76. AI Is Revolutionizing Distribution Strategy! What You Really Need to Know (with Brooks Hamilton)

Karthik Chidambaram: We are joined by a very special guest in Austin, Texas,
Brooks Hamilton. We're going to talk about AI strategy for distributors.

Brooks, welcome to the Driven podcast. Great to see you in Austin and
welcome to the show.

Brooks Hamilton: Thank you, Karthik. I've enjoyed so many of your episodes
before. Learned a ton. Very excited to be here.

Karthik Chidambaram: So Brooks, in preparation to the podcast, I was reading
up on your work. And I read your Medium blogs, and I was fascinated by it.
The different things you wrote on AI, Microsoft, Google, OpenAI. I thought
it was really, really cool. So thank you for writing.

Brooks Hamilton: Thank you.

Karthik Chidambaram: So let's get started with AI strategy. I mean, AI here,
AI there, AI everywhere, and that's what we are hearing. Sometimes it's
like, hey, what's really happening? You know, it's just confusing, right?

So, I mean, I'm really glad we are talking to you today because every
company needs an AI strategy and especially distributors. That's who we
primarily work with and you also do a lot of work for them. And they need an
AI strategy. They're not tech companies.

So what is an AI strategy and how do I go about, let's say I'm a
distributor, what is an AI strategy and how do I go about it?

Brooks Hamilton: That's a great question. I think AI strategy is going to be
one of the most fundamental aspects of distribution, as well as large
sections of the rest of the economy, over the next five to ten years. What's
a little bit different about an AI strategy as opposed to some of the tech
strategies that we've thought about previously, so as we moved to mobile,
when we had to e-com, is that the pace of this is moving significantly
faster than those prior waves.

Luckily, I've had the opportunity of watching both of those. And even as a
very tight watcher, the field's moving very quickly. So I can certainly
understand how somebody who does not look at this day in, day out, as my
team and I do, can really get overwhelmed by the amount of media that's
heading towards them saying AI, AI. Really what it comes down to for
distributors, is to understand, where do I have opportunity to drive
significant sales increase and where do I also have those opportunities of
taking costs down in a way that's going to yield both better margins as well
as potentially and unusually increased customer and Employee satisfaction,
which usually we needed to give on one of those aspects before.

Now we find we can actually go execute on all of those at the same time and
have a positive outcome. So, I think that's a little bit like the backdrop
and the opportunity that we see out there. But, when we think about an AI
strategy, it's more akin to, Where do I get started? How do I build the
message?

Both internally for my executive and larger employee team, but also, how do
I go about building the technical capabilities and organizational
capabilities that will allow me to compete in the market going forward with
my traditional set of competitors as well as potentially a new set?

Karthik Chidambaram: Let's say I'm a distributor, right? Let's say we do
about 50 million in revenue, 100 million in revenue.

Brooks Hamilton: Yeah.

Karthik Chidambaram: So what do I do? Why do I call Brooks? Hey Brooks,
everybody's talking about AI. So let's say you come in. You come in. What do
you do?

Brooks Hamilton: Good question. There are usually two big aspects that we
work on first.

One is, do we have a common set of understanding of - what is AI? And where
is it today? Usually when we look around our executive team, there may be a
few people who are saying yes, we've really got to jump on this. Another set
of people saying, I don't really know what this thing is. And then, maybe
still another that says, I think I know what this is, and I don't think we
want any part of it.

So, at least trying to get that team all on the same page so that they're
working off the same set of facts, and then drive towards a consensus about
what do we do now that we have these facts, is step one. What is the kind of
thread of projects? That we're going to go work on that yield both a great
financial ROI for us as well as setting up that communication within the
organization of how are we going to go do AI as an organization.

A little bit more to say about that, but maybe we can get to that in a bit.

Karthik Chidambaram: So how am I going to use AI, right? So you talked about
strategy and the key thing is I was actually chatting with a customer of
ours, you know, this week in Dallas, right? I was meeting with a customer.
He's also a friend, Lijesh Shetty. He's a technology director, and he was
asking me, hey, what are you guys doing on AI? And I told him, we are
experimenting a few things. And I was just talking about it. But he asked a
great question. Hey, I'm really glad you're experimenting, but how is your
experiment going to impact my business?

