Latest / Revenue Problem Solvers / The AI Champion at This $6M Company Is Not the CTO. Here's Why | Chris Fossenier

Transcript

The AI Champion at This $6M Company Is Not the CTO. Here's Why | Chris Fossenier · Revenue Problem Solvers

0:00
You literally have an unlimited budget to hire any individual,
0:03
living or not living to work for you.
0:07
Welcome back everybody to the Revenue Problem Solvers podcast,
0:09
where we look to crack the code on growing your business by
0:12
tackling the challenges to keep revenue leaders up at night.
0:15
I'm Carl Ortmans, and on today's episode we're going to talk
0:19
about something that is very topical and also a little bit
0:23
confusing, AI and revenue. Now today's podcast is sponsored
0:28
by the great team at Virtual Causeway.
0:30
Imagine your sales team spending less time chasing and more time
0:34
closing because buyers already know your name before that first
0:37
call. That's exactly what Virtual
0:39
Causeway does. They help B2B companies get
0:41
found earlier in the buying process by the right people
0:44
through the right channels that actually influence decisions
0:47
today. What's the result?
0:49
Well, fewer cold calls, better conversations and pipeline that
0:53
moves faster. You can visit them at
0:55
virtualcauseway.com and don't forget to mention the show for a
0:59
special rate. Now my guest today is my good
1:03
friend Chris Fosner, the Chief Growth Officer at Member
1:07
Solutions. Chris started his career as a
1:10
network engineer in Cleveland where he became Chief Engineer
1:13
within nine months and then became a partner.
1:16
He then moved back to Canada and spent nearly 20 years running
1:19
Rocket Inbound, an IT and digital agency he founded and
1:22
built from scratch in Saskatchewan, more specifically
1:26
Saskatoon. In 2013, he founded Social
1:30
Spiral on the side. It's a loyalty and and reviews
1:33
platform built on SMS and he ran that for a decade.
1:37
And this is also where I got to meet Chris and I'll say share
1:40
some opinions with him. Impressively, he had under a 2%
1:44
annual churn and decided to shut it down on purpose in 2024.
1:49
When he was ready to move on from there, he joined Vandasta,
1:52
one of Canada's largest tech companies, as the head of
1:55
special projects for an 800 person company.
1:58
It was here that the executive team chose him as their
2:01
corporate AI champion. Not the CTO, not the developer,
2:05
the operator who spent decades figuring out how to make
2:09
technology actually work for small businesses.
2:12
And today, he's focused on two major things at member
2:15
solutions, growth and AI. And in my view, this is what
2:18
makes it worth listening to the topic because it's simple.
2:21
He's never just talked about adoption.
2:23
He's actually driven it. And if you follow him LinkedIn,
2:26
you'll know he's on the water every morning.
2:29
Kayak first, then work. And sometimes it's kayaking to
2:32
work. Chris, welcome to the podcast
2:35
and thank you for joining us. Morning, Carl.
2:38
Thanks for having me. So let's get kicked off and if
2:42
you can tell everybody or share with the world a little bit
2:44
about your world. What industries, verticals do
2:47
you work with? What are the selling motions?
2:49
What size companies do you work with?
2:51
Sure. So member solutions not unlike a
2:55
lot of companies within the Jonas Constellation sphere are
2:59
fairly niche. So we deal with membership
3:03
software, mainly martial arts, but we have gun ranges, Botox
3:08
clinics and the interesting thing about this company which
3:13
really I found very intriguing when I came on board, it's been
3:18
around for over 25 years. So and some of the customers
3:20
have been around for that long as well.
3:24
So not a, a large company revenue wise right around
3:27
6,000,000. And our net revenue is there's
3:32
kind of a mandate of 30%. That's the the bar that you're
3:35
help to. So we're dealing with with SM BS
3:37
really is what we're dealing with and we're helping them
3:40
solve a problem in the membership space.
3:42
Those companies that I mentioned, you know that's,
3:45
that's how they drive their revenue.
3:47
And we have a, a special arm that helps them with invoicing
3:53
and collecting those delinquent payments.
3:56
So that's where we what we're focusing member solutions and
3:58
within our greater group we have related companies.
4:00
We have a website company that caters to the same type of
4:04
businesses and then we have a facility management company that
4:08
is more specific on recreation facilities.
4:12
So not a lot of things to help one group do all the things
4:16
you're, it sounds like you're covering all angles for those
4:19
SMBs to to make sure that they're able to run their
4:21
business. Yeah, primarily.
4:23
And the ESMBS would be a membership type focused
4:26
business. Yeah.
4:27
Nice. So how we got today's topic of
4:30
AI is I asked Chris, just like I asked everybody what's the
4:33
successful revenue initiative you've LED or an existing
4:36
challenge you're solving or have solved?
4:38
Well, it didn't take us long to know that we had to talk about
4:41
AI and more specifically AI tooling for revenue impact.
4:46
So let's kick off here, Chris. Most teams are adopting AI right
4:51
now. I think that's a fair assessment
4:54
across what's happening. What's also what I'm also seeing
4:58
is that this means that we have to change our workflows or
5:01
behaviors on how we operate. So from your perspective, what
5:04
does real integration actually look like, and how have you
5:08
helped teams change behaviors? So that's a great question and
5:14
real integration. Often people think it's like
5:17
adding a tool to the toolbox, maybe taking a, a Udemy or
5:22
something on, on how to use AI, but it's actually a patience is
5:27
probably the most important thing unless you're on, even if
5:31
you're in an, on an R&D team, there's still a learning curve.
