From Data to Decisions: Building the Self-Multiplying Business | Peter Caputa

Peter Caputa, CEO of Databox and the architect behind HubSpot's legendary partner ecosystem, returns to unpack why having more data has not made decision-making any easier, and what it actually takes to close that gap.

Brendon and Pete dig into the four pillars of revenue operations, why franchise systems are uniquely positioned to win with AI, and how the most forward-thinking organizations are moving from dashboards to autonomous, self-correcting business systems.

If you're a RevOps leader, franchise executive, or operator trying to turn data into a competitive advantage, this episode reframes what's now possible.

Read the full transcript.

What You'll Learn

• Why more data doesn't automatically mean better decisions
• The four pillars of RevOps: people, process, data, technology
• What anomaly detection can do for multi-location brands
• How to daisy-chain processes into a scalable system
• The franchise visibility gap and how to close it
• Writing down processes before automating them
• Why AI makes this a bigger shift than the internet

Resources Mentioned

• Databox
• "The Cold Start Problem"  by Andrew Chen

Listen

About the Guest

Episode Transcript 

Introduction

Brendon Dennewill: Hello and welcome back. Today I'm excited to welcome back Pete Caputa to the podcast. Peter is the CEO of Databox, former vice president at HubSpot, and one of the most influential voices in partner-led growth, agency ecosystems, and revenue operations.

Pete helped build and scale HubSpot's agency partner programming into one of the most successful partner ecosystems in SaaS, and today leads Databox, an analytics platform that helps businesses turn performance data into better decisions, which is one of the reasons I wanted to have him back. Throughout his career, he has been passionate about helping agencies, consultants, and services businesses create measurable value for their clients through data partnerships and operational excellence.

What makes Pete's perspective especially valuable is his ability to connect growth strategy, partner ecosystems, content marketing, and business intelligence into a cohesive approach to scaling modern business. Whether discussing agency growth, SaaS leadership, or revenue operations, he consistently emphasizes one principle: help others succeed first, and then growth will follow.

Pete, welcome back to the RevOps Champions Podcast.

Peter Caputa: Got me in my feels, Brendon Dennewill. That was probably the nicest intro I've ever had. Thank you.

Brendon Dennewill: Well, I'm just stating the facts. Pete, thanks for being here.

Peter Caputa's Career and Growth Philosophy

Brendon Dennewill: You've had a unique journey, from engineer to entrepreneur to a HubSpot leader and now CEO. Looking back, what experiences most shaped how you think about growth and leadership today?

Peter Caputa: I think you kind of covered it there. Focusing on customer value is the one area I've obsessed over in my career: building ecosystems and then focusing on the value to the people or companies in that ecosystem. That's what motivates me, because I can see a direct impact on the partners of the companies I work at, whether it's HubSpot or Databox.

I even had a startup beforehand where we built a partner program. That startup is long gone, but it's the same principle: build a system that helps other businesses succeed, and when they do, your company succeeds. I love that, because I can build systems that other people use to become successful. That's the one theme I haven't gotten over.

Brendon Dennewill: Yeah, and what an exciting time to be doing that right now.

Peter Caputa: It is. I feel like with AI especially, there's so much potential for individuals to build cool things. I don't know if I've told you this story, but my son had a class in high school. He just graduated, but as a senior he took this entrepreneurial business management class where they actually got to build a business.

As an eighteen-year-old with no coding experience, he and a few buddies built a two-sided marketplace where restaurants could list deals and students could take advantage of them, then show their phone at the cash register to claim it. That's the kind of thing that a few years ago would have needed a handful of engineers, not just one.

So I think the world is changing so fast, and Databox has moved from being a product to more of a platform company, where people are building really cool things on top of us. It's a great alignment for my career: everything I've always worked on is coming together with the technology and the opportunity for people to really innovate on top of the product we've built here at Databox.

Why AI Is Changing What's Possible for Businesses

Brendon Dennewill: Very cool. That's part of why you and I have been talking again more recently. Pete, for the folks who don't know what Databox does, it's been two years since I had you on the show, and there's so much more you can do today. You and I, and thousands of other people, have been dreaming about things that are now possible that weren't before, and then there are things we didn't even dream of that are now possible.

Peter Caputa: Right.