So, experimenting is one thing, and experiments lead to results, and that
essentially has an impact. But one question a lot of distributors have is-
you know, one of the things he talked about is, hey, you know, you need to
drive down cost, or you have to improve efficiency. It needs to be tangible,
and we need to be able to see it. So how would you answer Lijesh?

Brooks Hamilton: Yeah, first what I think about is, how are you going to go
about using AI in your company is a big deal. There are a lot of different
directions and areas in which you can go use it. But, there needs to be a
guiding principle behind it. And ideally what that should be is, you should
have great alignment with what is your core value prop that you're bringing
to the market in general.

And then, how do you want to align AI within the market, within your
organization? So, a quick example there is, do I want to make AI something
that my customer base interacts with regularly? Or, is it something that I
only want the internal portion of my organization to interact with
regularly?

There are a lot of different ways of applying that. Each organization is
going to try to figure out the one that's right for them. And then the
second part is, how does it match up with my value structure? So if I'm an
organization that is extremely financially KPI driven, then I need to push
for those cost changes and revenue put and revenue growth in order to meet
those financial goals.

There are other organizations that may be family owned, employee owned, have
long histories of engagement in between ownership. and employees where the
way in which it's going to be used in the organization is going to be quite
different. The way in which they may decide to retain employees, to re
skill, it is a very different story there.

So figuring out right from the beginning. One, how do I enhance my value
proposition to the market? And two, how do I stay true to the value set that
I've had as an organization? is going to matter a ton. Because the first
couple of projects you go do as an organization is going to communicate an
enormous amount about how you see this playing out.

Because the advent of AI is like outsourcing to third party countries and a
technology revolution wrapped into one. Now we can kind of imagine that's
going to create A lot of emotion for people who are in the organization.
What's going to happen to me? What are the expectations for my job? Am I
still relevant?

They'll, at every point in the organization, they will ask themselves that
question. And then depending upon the answer or their perceived answer, they
will decide to stay or to walk. And so it's excellent to have a plan in
place about how you're going to approach that value prop and your value
structure.

Before you ever go launch into what a neat application of the technology may
be.

Karthik Chidambaram: Yeah, really determine your values first and based on
the values, determine the strategy and then go ahead and execute it. You
also talked about people. And during lunch, we were chatting about one of
the distributors who is doing so well.

They had a great culture. They have a great culture, but then they are
implementing AI. And people are very happy about it, right? Can you talk
about that? What exactly is going on there?

Brooks Hamilton: Trying to identify the areas of people's roles, inevitably
a chunk of work is crucial to the business and the operation of the
business, but nobody likes doing it.

And every role has these, but sometimes there are entire roles defined for
these. We spoke to one electronics distributor. We were talking about part
crosses, so, or part matching in between, say, market basket RFQs that come
in or price updates from vendors. And what they shared with us was they have
a hundred people dedicated to part matching, which was shocking to me.

And what we recognize is in smaller organizations, part of somebody's job is
that they don't enjoy it. So go find parts where people really enjoy. the
other parts of their work that they may even have certifications and
training on, let them do that part rather than compare one cell to another
in Excel as a way of passing the time at their, their job.

So there's an enormous opportunity here of finding the areas that are drags
on overall productivity and key KPIs. And really go about alleviating those
to drive that progress and getting rid of a lot of that busy work that's not
actually adding any value to a customer transaction.

Karthik Chidambaram: Brooks, we talked about strategy, but then execution is
the key thing.

So how do you go about executing it? Let's say you come in and we come up
with a strategy and we understand. So you talked about patterns. You find
the pattern. Let's say people don't really like to do the task. And then you
take that, you try to automate it. And what do I really love doing? You
talked about certifications, right?

I might be really good at it. So let me focus on that, but then whatever I
don't like, find the pattern and automate it. Great strategy, and I think it
looks great, right? But then, how do I go about executing it? Actually, do
that thing.

Brooks Hamilton: Yeah. That is what so much of this year is going to be
about in 2025.

So, let's kind of go back. We're all familiar with the major software
producers that are out there in traditional software. You know, we know our
Prop 21s, our, you know, SAP, Microsoft, so on. There have been a new set of
vendors that have come to the fore. Related to A. I. Infrastructure. So
those are the most popular one.

