'm a top Gartner player
and they're like, okay, what's your price?
What are you gonna do for me today?
Right?
Like they don't care.
right?
Like it doesn't matter.
But in enterprise or
even mid market, right.
Then the analysts are more there
because if you think about like a
mid market or enterprise buyer, they
buy so Very infrequently, right?
Like maybe once, twice a year.
They don't know how to
buy very well, right?
And they definitely cannot skill up
enough to understand the nuances of
your solution, unless it is right down
the middle of their wheelhouse and
something they've worked in before, right?
And so they have to like trust
something and they either trust you.
That's harder to do when you
get to mid market enterprise.
Yeah, because mid market
enterprise, they can't, right?
Because if they trust you and they
fall, Oh, I trusted that sales guy.
Well, you're an idiot.
But like, if you say, Hey, I'm
going to, I'm a senior IT director.
I trusted what Gartner said.
You're like, all right, well,
Gartner you know,
Gartner things.
Like you can't fire me.
Come
on.
yeah, I'm doing the things right.
I'm doing the way that I feel like
it's covering your ass essentially.
But
incumbents, man.
It's crazy.
Okay.
Actual cool.
I'm always interested in
your perspective on things.
Cause you see parts of the world
that I don't even get exposed to.
What are you working on right now?
Or what is the coolest, like, so
for context for anybody listening,
I may have told this in pre show.
Maybe I told it now, but Nate's
been listening to the show.
He's like, Hey, I'm doing
a lot of cool stuff.
I really want to talk to you about it.
And I was like, Oh my God, I'm really
excited, but I am going to record
it and share it with the world.
He graciously agreed to
come and tell me about it.
And so I'm exceedingly interested in.
What are you working on?
What's going on?
Tell me what's up.
I've been using AI probably
more than I should be.
Candidly.
I actually spend sometimes after
hours, like five o'clock, six o'clock,
I'm still just messing with models.
And it kind of started with, you
know, me messing a little bit with
ChatGPT when it came out, right?
And not long after that, we ended up
buying Jasper for internal use at Syncari
and using it a bit, right?
I never
Jasper is copy primarily, right?
Yeah, it's copyright.
Yeah, for the most part.
Yeah, but you know, it's a reg, which
I forget what that stands for, but
essentially it's like, hey, it's like
a layer in front of it where it goes
and retrieves data sets and brings
it back and uses that as a knowledge
base essentially which I sound super
knowledgeable in AI because I can't
remember what that means anyway.
So I use that for a bit, right?
But I never found it like super
helpful to what I was doing.
And so I didn't really use it too much.
I used it to test some stuff.
I also used Lavender a bit, right?
And I'm like, and then maybe
it's something here, right?
And so fast forward, try GPT a little bit.
I hear about like these
custom GPTs coming out.
I'm like, maybe I can train something.
And so I started just messing with it and
it started out with just like, Hey, I'm
doing outbound prospecting to partners.
I wonder if I can just have something
that's going to edit my emails or make
them better or just like I can drop
in a LinkedIn profile and that'll like
tell me kind of what to say to them,
right?
And this interest actually started
came about at the same time as
I started using Amplemarket.
Amplemarket
has a little bit of that and
I was like, Oh, it's not quite
there, but I see the value, right?
And so, Yeah, I started training
this GPT and I first started off
by training it on just a bunch
of emails and stuff like that.
And it did okay.
What, when you say I'm training on
these deals, is it like, these are
ones that worked make more like this.
These are ones that are
bad, make less like this.
yeah yeah,
I don't mean to just
that level, you know, but like,
say.
No, there's so much to say.
Like, like literally, like, so most
people don't realize how easy it
is to go off and train an AI model
just for any specific task, right?
You can go get an open AI
license for 20 And you can go
train a GPT in something, right?
And so what that means is essentially
going off and picking a data set, right?
If you're really good at it,
what I did was I created a GPT to
help me create other GPTs, right?
Which I trained on best practices
around building a GPT, right?
I've got one that I built to
write prompts for me, right?
But that kind of stuff.