Brendon Dennewill: So maybe just explain to us what Databox is and what the value proposition is.

What Databox Does Today: From Dashboards to AI-Driven Analysis

Peter Caputa: Since we talked two years ago, I'll give you a sense of what's happened. Our vision hasn't changed: it's still to help businesses make every decision, or as many decisions as possible, a data-informed one, across the world. What's changed is the technology we've built and can leverage in our product.

It's really three things, all in pursuit of enabling regular people to make data-informed decisions. The first is that we've built a really sophisticated data layer that pulls in data from any system a company uses. We've made it easy for hundreds of popular software tools: log into your HubSpot, your Google Analytics, your Facebook Ads account, and we pull the data in. We pull it in a way that's defined, so it's easy to build a dashboard or start analyzing the data using AI in our product. You can also push any data into Databox.

Step two is what most companies do to analyze data: build metrics, build dashboards, build visualizations, maybe add context to a report and send it out to people. That's a core part of our product and always has been. The newest thing, which has really changed in the last year plus, is the AI layer on top. Once you aggregate your data and define your metrics, even build some dashboards, AI is in a really good spot to not just tell you what happened, but help you understand why it happened, whether it was good or bad, help you predict what might happen next, and, most importantly, tell you what you can do about it to change future performance based on historical patterns.

If you structure, aggregate, and define your data well, AI can actually do the work that most companies rely on their smartest, most mathematically inclined person to do, and now AI can do it in minutes. That's what we've focused on for the last year: the value of our product has changed from just visualizing and reporting to actually doing in-depth analysis and helping you understand and decide what to do next based on your data.

The Franchise Data Gap: Why Real-Time Visibility Matters

Brendon Dennewill: One of the things we talk about a lot in the franchise space, because unfortunately it still happens too often, is that something happened six weeks ago at the franchisee level, but the franchisor at the corporate office only found out about it today. As you were explaining, in that scenario they don't have all their data aggregated, so they don't even have access to all the data across different systems, never mind having it in real time. As far as I'm concerned, there's no reason for any business today not to have data available in real time.

Peter Caputa: Right, pretty much real time.

Brendon Dennewill: I mean, within a day.

Peter Caputa: Exactly, there's no excuse not to have it within a day. I'd even say within an hour. Most businesses don't need to refresh most of their data every fifteen minutes, but refreshing every hour makes sure you didn't miss anything.

Brendon Dennewill: I think you touched on some of the realities, and most products are built to solve real business issues, which is what Databox is doing. So if a company, no matter what industry, has multiple sources of data, which means multiple technologies, because that's just how they're set up, and most businesses are, there has to be a way to aggregate and pull all that data in, because you often need to do that to make decisions.

Aggregating Data Across Franchise Systems

Peter Caputa: I think that's step one. In franchises, there are franchise groups that might operate a handful of locations and be sophisticated enough to have their own systems in place that the franchisor doesn't even have access to. Then there are franchisors with hundreds or even thousands of franchises, and being able to access that, let alone monitor it, is usually a challenge.

We have a handful of franchise systems that use us to aggregate data from different CRMs, different ad platforms and accounts, other marketing tools, financial tools, and so on. Then there's a matter of managing access: who has what. Of course you're not going to let Franchise X access Franchise Y's data, but the franchisor would like access to all of the franchise data, and the franchisees would want to see how they compare to other franchises, which you can do in an anonymized way. So there are different ways to set that up so the right people have access and there's transparency, plus cross-franchise insights so everybody can make smarter decisions.

That's usually the goal with a franchise: when one, or a group of them, is more successful, you want to emulate that across the other franchises.

Brendon Dennewill: Absolutely. And maybe there are franchise leadership teams that just aren't aware this is possible. That's essentially how we set it up in HubSpot, where corporate uses HubSpot for their franchise development, but they also make the same CRM available to all their franchisees.

Peter Caputa: Yeah.

Brendon Dennewill: And it's all permission-based, so franchisee one and franchisee one hundred fifty, if they don't have the same owner, won't have access to the same marketing materials, and they know what data they're tracking, and it all goes back up to corporate so corporate has the visibility to make decisions. But then we'd potentially bring in a business intelligence or data aggregation tool, something that does it all in one place, like Databox, where we can make sure they're able to analyze that data, which is a massive opportunity. It avoids that scenario, because every franchisor, and anyone who's ever worked in a corporate role at a franchise, understands you want to solve for an unhappy franchisee as quickly as possible. Six weeks is obviously not acceptable, and in many cases, the longer it takes to solve their issues, the worse it gets.