Open AI. As we all know, with our buddy Sam Altman. And then Google is a
major provider as well. And then finally a company by the name of Anthropic
that's done some great work. All three of these vendors had held back a lot
of their release of new capabilities until the end of the year and the new
year.

And what we see now is just an absolute wealth of capabilities that are now
provided into the market. So, for example, there's one called Computer Use
that's offered by a number of these where you can simply say, listen, I
would like for you to take the invoice that is in this email and I'd like
you to match that against the prior invoice that we received as well as a
general ledger entry.

Great. Now the application can pull open my email, take a look at that
document, go over to the application. Take a look at this, say, within Infor
or whatever application I'm necessarily using, Oracle Financials, and do
that matching activity, all without me having to walk through that process.
It can instead do it itself, using the set of tools that are available today
that a person has, rather than having to entirely redo their technical
suite.

So if we think about it, there's one layer of applications, which we're all
very familiar with from our standard vendors. And we're beginning to see the
start of a new layer of capabilities that are able to use those original set
of tools in much the same way a human would. To complete those tasks,

Karthik Chidambaram: So as a distributor, as you rightly pointed out, I'm
better off using the existing tools and technologies rather than trying to
figure out something on my own. Would that be the right approach?

Brooks Hamilton: So the first thing that we want to understand is what kind
of capabilities do we have as an organization? If I'm coming from, say,
somebody doing 10 to 50 million, chances are I'm really going to focus a ton
on my sales organization, my warehousing organization, delivery, and
probably not quite as much on my technology organization.

So in that case, I'm really going to lean on other vendors, tool vendors,
and then potentially also Vendors to come in to help me set this up if I
need. I'm not going to go try to do it myself. Whereas we know other
organizations somewhere in the hundred million and above may have a more
robust IT organization that can try these things out themselves.

And we know that there's a broad set of tools that are out there. to begin
to use. These tools are materially easier to configure than what we had in
the prior realm of technology and technology providers.

Karthik Chidambaram: But how do I know which tool to use? There are a
zillion tools out there. And how do I know which tool to use?

Brooks Hamilton: One, I think it goes, there are a few pieces and this is
really the work we do to try to help our clients figure out the right path
for them. Oftentimes, it has to do with- where are they in their installed
base of applications, their current technical stack? So if I'm a big
Microsoft user, maybe that's who I'm comfortable with in terms of where my
data is from a security basis.

Similarly, if I'm a Google customer, maybe that will make more sense, but
there are also an entirely new set of tools that are coming to market like
Bland AI, Happy Robot, and literally hundreds of others. That have started
to pick off very specific use cases within the distribution industry to try
to help with.

And so those are things like order entry. Customer service, great things
that kind of span, not just the distribution industry, but accounts payable,
accounts receivable, financial tracking, and then some really neat
applications for how you can create technology in ways that are far simpler
before. So, I think that all those are the ways that we see distribution
organizations picking some of the more mature vendor areas.

Karthik Chidambaram: Yeah, I think small wins are very important, right? You
have small wins, then that motivates the business to invest more to larger
projects. So in your experience consulting with multiple companies, multiple
distributors, can you talk about, can you share a story where You went on to
a distributor, you did some strategy for them, and there was a small win,
you know, it's not big.

And then, that really built confidence, and then they are ready to invest
more. Can you share some real examples, you know, which made a difference?

Brooks Hamilton: Yeah, I think one had to do with a rebating application.
So, they would receive rebate information from a major vendor of theirs.
They would receive that on a monthly basis, it was usually a big raw file,
they would have to then go through multiple activities of trying to get the
information down right, and to the minimum amount possible, to then create
the hundreds of credit memos that would then in turn be sent out.

Back to that same vendor to go about collecting on rebates, which in a sense
like that's great That's how it kind of should work But it was costing the
multiple people of time to go capture this benefit. So there was a real cost
associated with it and instead what we said we looked at it, we thought,
wait a minute. We can just collapse this whole thing down.

So that this process is done very quickly so we could use just a typical set
of automation slash ETL tools. And for some of the gray business logic
parts, that's where we could go about inserting AI capabilities in really
focused ways. And then bang. Now the process works more quickly.