But in general, Yeah, taking a bunch
of the emails that have worked,
taking templates and sequences and
things like that, that I feel were
successful or good writing, or at
least has a good writing example, and
then putting it into the model, right?
And saying, Hey, this is your
job is to edit emails and
tell people best practices.
It worked okay, right?
It wasn't, it was saving
me a little time, right?
And I was like showing it to one of
my SDRs and they thought it was kind
of cool, but it wasn't game changing.
I started editing it pretty
heavily right around Christmas.
I got some downtime, right?
Around the Christmas holiday.
And I started adding real product docs
to it, like getting it really smart.
And it got to a point where like,
at our SKO this year, I know none
of my senior leadership had seen it.
And, someone mentioned it to my CEO.
My CEO was like, what
are you talking about?
What's this bot thing?
Right?
And so he tries it and he just asked
it to write a blog on something like on
the spot and it pushes out this thing.
that's You know, like 70
percent it's good, right?
It's got like a bunch of, and he's
like, this stuff's not on the line.
How does it know this
stuff, all this stuff.
And I was just like, all
right, so I got Valley, right.
And now the company's using it a bit.
My SDR has got a license.
My licenses, they're all using
the bot I built to like write
stuff, to like pull stuff down,
Right.
And so I'm using it heavily.
I'm also using Claude by Anthropic because
I feel like it has better reasoning.
As soon as I actually read the article
about it having better reasoning,
I went off and checked it out.
And within a couple of days I
ended up purchasing a professional
license because it is better.
It is fundamental.
I'd say like, it's maybe like a 1.
4 X of like a GPT 4.
At the moment it's got good reasoning.
It's got ability to like take
bigger tokens, all this stuff.
So basically now what I do, I have my
Syncari product expert in GPT, right?
And so I'll run research
potentially out of Gemini because
Gemini is like the free one.
Maybe I do it out of GPT,
like a normal GPT four, right?
Like, Hey, I do this all
the time with partners.
Hey, I'm building partner
profiles right now.
So Hey, take this partner.
Here's their website.
I'd like you to list out all the
practices, list out their history
their verticals, what they focus
on and some solutions they have.
And then the second prompt is, okay,
now how does Syncari get aligned
with each of those practices and
how are we going to help them?
Right?
And so it goes, here's the practice.
Here's how some great fits.
Here's the practice.
It gives me this whole cheat
sheet on basically how we align
with their practice, right?
And I use that to help build
a business case and help me
reach out to these partners.
We're also using it for
direct sales, right?
So like, Hey, if we can drop in a Gong
call or if we can drop in like a thread
of emails plus a Gong call and say,
Hey here's how we're thinking about
this.
What do you think?
doing that yet?
Native
Gong is doing a bit of it,
but it's not,
it's.
In my opinion, it's
narrow and it's closed.
So like you can use AI with a
couple of their buttons, but it's
not like flexible at all, right?
It's,
I just, the thing that I think is so
crazy about where we're at right now
that I think we'll just look back on
and be amazed by as we're in this very,
like I, use this software product for
this, like particular workflow or thing.
And then I take information from that
software product and I put it into
the magic box and then it comes out
of the magic box and I put it back
into the software product and it's
like, that's like a huge workflow
that everyone is doing right now
across every part of their job.
And it's just like wild.
It is.
And you know, like one of the
things I talked to my CTO.
He's not going to do it because he's like,
Nate's just doing weird hacky things.
But I'm like, we don't
need you to do this.
And I could just run it through a doc.
And he's like, okay, dude.
Cause he knows I'm probably going to
change the way I do it in the next
Sure, sure.
I'm adapting these all the time,
but yeah, I agree with you, right?
It's really interesting because like
a lot of these tools are brought
in AI, but it's narrow and it's
not very powerful in what it does.
And it really, actually,
I think it cheapens AI.
I think I
I agree with you a hundred percent.
the idea of AI has been cheapened
dramatically by, you know, these kind
of onesie twosies little like query
thing approaches that really are not.
Valuable, right?
Like
the stuff I do is
fundamentally valuable, right?
Like I can build out a full
partner profile that has crazy
details in like 10 minutes, right.