Anomaly Detection: Knowing What Changed and Why

Peter Caputa: Agreed. I think, as you know, we have a very strong integration with HubSpot, and we have franchises using us to pull in other marketing data, ad data, financial data, and so on, so they get that global view. But the work you guys do getting both franchise development and the franchisees using the same system, so the franchisor has visibility and can spot when something isn't going well, is huge.

The other thing our customers do is set up anomaly detection. Once data is aggregated in Databox, they define what metrics mean: literally saying this metric means this, or franchise pass-through revenue means this. They define what it means, and the system also captures how to calculate it, so when you say, show me franchise pass-through revenue, it does it the same way every time.

Once those metrics are defined, our system can detect any anomaly, positive or negative, and alert people to it. If you have hundreds of franchises using different systems, there's no way one human can pay attention to all of that. You'd need somebody to sit there and go through dashboard after dashboard every day just to catch something before it got worse. Instead, you can rely on AI and statistical techniques to say, this number is out of the ordinary, alert the team, and address it.

You can even go so far as, if you define your metrics well enough, alerting the team with the reason a metric is off: you've had fewer sales calls this week with potential franchise buyers, and here's why, this ad campaign stopped working last week for whatever reason. By having all those metrics aggregated, you can go a level deeper and understand why.

Brendon Dennewill: Which is that whole cause-and-effect thing that I know is part of how you think, and I'm guessing you even do it in your sleep.

Peter Caputa: Yeah. So anomalies are one statistical function, detecting something out of the ordinary in a time series. But there are also correlations. Correlations look at two different metrics and say they have an eighty percent correlation, meaning that when one changes, the other probably changes in a similar pattern. There will also be metrics that aren't correlated, in which case you'd never say this went down and therefore this went up. By doing that statistical technique, you can say this is most likely why something occurred.

Systems Over People: The Golden Age of Business Systems

Brendon Dennewill: Pete, let's go back to what else has really changed for you as a business at Databox, and the parallels between what you're seeing and the opportunity that exists now for businesses, whether they're in the franchise or multi-location space or not. What else are you excited about that leadership teams can do today that they couldn't do a year or two ago?

Peter Caputa: As you know, I'm an engineer, so I think in terms of systems. I think we're in a golden age, I'll just use that term, of systems running businesses. I don't want to downplay the importance of people; people are obviously very important to running a business. But I almost think systems are even more critical, at least than individual people. Systems can survive.

When you look at some of the biggest companies that have grown over the last decade or two, take Uber, or even HubSpot, though it's not Uber-level, or Google and Facebook, but even newer ones like Claude, they don't run on people as much as they run on systems. I don't know if you've read the book by Andrew Chen on network effects, where he breaks down how businesses like Airbnb and Uber grew. The people running those businesses are almost like they're sitting in a room, Star Trek style, with data everywhere, and that's what they're monitoring and trying to influence through their sales and marketing activities.

That was about ten years ago, and it wasn't feasible for most businesses to operate at that level. Now, with AI, it's very feasible, even for the smallest businesses, to look at data, run campaigns, run experiments, see what works, and build that knowledge into a system so the system learns and improves over time. Every business, franchise or not, should be thinking about how to build systems. Franchises are systems builders, but most businesses are still missing the sales, marketing, and to some degree customer service systems they could be running to capture data and optimize performance.

That's the opportunity in front of everybody: nearly automating everything. Of course, a lot of work goes into building and designing the system. You have to capture context about your business and your customers, figure out how to keep that context fresh, and set up systems for agents or people to do things repetitively so they can be measured and improved. But setting up those systems is really the opportunity.

People, Process, Data, Technology: The Four Pillars

Brendon Dennewill: You touched on multiple things there, but I want to come back to one. In the franchise space, private equity is growing at an incredible rate, and what you just talked about is essentially the reason private equity keeps investing heavily in franchises, because private equity invests in two things: people and leadership, and systems.