5:34
There's still an adoption curve. I think we're almost past the
5:39
belief curve, but that was in 2023.
5:41
That was a big challenge, just people believing that was even
5:44
real. You know, they thought it was
5:46
great for like writing maybe a Limerick or something like that,
5:48
but they didn't think it was going to really help them in
5:51
sales or something. So with our team, what we do is
5:56
it's just working with the team one-on-one, very specific in
6:02
what they do. So in our revenue engine, our
6:05
team goes right from the marketing aspect where you're
6:07
generating leads and and creating content to your sales
6:12
demos. And then there's an
6:13
implementation phase and then you have your account management
6:16
phase where you try and maintain customer happiness for, as I
6:20
mentioned, 25 years with your, with your customers.
6:22
So each stage has a very specific use for AI.
6:27
It's it's not the same tool across, although you, you would
6:30
be using some of the same tooling.
6:32
So what I've found is the approach is a very hands on
6:35
approach. So I'm lucky that I came from a
6:37
technical background. So I understand some of the more
6:42
technical aspects of AI, but I've also worked with SMBs and
6:46
individuals who are non-technical.
6:48
You mentioned I was in Cleveland.
6:49
So there I was working with very technical networks back when
6:52
people had servers in their offices and I was dealing with
6:55
accountants and hedge funds and and those people, although
6:59
intelligent, they didn't work with technology.
7:01
So just being able to communicate that.
7:04
So the cadence that we have is I will sit down for an hour
7:07
session with my team and I've done this for a few quarters now
7:12
and really focus in on a problem they're trying to solve.
7:15
I'll give you a specific example.
7:17
Mary Margaret is my content writer and it took me a while to
7:21
get her on board 'cause she was on on mat leave.
7:23
But I I'd worked with her before and she's fantastic, but she's
7:27
not tech savvy. It took us almost two quarters
7:31
to get her really ramped up on on AI.
7:35
She just told me we were at a Cancun off site for my team just
7:38
recently. And at that off site, she
7:40
presented her content engine that she's created and she had
7:44
all of Q2 content done about two weeks into Q2.
7:50
And now she has the ability and the time to work on Q3 side
7:55
projects, enrich that content. And that wouldn't have happened
8:00
without spending the time with her and obviously her having
8:03
the, the interest in, in embracing AI.
8:06
So I believe that the most important thing is that either
8:10
you or you get someone that has the ability to really dig into
8:14
the problems and move beyond things like prompting or just
8:19
using. You might see it a tool like
8:22
Lovable or something like that, where people will go on and
8:25
they'll create something, but they they don't actually use it.
8:28
The true test is if you've done something in AI and you actually
8:32
use it every day to, to accomplish what you're trying to
8:36
do within in your job. And the people on my team are
8:39
involved in that, that revenue engine.
8:43
How did you get your content writer to jump on board?
8:46
Because there's, there's a lot of noise on LinkedIn, I'll say
8:50
it this way. And, and sort of in the market,
8:51
oh, AI is going to replace people, I don't need them,
8:54
etcetera. But the, the challenge statement
8:57
really comes, there's still a challenge statement of, you
9:00
know, the creativity and the thinking and how we're going to
9:02
do this and where's it going to go and etcetera, etcetera.
9:05
When you brought this to your content writer, what was the
9:08
conversation? How did you and how did that
9:11
conversation go to say, hey, really want you to start to
9:14
invest in AI and use this because it'll help you to
9:18
maximize your output to work on other other higher value tasks?
9:21
Or what did that conversation sound like for her to go or that
9:24
person's right? I don't want to say it's her,
9:25
but that person to just go, OK, I'm in, I'll figure this out
9:28
with you. Yeah.
9:30
So in this case, Mary Margaret is a her and the conversation.
9:34
So she's always been passionate about content and creating
9:38
content. So I, that was a, that was an
9:42
advantage. That's one of the reasons I
9:44
hired her is she just loves to create content.
9:46
And I, and I saw that she had the ability to create massive
9:49
amounts of quality content without using AI.
9:53
So the other thing that was that helped when when I brought her
9:57
on board my entire team, I give them a lot of autonomy.
10:01
I say, this is the goal. We we need to hit it.
10:04
It might be a number, it's a metric of some sort.
10:06
We use OK, Rs I, I don't overly care how you get there.
10:11
As long as we're on brand, we're professional and, and we get
10:15
there. That's all I want to see.
10:17
And so I explained to Maria Margaret that the way that she's
10:19
going to get there more quickly is with the use of AI.
10:22
I had already set up a, a toolbox for her of, of some
10:25
common tools for, for video content.
10:28
But I said, it's, you know, you're, you're in charge here.
10:31
You're going to use these to shape whatever you want to do.
10:34
And so initially it was just some tools.
10:37
We use tools like Vizard for video, Gamma for content
10:41
creation. And, and then, and then I moved
10:45
beyond that to a more in depth stack where actually every
10:49
single one of my employees, even even a content writer works out
10:53
of Visual Studio Code using Quad code.
10:56
And she built a team of agents. So another thing I tell my team
11:01
is that you literally have an unlimited budget to hire any
11:04
individual, living or not living to work for you.
11:10
And I showed, I showed them how to do that.
11:13
And her content engine that she created, which is a team of
11:17
agents that she communicates with, she basically acts as the
11:21
final editor because having a human touch to her content is
11:25
very important to her and also to me.
11:28
But when she presented at our, our off site, we had a few CEOs
11:33
from within the organization and she was asked to present
11:36
organization wide on her content engine.