With like a profile for
my executive team, right.
A profile and a sheet to
go off and pitch them.
Right.
And then, you know, all kinds of
other stuff alongside that as well.
As well as like an evaluation
on if they're even a good fit.
Right.
And like, it does all that stuff for me.
So I don't, I like save
myself hours of research.
What's the level of like cost basis
beyond your team just all having GPT
licenses for you to create something
like that and then distribute it
across your whole organization.
And like, is there, a,
The real, the reality is.
Go ahead.
we'll do that.
And then I'll ask more questions
about it because I'm just
interested in like,
what you guys are doing
it's cheap, right?
Like it's, the reality is like
for a long time, my SDR would just
ping me and he'd be like, Hey, just
run this through your model right?
Or like, it'd also be like, run
this through your model, right.?
I don't know if that's a company
that even bought a license.
I think each of these people are just
like, that's going to save me more than
20 bucks over
that alone is like, pretty crazy.
Actually, when you
Yeah.
like, I can't think of anything
else, like maybe superhuman, maybe
that
I'm paying out of pocket for mine.
I'm paying out of pocket for chat
GPT and Anthropic today because that
saves me.
Totally worth it.
Yeah.
I don't want to sit
there and like evaluate.
like, how does somebody so we
have like a centralized account.
1 of the things that hopefully none of
our software vendors, like, listen to
this, but 1 of the things that saves
a lot of money is by nature of our
work, doing systems work for customers.
Password
sharing, password folders and
password distribution is like a
part and parcel of what we do.
What that means is we get to use a lot
of software for single logins because
everyone in the organization has to know
how to navigate, find, share, access,
and manage passwords to different things.
And the side effect of that is
we're just like, I'm not going
to buy fucking 30 loom licenses.
Like I'll buy a loom license
And everyone can log into it all the time.
But so we're doing something like
that with some of our custom GPTs.
And so I have no concept of like
what the level of lift it is
in order for you to say, Hey, I
have this, I have a custom GPT.
I want
Almost no lift.
It's like 20 bucks literally.
And like, I can
No, but how did, like, if you
have a new SDR and you want to
give them access to your custom
I just sent him a link.
I just sent him a
link
is it private or is it public?
You
can choose I can choose.
I can choose to make
it so I can choose it.
It's like Google drive.
So I can be like anybody with a link
or anybody in the world can find this.
And they have a marketplace where
you can like, you know, if you do
public, it goes on the marketplace,
right?
So I wouldn't do that with
anything that like has, you know,
Product information in it, right?
But like things like the email
one, if you want to, right.
But I don't care about that,
Yeah.
Of course.
Yeah.
It's not like an issue for sure.
yeah, but my team just goes off,
spends 20 bucks and I expose the
link to them and they just go use it,
right?
and they get, you know, their own, the
usage doesn't come out of my license.
So they get their own usage on it and
it really doesn't affect me at all.
They just get, you know, as I
tune it, they get better results.
I think everyone.
So like we have two people in our
organization who are building these things
and sharing them with people and doing it.
And I think anyone who wants to have
outsized impact on the organization
that they are employed by should
go and do, create a custom GPT and
just start sharing with people.
And the price point is low enough
and the accessibility is high enough
that it becomes my assumption is
that you've just become the AI guy.
Yep.
Yeah.
It's gotten to Right
A
blessing
and a curse, I'm sure.
It is.
And I think that I probably upset my
marketing team a little bit sometimes
because the, you know, someone's like,
Oh, we gotta write a blog on this.
And I'm like, 15 minutes later, I've
got like, because it's not like I
just type in a query and it's done.
I got to do a little tweaking,
Yeah, of course, of
change it like this.
And, but like, you know, maybe
20 minutes later, I got a pretty
good, you know, I get 80 percent
a tweak this and it's there.
And you know, I don't think that
makes people feel good sometimes,
but it is what it is, right?
It's the tooling and it's super powerful.
There's like no reason you shouldn't
be using it and kind of customizing
it for whatever you're doing.
Like it doesn't take very long.