I think they see the gap in the franchise ecosystem, that there's so much more that can be done with technology and data systems, that there are massive opportunities, and they know how to fix that. They also know how to train and bring in the right leadership to run those systems in a very data-driven way.

That touches on something else we talk about on the show all the time: the four pillars of the revenue operations model we implement for a franchise brand, in this order: people, process, data, and technology. We see this all the time. When we're speaking to a prospective franchise client for the first time, they're often under the impression that if they implement a really good revenue operations system around the HubSpot CRM, it's going to fix their data, process, and people issues.

Peter Caputa: Not at all.

Brendon Dennewill: And of course that's not how it works. Typically we have to go back and ask, do you have the right people doing what they need to do, running your processes? Are you clear about what your processes are? And do you know what your KPIs, metrics, or OKRs are? Because we need to map that manually, and then we can build the system for you.

Peter Caputa: Right, totally.

Brendon Dennewill: And when we talk about that, a lot of models lump data and technology together as one thing, and we purposely separate them. I'm really happy we've kept those separate, because the data layer is so distinct now from the technology piece, which brings us back to what you do.

Peter Caputa: No, absolutely. I love your step-wise approach: people, process, data, then technology. Data is really a byproduct of running a process effectively, and similarly of repetition; otherwise you don't have an understanding of what happened and why, or you can't measure it at least.

If you implement a piece of technology that forces people to change their processes, that may or may not work. I do think there are some best practices out there on how to measure things and leverage technology, which should factor into your processes. But starting with your existing people and existing processes, and building things that are repetitive, measurable, and efficient, because the technology should ultimately make them more efficient or more effective, I think that's the right approach.

Franchise Development vs. Franchisee Success

Brendon Dennewill: One of the most common friction points in the franchise space is the friction between the success of the franchise development team and the success of the franchisees. You talked about this earlier: the original model is marketing, sales, and customer service. In the franchise space, customer service splits into two, because the way we think about it is franchisee service: how are you supporting your existing franchisees? This is transferable to any business, because the businesses that do really well are the ones that take care of their existing customers.

Peter Caputa: Yeah. You shared a stat with me a few months ago on the number of franchise systems that fail, and I was shocked at how high it was. I think any business thrives or fails based on the success of its customers, so putting that first is key. I know enough about franchising to know that often franchisors are so focused on acquiring new franchisees that they spend all their sales and marketing budget on that.

Meanwhile, if the existing franchisees aren't successful enough, it's going to be hard, because nobody spends a lot of money on a franchise without talking to a few existing ones first. And vice versa: when the franchises are doing well, it's a lot easier to market and sell to a new franchisee, because word of mouth will be there, and the market, the buyers, the consumers, will be talking about the franchise and its products. So the work you're doing, using HubSpot to enable franchisees to do sales and marketing consistently to their end customer, whether B2B or B2C, is huge for helping the franchisor and the franchisee succeed.

Giving Teams Data Visibility to Act Autonomously

Brendon Dennewill: Absolutely. The simple analogy I use all the time, and I learned this in my early twenties when I went to Europe and started working odd jobs, one of them as a longshoreman in the Port of Hamburg in Germany: coming from Africa, we didn't always have the best tools available to do our work. But in Germany you have the best tools to do your work, no matter what work you're doing, even as a longshoreman.

That's stuck with me for all these years. One of the things I think about all the time, especially at this time of year in the US, is family barbecues, whether Friday, Saturday, or Sunday. You get together with friends and family, and everyone's catching up on how things are at work. One of the things I'm always solving for is that I want the people working for our clients' businesses to be the ones saying, I love my job because I have the tools to do my work. Having the tools is one thing, but if you don't have clarity on how your success is measured in your specific role, no matter what that role is, it's really hard.

Peter Caputa: Yeah, totally, absolutely.

Brendon Dennewill: So providing that visibility, and providing the tools, which in our case is the technology they use every day to do their jobs, makes for much better, happier employees, who then tell their friends and family, I love my job because I have the best tools. You should come work for our company. Otherwise it's cousin Joe saying, I'm not really loving my work right now because I don't have those tools, I'm still using stuff built for the nineteen hundreds.