11:39
And they want to, they want to leverage that at other business
11:41
units. So by empowering people with a
11:46
goal and then and listening closely to the problem they're
11:49
trying to solve, not just, you know, slapping a piece of
11:52
software on and say, use this, you can get some pretty amazing
11:56
results. And she's from this agentic team
12:00
that she's built. She's going to build another
12:03
team, very specific. And she's even gotten to the
12:06
point where she hooks directly into the different tools via
12:10
API. And she's not a programmer.
12:12
It's clear for people listening here that she I still get calls
12:16
from her about not being able to log into a piece of software.
12:18
So she's not real technical, but she's solved a very technical
12:22
problem, but helping people understand that you can
12:25
literally talk to AI in a conversation these days and you
12:30
can get a very technical result. The goal isn't the technical
12:34
result, it's what the technical result can do for you.
12:38
And I might have already mentioned it, but she's, you
12:40
know, created her Q2 content in about in about two weeks.
12:45
And now she can just refine, she can add layers to it.
12:48
She can host events. There's it just frees up.
12:53
She's basically doing the the work of a 5 to 8 person
12:56
marketing team right now. That is awesome there.
13:00
The one thing you said I do like is that there still needs to be
13:03
a human in the loop. One of the things that I've I've
13:06
talked a lot about is that AI sandwich, you know, the human
13:09
tells the AI what to do and then the human validates whether the
13:12
AI did it correctly or not. I think there's there's an
13:15
important thing there. And actually just kind of a
13:18
sidebar because I've got a couple of questions.
13:20
As you said, this one of the I took this AI in GTM course
13:25
through Pavilion and one of my team members, I'll say team
13:28
members because we are part of the same cohort.
13:31
He said, oh, our CEO brought in open claw and gave it full read
13:35
write access to everything. And he's a revenue operations
13:39
person. And so he, he said it that they
13:43
launched it, it went on. He was like, OK, let's let's see
13:46
what this is about. And then it was changing banners
13:51
and headlines and all these things on the website and it was
13:53
just constantly updating, iterating.
13:55
And he got questioned like what happened?
13:58
Why is this happening? He's like, I have no idea.
14:01
Like I don't know. There's no logs.
14:03
There's no This is what we did. Here's.
14:04
What we changed? He's like, I have no idea what
14:07
it's doing or what. So he was quite literally
14:09
pulling his hair out because it's doing too much, it's too
14:13
much too fast. And there is a little bit of I
14:16
think concern, I'll say it that way, concerned what I'm hearing
14:19
with others is like, oh, we went from no AI to full AI or we just
14:24
try, we're, we're trying to learn so fast.
14:27
And there is that little, there's the element of this
14:30
tooling dropped in our laps all at the same time.
14:34
Everybody's learning it together.
14:36
It's iterating at a pace that we've never seen in terms of
14:39
technology in my opinion. So there's, there's two
14:42
questions. There's there's one question
14:44
that I think I I want to ask you about Mary, did you say Mary Lou
14:48
was? It Mary Margaret.
14:49
Mary Margaret, Sorry, sorry, Mary Margaret.
14:51
So the question is, how, like, how did she learn and how did
14:54
she change her skill set to do all the things she's doing?
14:57
Because building a team of agents is not super simple, but
15:02
there's lots of resources out there now.
15:03
So how did you enable her to do the those things so that she can
15:07
create such a great output? Yeah, so the this works.
15:12
I found this works with anyone who is interested in their job
15:17
and in their career. So everyone has a problem
15:22
they're trying to solve. They might, they probably have
15:24
10. So in those in those initial
15:27
sessions, I quite literally just when I'm sitting down with the
15:31
person, I'm like, what if you could change one thing, what
15:35
would it be? And, and, and we pick that and
15:39
sometimes when I meet with people, they don't, they have a
15:42
hard time coming with an idea specifically to their role.
15:45
And I'm like, well, it can be anything.
15:46
It can be, you know, it can be a stock picker, it can be, you
15:50
know, your fantasy football app or something like that.
15:53
But the, the key is to create a paradigm shift in that session
15:57
because the, this is the paradigm shift that happens.
15:59
People are comfortable with AI in the sense that they write
16:02
prompts, they do a couple things like that.
16:05
They maybe use perplexity and but then they hear about these
16:10
things other people are doing that.
16:11
Well, I can never do that. Like that is that's the, that's
16:14
the other folks doing that. I'll just stick with my
16:16
prompting and whatnot. So in this half hour to an hour
16:20
session, I can demonstrate to them like you can literally do
16:23
anything. The, the way it's more of a
16:27
pattern, teaching them a pattern of how to use AI, just kind of
16:31
almost like scientific method. And then when, when that light
16:36
goes off, that's where you, that's where you have them, then
16:38
they're OK. I, I get it now.
16:41
And, and I reassure them like, you're not going to break
16:43
anything. We're not like, whatever you
16:45
build, we're not going to hook directly like you said to
16:47
production or you're not going to hook it right up to the
16:49
website. We can sandbox this stuff, but
16:51
play around with it, get creative.
16:55
I have show and tells with my team when they create stuff.
16:57
We meet every week specifically with my team.
17:00
We also meet with the R&D team and, and I encourage show and
17:03
tells there. So that's really how you do it.
17:06
You get a paradigm shift, you get them excited and you give
17:10
them ownership of, of what they're creating.
17:13
And because they, you know, they're a professional and they
17:16
know that they have to do their job.
17:18
They're, they're creating things that are just making them more
17:21
productive. And I also encourage all my, my
17:24
teammates to create a portfolio, like on a regular basis, add to
17:29
your portfolio of what you've created, whether it's AI or not.