Is there anything that's not, so you
talked about like, Anthropic, you talked
about Claude, you talked about some of
these other ones you're getting exposed
to, is there anything that is, you briefly
touched on like, Ample Market, is there
anything that from a Is a GTM specific
tool you're doing really cool stuff
with that you think is like Exciting or
is it just all wrappers that are like
Today it feels like all wrappers.
It doesn't feel like
anything special to me.
every time I tried to use one of those
tools, I'm like, You know, maybe it's
a little bravado, but like I get it
and I'm like, I think I can build this.
So I don't know, you know, obviously
like when people are, where it
starts to be different, right?
Is when people have built
their own LLM, right.
And it's not a wraparound TPT
and it's a custom built LLM
to do a specific task, right.
So I recently came across a company called
call sign, really interesting, right?
They're like
What is that?
What do they do?
they're like a Reggie competitor, like
a lavender competitor, but they do it.
They have their own LLM.
And you basically can
enter personas, right?
Persona types and things like that.
And like some descriptions and
they'll go write emails, you know,
for different personas that you might
sell to with different value props.
The way most of these other tools work
is it's one generic set of instructions
and that's it for everything, right?
And you can kind of go off and customize
instructions, but it still feels
like it's a wrapper on top of GPT.
Because if you ask the wrong
prompt, it'll be like, Oh, Syncari
cares about blah, blah, blah.
And I'm like, I didn't say
anything about Syncari right?
And that's most of your wrappers.
If you train them on
something, they often spit out.
erroneous things because they're just
wrappers on top of something that wasn't
meant to do what you're doing with it.
And so even me, when I'm using like
these bots, I've built, like I'll
use the same query three times.
It'll work one, right?
Do it again.
Okay, cool.
It worked, right?
And it's like, no matter how
much prompt tweaking and how much
time I spend working through it.
You just got to know that like a
bunch of the times it's not going
to work out the way you want it to.
And you just got to tweak it.
Unless you built something
custom for it, right?
Like, like these companies are
starting to do, but it's really hard.
And it's also like really hard to
keep up with the speed of evolution.
And so I would be really scared
to go off and build an LLM because
How
do you differentiate?
So the thing that's interesting, right?
Like i'm not I'm an
analyst of all this stuff.
I'm not as much the buyer
right and I think the
thing that's
of analysts, man.
Boom.
There you go.
Logging time to this
Yeah, there we go.
The, so I think most of the AI
products that seem to be coming to
market and being loud and doing all
these things are largely in GTM stuff.
And the reason is most GTM
stuff is written and or
verbal and then transcribed.
And so it's really easy
to put into AI tooling.
And.
A lot of GTM stuff is shallow and
that you don't really need to go that
deep or do that much in order to have
something that like, look, we have
a product, we're going to sell it.
We're going to do stuff with it.
It exists now which I think
is sort of what you, Right.
Exactly.
Like so easy to spin up.
And it's kind of what you're
describing of like I, I subscribed
to this great, um, newsletter.
I think it's called super.
I never, cause that's super human.
But if you try to subscribe
to it, does something else.
I don't remember.
I can find it.
So sorry if I missed the plug for
these folks, but it's really good.
But every week or every day they have a
section of like, Oh, all these AI tools
after a couple of weeks, I'm like, Oh,
Wasn't there something else by a different
name, like two weeks ago that was doing
this and it all gets really noisy.
And I think to your point, the reason is
because it's really easy to wrap a basic
front end around AI and GPT and say, Hey,
we made this product for this thing, and
it has maybe some custom instructions
and a UI and it's nothing else.
How do you differentiate or even
like filter out the noise of.
What is BS useless stuff and
what's actually interesting.
It's actually been something I've
been, so we're You know, thinking
about doing a lot more AI partnerships
is where he is right now, right?
And partnerships is my wheelhouse.
And so I've been diving in and,
you know, regularly, I think I
found something kind of cool.
My CTO is like, that's stupid.
And somebody else has
done like four of those.
I'm like, ah, yeah, I gotcha.
So I'm getting better, right?
And I think what you kind
of have to realize is like.
All right.
Is this like, does this
look like it's a first off?