Those people do not have to spend time on it. It's just a simple review
function. And now that rebate activity has gone from something that costs
them money to make money, to actually enjoying much more margin on their
rebate program, which is fantastic. That's what they would really want from
it originally.

Karthik Chidambaram: Yeah, I love this. You know, I hate filling out rebate
forms. I don't know anybody here and I want to stay away from it, but I
can't really relate to how I can make a big difference here. You know, I
think that's a great example. And once they saw that, then that builds their
confidence and then they want to do more. Is that typically how you see it?

Brooks Hamilton: That's exactly right. We're also big proponents of just
take small steps. If all you want to do at first is have some of your core
corporate team become excellent at using a language model, great. Go do
that. Go capture the 10 to 20 percent productivity improvement that they're
going to get out of 20 bucks a month and start there.

If you instead want to try out like a small project, fantastic. That is the
type of learning that needs to occur within these organizations. And the
reason that I use the word need there is that many of them don't recognize
that they're on a clock. This is not an opportunity where they might be able
to use AI capabilities for improving top or bottom line over the next five
to ten years. This is, instead, going to be something that happens quite
quickly.

And they need to begin to experiment to find out what works for us, what
doesn't, and what creates a positive relationship with our customers. Also,
what may generate a negative response from our employees. And how do we
balance this out in terms of developing this new capability set, which
really nobody can do.

Karthik Chidambaram: And everybody's talking about AI being a wave, a
monstrous wave, or some say it's a tsunami. Again, all that is true, or but
then let's say, how, how big a team do I need? You know, so let's say if I
want to start implementing and testing the waters. How big a team do I need
as a distributor or as a manufacturer?

Brooks Hamilton: Of course, there's going to be quite a bit of variation
depending upon the size of the organization. Let's go back to that, like I'm
somewhere between 10 and 50 million dollars a year. I would say don't invest
in a team at that point. If you have an IT person, really ensure that they
deeply understand what is going on with AI.

But for teams that are larger, we think that this is something that needs to
translate into an AI center of excellence. And that is something which
should be staffed with individuals that have a technology background, but
also a deep understanding of what it takes for successful change management
in an organization.

Since the change management aspect is likely going to be the part that is
more crucial to success. compared to project management or even technical
capability at this point.

Karthik Chidambaram: But let's say I am working in a warehouse or I'm
working for a distributor. What AI skills do I need as somebody working in a
warehouse? Or how should I be approaching AI?

Brooks Hamilton: The thing that I probably want to keep my eye on is the
state of robotics, which is moving very quickly at this point. If we think
about, like, what was so difficult about robotics before, it's programming
every single little activity. Well, what these language models have given
us, and what the last round of AI has given us, is a way of translating
intent to software.

So, the robots, or at least the language models can program the software
dynamically for these robots to take action. And what that means is the cost
of producing these robots, the cost of software, the cost of training, is
coming down by orders of magnitude every six months. So we expect to see
things like 100, 000 units available towards the end of this year.

That will be probably leased back for, say, 25, 000 to 30, 000 a year of
robotic workers that are available to do quite a bit of work for about 20
hours, 18 to 20 hours a day. So, if I'm working in a warehouse, I'm going to
keep my eye on that. On the flip side, if I'm working on the corporate side,
what I need to do is really assess my job skills and understand how am I
actually adding value to the business in terms of being highly contextually
aware.

So what is going on with my clients, the market and the company? And then
second, how do I go about creatively approaching? Certain questions, because
the value is not going to be in how well the email is written or how many
pages the policy manual is, or how in depth the process diagram is.

Instead, AI is going to do all of that for you, but AI needs direction. And
it needs creative, context aware direction. So, as I begin to build my
skills as somebody working on the corporate side, I need to think about how
I am making that difference. How am I going to use AI to get better at my
job? And then, how can I go about pushing it?

To solve problems that I might not have even seen as problems before.

Karthik Chidambaram: Yeah, on the corporate side, be better at prompting,
right? So, give better prompts, you get better results, and work through it.
I think that part is pretty clear, but on the warehouse side What do I do? I
mean, do I upscale myself or where do I upscale or should I look for another
job?