17:33
Because although I'd like you to stay on my team forever, it's
17:36
important, it's important to do that.
17:38
So just that genuine assistance to help them develop their
17:42
career. And whether it's AI or not,
17:44
they're going to buy into it. And with AI, it's, it's just a
17:47
lot. It's just very exciting right
17:48
now. So the buy in, if you put in the
17:50
effort happens. I've seen a couple of memes and
17:55
IT I think they're quite funny. It's like wait until the CFO
17:58
sees their token budget compared to the headcount budget that
18:03
they had before, right? Like on that sort of parallel of
18:06
the amount of work that's being done.
18:08
So when, when you've sort of helped to, to enable your team,
18:12
where do you have them start? Do you say pick a problem
18:15
statement? So as you said, the content, and
18:19
I can't remember who said it to me on the podcast, but anyways,
18:23
they said attach yourself to an invoice.
18:25
And the, the notion there is see all the steps and all the things
18:28
you have to do on going through from I created an invoice to
18:33
getting paid. And what are all the steps to
18:35
see what, what the things are that you can either maximize or
18:38
you optimize or remove. And so I'm, I'm sort of using
18:41
that as a parallel for this of do you have somebody sort of
18:45
attach themselves to an invoice, Take the problem from or take
18:48
the situation, walk it all the way through to the very end and
18:52
say, OK, what are the things that you really don't need your
18:55
full focus? So you can do a higher value
18:58
tasks while it's doing these things?
19:00
Because you know what, it's pretty standard.
19:03
Like how did you, how do you do you use an analogy like that?
19:06
Or how did you get them to see the problem statement for the
19:09
outcome? You, you talked about the
19:11
outcome and working way backwards.
19:12
So how do you how do you enable that whole process?
19:15
Well, it really ties into that. My method anyways is more of a
19:20
one-on-one and then I then you get to the point where you'll
19:23
have a couple kind of champions on your team and then they can
19:27
help you with that one-on-one. So you turn your one-on-one to,
19:30
you know, to a 212 and a three on three kind of thing.
19:34
But the probably the key is just hiring the right people.
19:37
If you hire the right people, then they're going to get it.
19:40
Like they're going to understand like this is just the next tool.
19:44
This is the next Internet. This is, you know, this is just
19:47
something that I'm that I'm going to embrace.
19:48
It just makes sense. So we, we do have that, that
19:52
mantra. We don't believe that AI is more
19:55
important than people, but we do believe that AI is important to
19:59
get the job done. And we, we take pride in the, in
20:03
the work that we do. We've been told that our team
20:07
is, you know, a couple of notches above other teams.
20:10
So that competitiveness is, is kind of fun too, if you have
20:13
people on your team that are competitive.
20:15
So I think it's, it's more about the people.
20:17
The method that I use is. Solving a problem for them.
20:20
And once they see that they can easily solve a problem and they
20:23
have the pattern in hand on how to do it, then they like they
20:26
can solve any problem really that they're, that they're faces
20:29
because in, in revenue, the problems have been solved time
20:35
and time again. The the landscape changes.
20:37
So now you have to solve them in a, in a different way.
20:40
So we're not, we're not trying to get to Mars or anything.
20:42
So the problems aren't, you know, crazy to solve.
20:47
So they, they realize that once they have this tool and this
20:49
pattern in place, then they can then there's really nothing that
20:53
they they can't do. And yeah, eliminating those
20:57
simple tasks is, is also one of the first things that they get
21:00
rid of like Avery and sales, you know, cleaning up his pipeline,
21:04
eliminating leads that are probably less worthwhile than
21:09
other leads, maybe putting those on to some sort of an autonomous
21:15
AI driven campaign, diving into the data.
21:20
That's a that's a really big chunk.
21:22
That's a whole conversation on it on its own, I think.
21:25
But one of the things we've done for our team, this also gets
21:28
them excited is we have a bunch of sources of data.
21:32
And in the past you had to ask R&D, hey, could you query the
21:36
database and see if we have this information?
21:40
I have it set up now using MCP, so they can just ask in English
21:45
what you know the numbers from any, any data source that we
21:49
have. So they can they can go across
21:51
data sources. Hey, could you check in our old
21:56
database versus our new database on the client that has this
21:59
client number to show me some revenue information and it'll do
22:03
those SQL queries and pull all that, all that information back.
22:06
So giving them data access and empowering them like that is
22:11
just, they just come up with a bunch of cool things.
22:14
And at the off site they were in their presentations, I gave them
22:17
some guidelines, but when they were presenting, I saw a bunch
22:20
of things that I hadn't seen yet.
22:21
So they literally had created new things I hadn't seen yet in
22:25
about a three-week period of time.
22:27
So it makes your team move fast and gets them excited about what
22:30
they're doing. There's a governance question
22:32
that that's underneath of this. Yeah, that I think I, I'd be
22:36
remiss if I didn't ask like where, how do you start them?
22:39
Because if you're giving access to all the system data, is
22:43
there, what's the level of security or what's the level of
22:47
governance that you say, here's the here are the, the guidelines
22:50
you can operate in, you can kind of build anything.
22:54
Go ahead. And then do they do they have to
22:56
come to you with like a proof of concept to say, hey, Chris, I
22:59
just built this thing, whatever The thing is and it's helping me
23:02
do XY and Z and I'm able to do a little bit faster.
23:05
And is that how you take that? I don't know if it's like an
23:09
internal sort of jam session or an internal internal competition
23:18
to create something to make their lives better to scale.