I mean, you can read what it does, right?
Does this, does it seem like something
that I could build myself potentially?
Right?
And then like to like,
what's the lift to do that?
Right?
Because there are.
There's value in something you can build
yourself as long as it's like too hard
to build yourself or it's not worth it.
Right?
But yeah, I mean, I'm looking for people
that have their own models and their
own things that they've kind of went
off and designed without, you know,
of course, like using training models
from Hugging Face and like all these
tools, like that's going to happen.
But if you're not building it custom
for something, then I really don't see
value in it unless you have built a
better way to String things together.
And so it makes my workflow better, right?
I mean, think of it like,
like outreach, right?
Like outreach was just helping you
send emails and sequence people.
But people are doing this
manually before, right?
In fact, I remember sending emails off a
document and being like, Oh, document to
Yeah.
Yeah, Absolutely.
that lift that first name on there.
No,
One of the early like sales consulting
things that we would do, this was ages
ago ages and ages ago, but post my VC
life and then doing like growth consult.
It's so funny.
Cause now I roll my eyes
at these fucking people.
And I'm like I was like,
I am these fucking people.
We were doing the growth consulting stuff.
We would literally just go and be
like, Oh, we'll interview sales teams.
We'll look at all the documentation.
We'll give you like a
Google doc of templates.
And then outreach came along and then
for a while we were like, Reselling.
I think they're doing the partnerships
thing now too, but they didn't really
grow that way in the beginning.
And we were like an early outreach
partner and we were just like setting up
outreach and putting sales reps on it.
And the difference between, Hey, we
emailed you a Google doc and like,
we onboarded your team to outreach.
We like could five X what we were
charging people for this thing.
And all we were doing is the exact
same, like you have these five emails.
yeah.
But now they're in a sequence
and they're all it's a workflow.
Now we just automated for you.
That's expensive.
Yeah.
I mean, that's a product
that I think I don't know.
I remember, I think it's
one of those things.
It's a good moment in time
where it was so impactful.
Like it really did change how everyone
was doing all of those functions and
shifted
I feel like they
of the sales workflow.
yeah, I feel like they squandered it
a bit, I feel like they, they went the
wrong direction with it, right, like,
in my opinion, we don't have to talk
about that, because it's so beautiful, I
That's another another good combo.
I don't necessarily disagree.
I, the only thing that I will say to
outreach, they think that they went into
the partnerships universe and I think
did not execute on that as well as, and
I think they had a lot of opportunity
to execute on it really well and to have
a HubSpot level universe around them.
And I don't know if that
was a strategic decision.
I don't know if it was a fumble.
Like, I have no idea.
Salesforce didn't do it either.
Right.
I mean, I've tried to partner with them
at both the companies I've been at.
And maybe like, well, there's
early stage startup work.
Workato is now like a
billion dollar company.
Right.
And I don't think it's still
probably not part of it, right?
So anyway,
I would say that the other problem
with these wrappers, right?
So when you're trying to figure out what's
real and what's not, the other problem
is like when you find a real thing,
like one of our customers is Deepgram
right?
They've been doing AI for six or
seven years around voice, right?
Just voice recognition, turning
that into text, translating, they're
adding features there, right?
But it does this, right?
It does this narrow thing.
If you need that, it's probably
the best out there to do it.
But that's the problem with these, you
know, like if you want a general AI
that can do a lot, that's a chat GPT.
But it
also,
think that people want, like, do you
think the general can do a lot is cause
like SaaS is, I don't know, maybe this
is the other interesting thing, right?
SaaS really started in the early days as
like, we'll pick a particular workflow.
We'll build a very
dedicated solution for it.
Like an outreach or like a sales
loft that does this particular thing.
And we will build a company and
build a product around that solution.
And now everyone is moving to, Oh,
we'll build platforms and the platforms
will have extensibility and operability
they can do a lot of things and
everybody wants a lot of flexibility.
Is, does that same trend extend to
AI as a technology and platform?
And we just say, we want something that
is capable of doing all sorts of stuff.
Or do I really want something that's
built for a particular use case?