Brooks Hamilton: Yeah, I think these are going to be some really key
questions we're going to start working through. In terms of 2025, I don't
think that we're really gonna see a ton of impact there but there will be
the question of if I'm working in a role which all the tasks in that role
can be automated or 90 percent of them can be automated with AI, I may need
to look for something else on the flip side if I'm an owner, let's say we're
two, three years down the road from today and I'm going to choose to divest
of part of my accounts payable department that has done that work for 30
years.

There may not be any account payable roles to go do in the economy. In three
to four years, how are you going to go about responsibly working? So this
goes back to the question of my core values as an organization. If my core
value is achieving the financial KPIs at all costs, I have an answer. On the
flip side, if I'm an organization that is employee owned, family owned,
where there's a very different relationship between the two, then I might
need to give some thought to how am I going to handle this as an owner and
what signal is this going to send to everybody else. Who may not be impacted
now, but they may be concerned about being impacted later.

Karthik Chidambaram: Yeah, have an owner's mindset. So be more transparent
about the whole thing. Hey, this is happening. Or let's say if they bring in
someone like you, you not just work with the executives, but for the entire
company, right? So maybe you can give a session, so people are really aware
of this.

So it's not really a threat. But then it's more of an education. I think
education might go a long way.

Brooks Hamilton: Yes, our firm has a strong commitment to figuring out how
this can be a net positive societally. So we're not interested in working
with organizations that are just going to cut and cut. We believe that's not
the answer.

There are many ways of driving value and growth in the organization other
than that. And what AI opens up for us is the opportunity of matching our
intent, like what we want to do, with the end product. Where before, the
thing that stood in our way is a certain set of skills that we may feel
before just impossible to surmount.

Like beforehand, you and I were talking about different languages. You know,
how do I go about working with an international branch? Where I thought, oh,
I don't know, they speak French, or Canadian French, or Spanish, or
whatever. Why is that a limit anymore? It's not. I can just use whatever
language I would like because a language model provides for that.

Similarly, if I'm developing, if I have an idea for, you know what would be
fantastic here, is an app that helps me with my route planning, that takes
into account my workout schedule or whatever it might be. Developing those
applications today is actually amazingly simple. So before it took years of
study and years more of practice in order to develop these applications.

That's not the case any longer. So going back to kind of the idea of how do
I go about adding value to my organization, solving the problems that are
there and then how do I get creative about it and adopting the mindset that
there are a lot of barriers that don't apply anymore. That can be immensely
liberating.

And think about what's possible in an organization where people are not
having to spend time matching Excel cells. And instead are spending time
thinking about what an awesome flyer look like. Hey, which customer should I
go talk to? How can I go prepare for my negotiation and have a back and
forth in preparation for it, and then get critiqued afterwards all in the
safety of my car on the way home?

Only talking to AI? That's an incredible opportunity. So, we can massively
increase in terms of the skills and the output that we have as contributors
within an organization that just wasn't even thinkable.

Karthik Chidambaram: Yeah, we were even talking about price optimization.
And as a customer, I don't have to call the customer service rep to find out
what the price of the product is.

And then with AI, you know, I can figure out the price or depending on what
I need. And then as a customer service or an account manager. You can
upsell. Hey, this person is looking at the price multiple times. So there's
an upsell opportunity there. So time well spent. Yeah.

Brooks Hamilton: That's absolutely. Time well spent.

Karthik Chidambaram: So obviously, I mean, these large language models and
all that, the big companies are an advantage over the smaller ones. Is that
a right observation? Or let's say even a small tech player Can just get up
there or who's going to lead the game or what are your thoughts there? Oh,
man.

Brooks Hamilton: What a question That's a multi trillion dollar question
right there Let's take it back to distribution for a second and then talk
about tech. How's that sound?

Karthik Chidambaram: Sounds good. Yeah.

Brooks Hamilton: I firmly believe that the organizations that are going to
prosper in an AI technology area are not the organizations with the biggest
budgets and the most impressive IT teams The ones who are really going to
get a ton of value out of this, and I think grow in a very nimble way, are
the organizations that can make decisions quickly.

And the neat thing is, a lot of owner led, employee led and family-owned
organizations, they can do that if they are all of the same mind and say,
yeah, this is a big deal and this is going to impact us from an existential
basis, they can move on it, whereas oftentimes if you're coming from a
larger organization, the organization may think, yes, that's something that
we need to do, but it's not.