23:21
Like how does that process happen in inside of your team?
23:24
So I think there were two questions there.
23:27
There was the data governance piece and then kind of the
23:30
process to create something. So on the data governance side,
23:33
my rule of thumb, because I come back from a technology
23:36
background is read only access. And then the they only have
23:41
access to whatever our R&D team has allowed us to have access
23:44
to. So we don't, we don't create the
23:47
logins and things for these databases that have, you know,
23:51
could possibly have sensitive information.
23:53
So we have a, a dev OPS team and R&D team, they give us read only
23:57
access to the things that we can see.
23:59
And so my, the MC PS that I create for my team can only see
24:04
that information as well. And it's also just within our
24:07
own secure VPN network. So none of this is, is exposed
24:12
publicly. Now they could take that
24:14
information, they could anonymize it for maybe some sort
24:17
of presentation they might do on a sales call or something like
24:20
that. So the, the security side, we
24:22
just follow the best practices and the mandate from our R&D
24:25
team. And I highly recommend that all
24:27
companies do that. You definitely don't turn on
24:31
Claude in ignore all, you know, prompts mode.
24:36
There's kind of a God Mode you can turn on and Claude, some
24:38
people turn that on because they don't want to keep hitting yes
24:40
to the, you know, am I allowed to do this?
24:43
Am I allowed to do that? I recommend not turning that
24:46
off. And then the the process for the
24:51
creativity is we do the one-on-one sessions, We do group
24:55
sessions. I encourage team sessions where
24:59
it might be a one-on-one between team members.
25:01
Someone's a little bit better in one area.
25:04
I have a this individual Isaac, he's really good with technology
25:08
in general and he's done a lot of work on some some agent
25:13
building so he can help people with that.
25:16
He's also very good at just configuring your machine to use
25:19
MCPS. And once you've done it a couple
25:21
times, it's it's not, it's not rocket science, but it is a a
25:25
learning curve. So yeah, just the the one-on-one
25:29
type stuff more so than you know, watch this video or read
25:33
this blog. I I believe in a hands on
25:36
approach because then you can dynamically change the the
25:39
meeting according to whatever questions people might have.
25:43
Once I create a thing and I tell you, hey Chris, I if I'm on your
25:48
team as the example, say, hey Chris, I just created the thing,
25:50
it's making me do X better. Is there a point where you look
25:55
at it say, OK, is there a business case effective that you
25:58
have to that you ask almost ask from from me, would you say, OK,
26:02
Carl, great, I love it. You need to tell me what is it
26:06
saving? How are you doing something
26:08
different? Has it created you more
26:09
opportunities or saved you more money or whatever the case is
26:12
like how are you looking at that to say, OK, I see how it'll
26:16
impact revenue so that we can actually have others do this.
26:19
And do you know either have a higher output, save more
26:22
revenues, make more money, save more time, whatever that that
26:26
metric is, But how are you thinking about this?
26:29
Yeah. So the the mechanism I use for
26:32
that is outside of AI, it's just using OK Rs.
26:36
So objectives and key results measure what matters is a book
26:40
by John Deere. It's kind of a system that was
26:42
built by Andy Grove at Intel, pretty common system.
26:47
And so everyone has their objectives and key results.
26:50
And if what they're working on doesn't feed into those, I just
26:53
say don't do it. Like if you want to do something
26:55
for fun, that's fine. I don't mind that.
26:57
But you wouldn't spend a bunch of time on anything that doesn't
27:00
feed into your, into your OK Rs. And if you use that as if you
27:03
have some sort of system like that to, to govern decision
27:07
making, it kind of just self heals, takes care of itself.
27:12
Another test outside of OK Rs is if you don't use it, then you
27:15
probably shouldn't build it. Like if you're not going to use
27:17
it on a regular basis to do something.
27:19
So Mary Margaret's content agents, she uses those every
27:24
quarter and probably throughout the entire quarter on on
27:29
specific things as a special content projects come up there.
27:34
We've built this thing called weekly OPS for the team where it
27:37
ties into your OK Rs and you literally just talk to in the
27:41
cloud app or if you use Visual Studio Code, you tell it about
27:45
what you're doing. It like goes to your calendar,
27:47
your e-mail, your team we use or Microsoft Shop goes to your
27:50
team's chats and it helps you compile your weekly update every
27:53
week. And so a weekly update that
27:56
might have taken you an hour now takes you a minute to kind of
28:01
put together because all of your notes and your OK R results are
28:05
in there. You do your weekly update, it
28:07
says, Hey, these are your OK Rs. It looks like these things kind
28:10
of move the needle on a couple. Is that true?
28:12
And then you know, so building tools that you use is a is a
28:16
really good way to measure it. Building something fun is is
28:19
fine, but OK. Rs and and usability are the the
28:25
ones that keeps everything on the rails.
28:28
Yeah, that that makes that makes a lot of sense and it is really
28:31
interesting that you can do all this now when so you talked a
28:34
little bit about of of also taking giving like that.
28:38
I don't know if it's read write access to is.
28:41
It read only, read only. Read only Are you?
28:43
Are you? Are you using clod or clod?