I think that, yeah, the best path
is going to be multiple specific
AIs that are good at specific tasks.
And that's how I found to work
with the AIs the best too, right?
I mean, even when I'm using a
GPT workflow, I'll call in three
different GPTs to help me, right?
One will be like a research, one will
be like a Syncari expert, and then
one will be an email writing expert,
right?
And I've got each of those and they're
each, you know, one after another.
Okay, cool.
Now that's enough.
You know, generate some
potential templates.
Now pick your favorite one, right?
One of the things I can go into
prompt engineering separately, but
like one of the, like, it's, it is
like definitely a speci specificity
game I think at this point, right?
Because you're never gonna be able to
build a general AI that's as good at a
function as a purpose trained ai, right?
And you know, those will
win out in the long run.
The question becomes like,
okay, do those win out?
You know, white label motion,
whether embedded in other
tools or like standalones.
And I think there's like a bit of both
being played out in the market right now.
Like I mentioned Jasper earlier, they're
doing a big partner play where they're
working to embed their thing in all
kinds of products and businesses, right.?
Who knows?
Right.
Yeah.
I'll go the same way as
all the other software.
How do you think about, I mean, you
are in a growing startup, you're
building a team, you're seeing
the organization scale and grow.
How has some of the stuff that's
happening right now, changing your
view of what types of roles you're
hiring when you're hiring for it?
What, like how you're thinking about
spend generally, but maybe more attached
to like staffing and the team that
you need in order to get things done.
Is it less people?
Is it different people?
Is it.
Like, how does it change how
you're thinking about what a
team looks like in the future?
That's a good question.
I mean, personally, I wouldn't want
to hire anybody that doesn't know AI
reasonably okay, because I feel like
it is a force multiplier and it's
like, Oh, I don't know, SalesForce.
Okay, cool.
I got to train you on that.
So you can actually be as functional
as everybody else across the business,
Right.
And you know, maybe
you do, but it depends.
But I would say like
potentially less hiring, right?
Like, we have an an offshore
team that we work with for some
of our data entry stuff, right?
Yeah.
The other day, my VP asked a question
about like, Hey, we're about to load
this thing on the website, what's
the right codec and blah, blah, blah.
Can you go to this and reach it for
dropped it in my bot, gave it all
the details, it spat out exactly what
they should do, and I dropped it.
And she's like, Oh, thanks for the output.
Can you go double check this?
Whatever double checks.
She's like, that's
exactly what we should do.
Like, okay, right?
But it's like, okay, well that might've
taken someone three or hours before,
right?
That kind of stuff is, by the
way, I don't think anybody
wants to do those Freaking jobs.
Those are awful jobs.
Right?
Like we say that I think it's easy to
say that from an automation place where
we're like, we're not replacing jobs.
We're taking the bad jobs away.
Yeah, but then somebody
doesn't have a job.
Right?
So it's it is.
It's hard to be a good steward of
humanity and lean into AI, but you I
think you got to do what you got to do.
Right?
Yeah.
I don't think it's that bleak.
I think to your point I think there
I don't, I I think there are very
few entire jobs and functions that
can get completely replaced by AI.
And I think what ends up happening
is you have people who Are a lot more
effective, a lot more productive,
and I think it's not a zero sum game.
And when organizations can get more
done, they'll just get more done.
They don't just like, have a
static level of investment.
And I think as people get more
done you find other ways to
deploy your human capital.
And it gets there.
But I think to your point, the
skilling that is I mean, I don't
know, someone should build a, let me
GPT that for you, for me, and I'll
use it as liberally as I, I probably
still use, let me Google that for you.
But I'll very often see questions
in our organization that's like,
oh, well how could I do this,
that, or the other thing in Slack?
And I'm like, I don't know, but
I'm like, pretty sure you could
chat CPT that and it'd be fine.
And they're like, really?
And I'm like yeah, a hundred percent.
And then we do it.
And they're like, holy shit.
I'm like, I know
I told you, I do it all the time.
I mean, I use it in my personal life.
I use it yesterday.
I was like, okay, I only have bell
peppers, onions that are going to go bad.