There are many pieces that need to be organized, aligned, budgets
restructured, organizations restructured, in order to make that a reality.
And so, they may be great in a longer time frame of being able to be highly
successful, but I think in the short term, the prize is going to go to those
who are fast and nimble rather than those who are big.

And I think we're going to see something similar in the technology space as
well. And let's just use one example. How much do you use a language model
for search now, as opposed to Google?

Karthik Chidambaram: I use it a lot right now.

Brooks Hamilton: And how about two years ago? Nothing, right? Nobody did. So
all of a sudden, we see these little upstarts from Perplexity, OpenAI,
Anthropic, and others Jump into the search market, which we all thought was
entirely owned by Google.

And, they are able to pull out 56 billion every year profit. So that's a
pretty good size opportunity right there that these other organizations
have. So I think it's just another way of seeing this race may go to the
nimble. Not necessarily the incumbents with the biggest IT and technicals.

Karthik Chidambaram: It's very encouraging, not just for the distributors,
but even for a tech company. Tech companies, right. I think it's awesome.
Let's say I want to up my skills, and I'm completely new to all these big
buzzwords, right? Maybe can you give me a book or two that I should be
reading?

Brooks Hamilton: Number one book that I recommend is titled Co Intelligence,
and that is by Horton Professor.

I think it's really one of the most applicable to business users and to
people who have really not touched AI before. Ethan Mollick does an
excellent job of keeping the language simple. Having very actionable use
cases and just kind of talking about what the implications are. What's neat
about that book also is while it's mainly focused on business, it's also
focused on education since she's an educator.

And I think for those of us with families, it's really interesting to think
about what are the skills that our children are going to need in addition to
what we need today as workers. So I think it's a great place to begin.

Karthik Chidambaram: Thank you for the recommendation. And the theme of this
podcast is driven.

Yeah. So how are you driven?

Brooks Hamilton: How am I driven? This really goes back to my love of
understanding intelligence. So let's back up just like a little bit and
think about what did we know about intelligence five years ago? We knew it
was something, maybe it had to do with language. Computers seem to be smart
in some ways, but terrible in others.

And since then, what we've learned about is that there are laws.
intelligence. We know that if we have models with certain characteristics
and data with certain characteristics and run these things for a certain
amount of time then we get a particular set of capabilities. We didn't know
that before. We had no idea.

So, just driving towards understanding and seeing how these can be Applied
more broadly have been a big personal motivator. What I think about in terms
of my work within distribution is there are many different types of external
competitors. Just like remember Amazon business coming into on the e
commerce side.

We have many new types of competitors who are going to be moving into
distribution. And those may be publicly held organizations that are going to
buy distributors and modify their processes and go about competing in the
market and kind of doing roll ups. But it may also be coming from the
venture capital.

groups who have formed partnerships with BlackRock and others for financing
to go buy industrial organizations and go do the same. Quite quickly, we are
going to find that there are entrants into distribution that are entirely
different from the types of companies who they were competing against
before.

The goal of our firm is to help equip those smaller distributors or even
larger distributors to be able to compete effectively against these new
entrants. And that's our mission.

Karthik Chidambaram: I would like to end with this question, Brooks. How do
you keep yourself updated or what book are you reading right now?

Brooks Hamilton: Yeah, great question.

In order to keep updated right now to a really constant stream of
subscriptions to newsletters and oddly enough, YouTube. Some of the best.
People who I think are really pushing the boundaries and doing a great job
in terms of getting facts out about how to use these applications and what's
coming out are coming from shows like Andrew Berman, Wes Roth, and multiple
others that are doing a great job of letting us know where the tech is and
how it's applied.

But the other thing that at least I keep my eye on a lot is I read a lot of
the research papers that come out of the frontier labs in the U.S., China,
and Europe as a way of trying to understand where the technology is going to
be, not just today, but where it's going to be in 6 to 24 months. And if
anybody tells you that they know where things are going beyond that time
frame, it's just pure speculation at this point.

Karthik Chidambaram: Brooks, thank you so much. I really enjoyed this
conversation. Thank you for being a part of the Driven podcast and great
chatting with you.

Brooks Hamilton: Likewise. This has been a lot of fun.