28:46
Have you anthropic? Are you You're using Anthropic
28:49
internally then? Yeah, that's the the main tool
28:52
that that we use there. There are some other AI tools,
28:55
but as far as the the main LLM that we use, it's, it's
28:59
Anthropics, Clod and my team, not everyone is 100% in Visual
29:04
Studio Code yet, but by the end of Q2, they will be for, for the
29:07
core of the things that they do. And it just, it's a little bit
29:10
of a learning curve, but there's a lot of advantages to doing
29:14
that. Eventually there's going to be
29:16
some new tools in AI for basic workflow and collaboration like
29:22
the office suite used to be. It'll be a completely AI powered
29:26
and integrated system. Google's kind of making some
29:29
headway in this. It's not quite there yet, but I
29:32
often think of it as, and we do build these strategic brains
29:36
within Claude where if you think of it like this, this call we're
29:40
having right now, if this was a meeting that would go into the
29:43
brain, you would do your weekly updates into the brain.
29:46
And then anyone on the team could just ask the our, our
29:51
brain, we call it the CXSM brain is, you know, what is Chris
29:55
working on this week? Boom.
29:56
It comes up and then it's like, oh, OK, how could that help what
30:00
I'm working on? And then it would, it would say
30:02
these things. And you can avoid having
30:04
meetings, unnecessary meetings, collaboration.
30:07
You can avoid the the creation of five documents that they're
30:12
saying the exact same thing but just slightly formatted
30:14
differently because someone likes different formatting or
30:17
something like that. So yeah, there's a lot of
30:20
advantages of it. So the the whole company or your
30:25
team, I guess says the Claude is talking to the different tools
30:29
to be able to pull those things together.
30:31
And the, the goal there, it sounds like is to remove that
30:34
administration work of let me write out my update, let me go
30:38
back to my OK Rs and create progress reports.
30:41
Who's the one that governs that? Like is, is it you?
30:44
Is it somebody on your team who actually goes and looks at all
30:48
the systems and processes that you put in place with the AI and
30:50
go, yeah, it's still on point or no, something broke.
30:54
I've got a RE, I've got a reformat, either the visual code
30:57
or I've got a refactor, the prompt.
30:59
Who's looking at that from your side?
31:02
So the things like that where people are interacting on a
31:06
daily basis, I find are are self healing because the individuals
31:10
on the team would notice right away if something's not right.
31:14
And then as far as repairing it, I have a all of my repositories
31:20
like the repository on on GitHub where you keep the code kind of
31:23
thing. We use two different GitHub and
31:28
ADO. But the for certain
31:31
repositories, I just give people access to update it because it's
31:36
not a big deal. You can, if something goes
31:38
wrong, you can roll it back. And the, the tools that we're
31:42
using are pretty much within a team.
31:44
Now, if you were to expand something like this through an
31:47
organization with thousands of people, that's a whole different
31:49
thing. And that problem will be solved.
31:52
That's just not going to be solved by my team.
31:54
So we build tools that are good for our team and we, and I
31:59
encourage my team to improve them to identify issues with
32:02
them. But the most of the things that
32:05
we're solving are are fairly simple.
32:07
It's just that they they create cognitive load if you have to do
32:11
them manually. And I want to remove that
32:13
cognitive load because I want to use your cognitive brain power
32:16
to do other things that you could never get to because you
32:20
had these admin things. So in doing so, then that gives
32:23
our team an advantage over other teams because if they're not
32:26
doing it, they're spending the cognitive load on those weekly
32:29
updates and you know, the admin things that you kind of have to
32:32
do to keep on. Track, I like that cognitive
32:35
load. What are the things that are
32:37
occupying your time that you know you can remove?
32:39
I really like you said that. The other thing that I don't
32:42
know that I've asked you, but we've sort of danced around this
32:45
with all of these updates and when your team's starting to
32:48
build things and etcetera and they bring it to light, how many
32:51
of those things actually go across the team where the
32:55
adoption is full team? Or do some of them stay just
32:58
within that one person? Because the underlying question
33:01
is, when you build all of these things, how does that change
33:06
your systems and processes on the way that your team is
33:08
operating? Yeah.
33:11
So it it would be probably the the majority of them would be
33:15
individual specific because their roles are unique.
33:20
One thing, and we're already seeing it is that as a result of
33:24
this, instead of having five people on a team, you're going
33:26
to have one because that one person is running up effectively
33:30
a six person team with, with AI agents.
33:35
There are some that go across the, the entire team.
33:40
And I think that that's fine. The some people get worried
33:44
that, well, there's going to be all this stuff and, and we can't
33:47
keep track of it and like, who knows what's going on?
33:49
But that's no different than when everything was kept in a
33:54
sheet or Excel spreadsheet were a doc since the beginning of
34:00
time, when there was teams working together, there would be
34:03
a a meeting. They would discuss something in
34:05
the meeting. The the notes would be in a Word
34:08
document, but three other people be like, you know what, like
34:11
that meeting. Those meetings were OK, but I
34:13
can make those meeting notes better.
34:15
And then they would put it in another document And pretty
34:17
soon, you know, you get an Ant or person company and they're
34:21
operating out of five spreadsheets that are people
34:26
think they're all the the main spreadsheet, like there's a
34:29
there's five pockets of people that think this one spreadsheet
34:32
is the right one when when it actually isn't.
34:33
And so it's, it's always been a challenge.
34:38
I think AI will will help rein that that in a little bit.
34:41
But I'm fine with individuals having their own their own tool
34:44
set. It's not like it's not like you
34:48
know, a professional hockey team in the NHL, they certainly
34:50
follow a system and they're, you know, there's a game plan, but
34:54
individuals have their their own tools and it just makes the team
34:58
makes the team better. Like the like McDavid, he
35:01
certainly got a different toolkit than the Macar, right?
35:04
But both of them are doing phenomenal things on the A.
35:08
So I look at it, I look at it that way.