I have to
I literally, dude, I did this
like a
it'll spit out recipes.
And then you're like,
well, I got this too.
Wish I had it.
And then like, if you do that, it
looked like this and like this quick.
And you're like, all right,
this is phenomenal, right?
But like,
That was amazing.
it's amazing for stuff like that.
But I think like, yeah, like.
Like the idea that like an agent
will ever be able to do everything
the best, I don't think is reality.
I think what we will end up having
is a group of agents that surround
each human in their personal life and
in their professional life, right?
And they are doing
Bezos is on a yacht,
just chilling right now.
And at this point, but I, God,
I feel like Alexa just had,
it was so, there's nothing.
Maybe like Siri, yes, but
like not as tied to home.
Like I just think what an opportunity
to
been trying, I've been waiting
for Google to release Gemini
for home, even though Gemini is
not.
My least favorite out of all of them
that has become like, yeah, not great.
It actually has gotten worse.
I feel like, but anyway not neither
here nor there, but yeah, I've been
trying to trying to get it on my
home and it's like, It's awful.
It's like an awful experience.
It's like, I've tried to
load it on there twice.
And like, every time I do it, like
just starts failing out on my phone.
And like, it doesn't answer
some of the questions yesterday.
I asked my phone to play Spotify.
It's like, I can't play Spotify.
I'm like the old dumb
one could play Spotify.
I don't understand.
time.
Yeah.
All the time.
But okay.
So I don't know this it's,
it is interesting to me.
Same thing, but I think it's because
it's so generally released that they're
just being very careful because you
can't today it's like, You know,
like I said, same prompt, different
responses three times, right?
Until I get the response I want.
You cannot have that at the
consumer level, which brings
in a prompt engineering,
which is something that a lot of
people should think about too, right?
And a lot of people are just
dropping in prompts, right?
But if you start looking into prompt
engineering and it's literally
like 15, 30 minutes, you can really
actually level yourself up to do
things like tree of thought, right?
Which is essentially like,
show me your work, right?
You just basically ask or like,
Hey, build a virtual scratchpad.
And work on that and
double check your work.
Yesterday.
I did one where I was like, Hey, bot,
before you send this to me, step back,
pretend you're your manager, review this.
Are you proud of what you are?
You're proud of what you put out, right?
Make it.
So you're proud of it.
Okay.
I have to tell anyone.
I have to tell anyone who's listening
to this, that like what you are
saying right now may sound ridiculous
if you are not paying attention to
this, but it is absolutely true.
And there is literal AI researchers.
That no one knows why, but if
you tell the bot that you will
tip it for good performance, it
literally has better answers.
I did one yesterday.
I, it responded and I was like, I
think you can do better than that.
And great response, phenomenal response.
Then I said, thank you.
And it's like, ah, thank you.
I really appreciate you.
And I'm like, it actually helps too.
Please and thank you to an AI model
actually get better responses.
People don't realize that either, right?
I'm getting more polite in my
writing too, which is good.
The side,
Yeah.
The side effect is I'm a much
kinder writer than another one.
I actually, someone on my team, when
I sent them the, to your point when
I sent them like a screenshot, Of
like a Chat GPT thing we were doing.
And they're like, wow, you're
way nicer to the AI than you
are to me on Slack all the time.
Well, I don't have to be nice to you.
Well, they actually find an inverse
correlation between my kindness
and your performance far better.
The opposite of the AI actually,
Well, one last question I want to ask
you before, before we run out of time,
which is something you talked about,
you mentioned both two terms that
I, one, which we hear a lot and I.
Can't decide if I think is brilliant
or cringy, which is prompt engineering.
And the other, which I actually think is
a novel, more interesting is AI literacy.
And where would you put when you're
thinking about hiring, working
somewhere, doing something like.
Is, I don't know, maybe define AI
literacy and also how important do
you think that is now and will become
I think it's super important.
I think AI literacy is
a spectrum obviously.
And it's between like, Hey,
do I know what an AI is?
And do I know how to write a very
complex prompt or a series of prompts?
Right?
And so like, yeah, literacy is probably
like, hey, do I understand what an AI
can do task wise and do I understand
how to get it to do that task for me
effectively on a regular basis, right?
That's probably AI literacy in my mind.
And yeah, prompt engineering I
think is probably going to go away.
But with the caveat that I think
most people are not used to thinking
the way computers think, right?
I'm much closer to that because I've
been at Mercado and Syncari where
I have to do decision trees all day
and think about that kind of stuff.
Right.
Most people don't think like that.
So they're like, why
is the AI not working?
It's like, because of what you told
it,
doesn't it, only knows those things a
does exactly what you tell it to do.
Right.
And if you did not do a good job
describing That's what happens.
Right?
A connection I just made that feels
inspired by what you just said is
like, oh, and I think of AI literacy.
I think about it as this, a low
bar, but I think what you, the
comparison you just made, right?
Is like, it's not that different of
a framework or way of thinking, then
like solutions architecture of what
is technology capable of and what do
APIs do and how can I connect them?
And that's not a, it's a technical
skill set, but it's not an
engineering or development skill set.
And I think, I don't know.
I'm wondering is my, am I wrong about
either how hard or how valuable AI
literacy is because I'm, cause when
I think of solutions architecture,
I think about it as somewhat like
higher on the difficulty chain
and higher on the value chain.
Whereas I think of AI literacy is
like higher on the value chain,
but lower on the difficulty chain.
And maybe that's just wrong.
Cause I'm coming from a perspective
of previously existing knowledge.
I think it's also depending
on what you're doing, right?
I mean, like some of the process I'm
doing yesterday, I'm using like two
different models and you know, like
15 minutes across like three or four
different prompts that I spent a bunch of
time working on the prompts for, right?
It's not like it's simple,
right?
But at the same time, it's
not like it's a writing code.
So yeah, I think data literacy
and data literacy, and that's a
term we use all the time, sorry.
AI literacy is is, yeah, it's
literally just like being able to
understand what's possible too.
Because like, there's a lot of times.
People go try to use AI and then
they're like, why is it not working?
And it's literally because like,
they don't, you know, whether they
don't have the grit to like keep
trying different ways until they're
like, Oh wait, I got the result.
Or they just don't understand that.
Hey, like there's nuances, right.
And they give up, right.
Or they just think it's a toy,
right?
I think there's tons of people
I talked to that are like,
Oh yeah, it's still nascent.
I'm like, cool.
I'm probably saving like 10, 15
hours a week from my team at least.
You know, like half a head count.
That's great.
We're not a big company, right?
Yeah.
Significant.
Yeah.
What last thing for you?
And then I can let you go, which is for
anyone who is, has listened to this, is
interested and excited about accessibility
for some of the custom GPT stuff.
Where should they start?
I would just start on GPT.
You know, it's funny.
I mean, the thing is most of the
AI's also are great teachers, right?
Like you can ask them
Go to GPT and say, can you help
me make a custom GPT, please?
Yeah, that's what I did my first time.
And I've asked it explain tree of
thought prompting to me explain
prompt engineering, right?
Like I use it to teach
me all kinds of stuff.
Like I level up so much faster now
just because I can ask it things and I
would explain things to me in detail.
that I couldn't before.
So yeah, I would say get started
by just getting started and
just start messing with it.
The more you mess with it, it might
feel like a waste of time at first.
The more you mess with it, the more you'll
start being like, Oh, I could drop this.
I just dropped this like entire.
picture, like it was a picture of every
single vendor at a trade show because
they couldn't pull out the list, right?
And so it was like 40 screenshots.
I just dropped those in AI.
It built a list as a CSV, gave me the CSV.
I'm like, cool, that was easy.
Would have taken me 30
minutes before, right?
All that kind of stuff.
It's like, there's some basic stuff it
can do without very much effort at all.
Just get started.
Nate, as always lovely to talk to you.
Always love being able to connect.
And I always both have so much
fun as well as learn so much.
And I appreciate you sharing
some of your time with us.
Yeah, of course.
Thanks for having me, man.
It's great.
I'll catch up with you soon.
Yep.
Sounds good.