35:10
I there obviously is responsibility within the team
35:14
to properly document. So as an example, using a tool
35:19
like GitHub. So if you create a tool, it's
35:22
very easy, especially with AI to make sure that that GitHub
35:26
repository is very well documented so that anyone could
35:29
walk into it, clone it down to their machine if you were to
35:32
give access to them and they could figure out how to use it
35:35
and they could get it up and running with the with the help
35:38
of AI. So there is, I think it is if
35:40
you put structure in place to make sure that things aren't
35:43
getting unruly. And if you also have that check
35:46
metric of it should feed into your OK Rs And if you're not
35:49
using it, kill it. It it's kind of a self cleaning
35:54
ecosystem. That's awesome.
35:58
If I just looked at the time and I said I'd get you out of here
36:01
in a reasonable amount of time. But, but if I ask you this
36:04
question, if a sales or marketing, you know, if a sales
36:10
or marketing team built needed to build an AI stack from
36:16
scratch, and it sounds like that's what you've done is
36:20
effectively trying to create this AI stack or giving people
36:23
the tools, where would you advise A-Team?
36:26
So somebody listening to this, where would you advise A-Team to
36:28
say, hey, start here and build, build on it.
36:33
Like is there, do you start with an LLM?
36:36
Do you start with an ancillary tool?
36:37
Like where would you suggest somebody to start to try to even
36:40
replicate some of the things that you're doing today?
36:45
Well, I could give you a link to a vibe stock that I recommend
36:49
that you could share out the but you know, there's there's
36:55
there's there's a a million different ways to kind of peel
37:00
the onion kind of thing. So I'm a little I like to be as
37:05
close to the the technology as possible, kind of like the raw
37:08
format. So that's why my team is in
37:11
Visual Studio code. So when they first saw it,
37:13
they're like what like because it looks like something from
37:19
this is dating me, but something from the movie hackers or
37:21
something like that. So like the there's like a
37:23
terminal screen and it and it's hard to, but then I show them
37:26
like it's literally the same thing as the cloud app.
37:28
The cloud app just looks pretty like they just have it's just
37:31
nice and it has categories and stuff like that.
37:34
But you can you can maintain that in Visual Studio Code.
37:38
So an LLM, certainly an IDE like Visual Studio Code, it could be
37:45
cursor, it could be something else, are the main things.
37:49
And then you would have a suite of MCPS depending on the, the
37:52
tool set that you use, you might create some custom ones.
37:54
So if you're, if you're using Google, they have a set of using
37:58
HubSpot, they have an MCP Office 365 does.
38:02
And then there's other tools you might build them for like for my
38:05
marketing team, I built an MCP for Google Analytics in our
38:08
Search Console. So Mary Margaret can just say,
38:11
hey, how have the last 10 blogs been doing on our website?
38:14
And it'll come back with, you know, all the analytics and
38:16
stuff. So MCPS are important.
38:20
And then there might be some specialized tools that either
38:24
they're mandated by your R&D team or you've picked them out.
38:26
So I mentioned a couple on the marketing side wizard for, you
38:30
know, quickly generating short content video.
38:34
You would want a tool like Veed where you can do more manual
38:39
editing to make something custom.
38:41
Gamma is a great tool for creating documents.
38:44
You can connect through an API. We use something called Publar
38:48
for our social posting. That's a pretty powerful tool.
38:51
You can also connect to that directly through an API.
38:54
So you can have Mary Margaret's agents writing their content and
38:59
pushing it right to the to the platform.
39:02
You'll have AI tooling within your CRMS and things like that.
39:07
So I think it's there. It's going to vary from shop to
39:11
shop, but I would recommend getting very comfortable with
39:16
getting close to an LLM, getting close to a Visual Studio Code
39:19
type thing, because that that does matter until they get it to
39:24
the point where there is a tool that is like that, that is more
39:30
abstracted. So right now you can talk to
39:32
Claude and voice and create apps with just talking to it.
39:36
So that's that's one layer of it.
39:39
But it's going to get to the point where you basically take
39:42
your idea, you drop it into the AI ether, whatever that is, and
39:48
it'll come back with some options for you.
39:49
So this is the idea that I have and you can almost do that right
39:52
now. It might just be a couple
39:53
prompts, but it'll come back with.
39:57
You could use this tool, you could use that tool, you could
39:59
use whatever. In a year or less, maybe it's,
40:03
it's just going to do it. It's just going to piece all the
40:05
things together. It'll be like, oh, here, if you
40:07
want to do this solution, here's the subscriptions you're going
40:09
to have to pay and then you'll be able to do that kind of
40:12
thing. So I think have someone, have
40:15
someone technical on your team is important or have access to
40:18
someone technical. If your team doesn't have a
40:21
really good grasp on things. A good temperature check is if
40:25
if there's someone on your team, that is, if there isn't anyone
40:28
on your team that can use Visual Studio Code and create a simple
40:32
app or integrate with your platform or your CRM using
40:37
Visual Studio Code, that probably means you don't have
40:39
someone technical enough. So get someone that that that
40:41
that's that is that technical, and that'll just accelerate your
40:45
game massively. I think there's another topic in
40:49
here for us to cover. It's like the, the amount of
40:51
time that we can you, you know, invest in how are we, how are we
40:55
going through and, and actually creating the space for us to do
40:58
these things and etcetera. So maybe there's another topic
41:00
in here for us to cover, Chris, if I can, if I can suggest in
41:04
the future. But I, I realize the time and,
41:07
and as we go to the wrap up, I typically like to ask everybody,
41:11
you know, everybody likes new. Nobody likes different and doing
41:13
both at the same time, like AI and happening can either cause
41:19
friction or uncertainty. So for somebody trying to
41:21
improve their AI stack or even the way that they're adopting AI
41:24
inside their own group, what are two or three actionable things
41:27
for you to do for you to tell them on Monday morning?
41:31
Start with these two or three things and that will put you on
41:34
the path for, I don't want to say AI native or AI fluent, but
41:39
AI curious. I'm not sure what the word is.
41:42
Well, if you're not already doing this, make sure that
41:46
you're having fun. Like that's the number one
41:48
thing. So everyone on your team is
41:51
going to be trying to solve a relatively different problem.
41:54
So make it focused and make and make it fun.
41:59
You know, let encourage the team to know that like this is AI is
42:03
not replacing you. This is not a boring homework
42:07
assignment. We're going to have fun together
42:09
and we're going to, and we're going to be able to do the
42:11
things that you want to do much more quickly.
42:15
That's super easy to do to have fun.
42:17
And hopefully everyone is already having fun on their team
42:20
#2 I would have a structure weekly session, at least for a
42:25
period of time, a couple of quarters where your, your focus
42:29
with your team on very specific fundamental AI learnings that
42:37
can apply to everyone. So that would be, you know,
42:40
setting up your environment with a vibe stack so that you can,
42:43
you can use these different AI tools and then solving a very
42:47
specific problem that is meaningful to the entire team
42:51
and paints a picture of like gives them an aha moment.
42:55
But doing that in a group session is great.
42:57
It saves time because you're all together instead of having five
43:01
different meetings. And it allows people to shoot
43:04
questions and get comfortable with the their level of AI
43:08
awareness and, you know, making sure that everyone understands
43:13
it's OK if you're not an AI expert out of the gate.
43:16
And then maybe have some competitions with your team,
43:21
Just challenge them to come up with an idea that would move the
43:24
needle. You probably have a A1 pager for
43:27
your team, like a high level. This is what we're doing this
43:30
quarter. And, you know, challenge them to
43:34
to have an impact on that. Bring it to that session.
43:37
Those are pretty 3 pretty easy things to do that don't even
43:41
require a specific technology. I think the technology choosing
43:45
is, is a secondary thing. I'm a huge cloud fan and like I,
43:51
I just, that's just where I'm where I'm at.
43:53
But if you like Gemini or some of the tool, that's fine too.
43:59
I do agree pick the one that you like.
44:02
I also realize, you know, going through and testing a bunch of
44:04
these different tools, they all do something different and they
44:07
they're a little bit stronger in certain areas, right?
44:09
Like a perplexity. If you want to do deep research,
44:11
it's a little bit better than summer or maybe ChatGPT from a
44:15
idea creation. But Claude is the one or
44:18
anthropic is the one that I seem to be going back to more and
44:20
more. And then even like Gemini, like
44:22
their their nano banana and image creation is really good,
44:26
right. So they all do something
44:27
different and you have to kind of pick and you know, play
44:30
around with. I do, like you said, have fun
44:31
because if you have fun and you can start to play around with it
44:34
and you're genuinely curious about solving something, there's
44:37
lots of tools out there. But hey, Chris, this has been a
44:41
lot of fun. And for those that don't know
44:43
you, where would you suggest they find you or or if they want
44:47
to talk to you a little bit more, where would they go?
44:50
So I'm on LinkedIn and yeah, so there, that's my LinkedIn link.
44:54
Feel free to reach out a virtual coffee if you want to do an hour
44:59
long AI session to crack the code on that until you open to
45:02
it. And then if you want to just
45:04
book right in my calendar, you can go to exten-x.com.
45:07
And that's where folks usually connect with me when they want
45:09
to, you know, they have an idea for their team and they just
45:12
want to sit down and kind of jam through how they might present
45:16
that and, and, and have an effective launch with AI.
45:19
Awesome. I'm I've known you for a long
45:24
time. I'll say it that way.
45:25
And it's funny how our relationship is sort of evolved
45:28
and adapted over over the years. And it's, it's you're 1 where I
45:33
can ask you a question and you'll always make time for me.
45:37
And I very much appreciate it, even though our, our passive
45:40
have been sort of parallel, but totally different.
45:44
And every time we jump in, it's like we jump in and have a, I'm
45:47
having a conversation with somebody who I just saw last
45:49
week. And I really appreciate you, you
45:51
doing all that. And also helping, you know, to
45:53
shape me a little bit too, because our conversations have
45:56
been all over the place, but there's typically learnings that
45:59
happen. So I really appreciate you doing
46:01
this with me today and talking through it.
46:03
And I do hope there's another conversation because I think
46:05
there is one that we need to talk to.
46:07
Yeah, no, I really appreciate Carl and we'll have to get we're
46:10
getting closer to the Stanley Cup, so there's something to
46:12
talk about there too. So it's.
46:14
Always hockey to talk about it seems like.
46:16
So, yeah. Yeah.
46:18
This has been a lot of fun, Chris.
46:20
That's going to wrap up another episode of the Revenue Problem
46:22
Solvers podcast. Obviously, want to thank Chris
46:24
for sharing these invaluable insights and some strategies you
46:26
can put into action today. Remember, success leaves clues
46:29
and every revenue problem has a solution.
46:32
Sometimes you just need the right conversation to find it.
46:35
Until next time, it's Carl with the Revenue Problem Solvers
46:37
helping you turn your obstacles into opportunities.
46:39
Thank you for investing your time with us today.
46:41
I'll see you in the next episode.