Peter Caputa: Right, and I think that's another benefit of aggregating performance data: it's about enabling teams to act more autonomously. Getting them the tools is key so they can do their work, but another thing to think about is giving individuals within your company visibility into how their work impacts other teams, or how other teams impact their work, or just what the company is focused on and trying to accomplish. Then they're unleashed to make smarter decisions that are aligned with the company's vision, decisions that might help somebody else in the company do their job better, because they see the impact on another part of the company or on the customer directly.

We've talked a lot about helping the franchisor get visibility into what the franchisees are doing, but the other benefit is giving everybody at the franchise access, respecting the privacy of other franchises through benchmarks or correlations, so they can make more informed decisions. Step one is always aggregating the data, but what you do with it, and who you make it visible to, is another big question, because that can impact productivity, effectiveness, and even what people decide to work on, because they're more informed.

Franchisee Education and the Weakest-Link Problem

Brendon Dennewill: Right, which is actually another growing space within the franchise ecosystem: franchisee education. You talked about the scary statistic, and it's not that different from starting a non-franchise business, but so many companies that decide to franchise never make it past ten units before they get out of it completely or fold. Even the ones that are successful, once they've got over a hundred units, other things break at that stage. Whatever got you to thirty, fifty, or one hundred units isn't going to get you from one hundred to three hundred.

Peter Caputa: Yeah, right.

Brendon Dennewill: One thing that's typically forgotten is that as leadership teams elevate and bring in additional leaders to manage franchisee relationships, a first-time franchisee has never run a business before. They were doing whatever their job was, they saved up money, in many cases their life savings, because they wanted to take control of their lives, but they've never had to deal with P&Ls or running a profitable business.

If you, as a franchisor, can provide the metrics for success, the things that lead to profitability, the things you have to measure that are going to be costs every month no matter what, that's not that difficult to provide. Now you have consultants who've been in franchising a long time going back and educating these folks, because it comes back to what we talked about before: if the existing franchisees aren't successful, the whole brand won't be, because a brand is only as strong as its weakest franchisees.

Why Having More Data Hasn't Made Decisions Easier

Brendon Dennewill: Pete, let's get back to something. A lot of people get scared when we start talking about data, and a very typical thing is that everybody listening to this is collecting more data than ever before, yet decision-making doesn't seem to be getting any easier. Why is the gap between having data and actually using it to make better decisions still so big?

Peter Caputa: If we could solve that one in ten minutes, I could retire. There are a lot of issues. Number one, numeracy is pretty low among the general population, even in white-collar work. People can maybe use spreadsheets at a basic level, but doing actual analysis of data is usually pretty rough. They don't know how to structure data correctly. You'd be amazed how often people connect a spreadsheet to Databox and some part of it has rows, and some part has columns over here, and they just don't know the basics of how to work with numbers. That's one issue.

The second thing is people don't think about data until they need it. It's like, we're running this process, how's it working, when in reality it needs to happen up front. You need to think through process first, then people, then process, but you have to say, all right, we're doing this process, this is the data we're going to measure, and this is the system we're going to use to measure it. Because they don't think about it as a system, whenever they go to do analysis, it's usually ad hoc, and it's hard.

Combine those two things, not everybody knows how to work with numbers, and the system isn't defined up front, and it gets hard. Usually what happens is companies are still operating at the pace that was possible before: we're going to do an annual plan, a quarterly review of the numbers, a report every month, though we're not going to change much, we're just going to report it out. Maybe we'll fire someone, maybe we'll go tell someone to work harder, but for the most part we're not changing the strategy.

Now, if you set all that up right, data analysis can be done in minutes, and in depth. I have skills built in Claude that work with the Databox MCP, which aggregates data from the fifty different tools we use. I can run a report and see how our business is running across everything. Not only that, I have every one of our sales calls, every one of our podcasts, every one of the surveys we run, vectorized into a database. So I can efficiently search that and say, what's going on with this feature adoption, and why? Because not only did I see that feature adoption is down, I can go figure out why. Did customers stop asking about it? Did they not like it? Did we break it? Did we roll out something that didn't make sense? What is it? And I can analyze that without talking to any human in my business.

Building Systems First, Then Automating with AI

Peter Caputa: But to get there, people have to really build systems. Without building systems, you're not going to get the benefit of AI doing analysis, of data moving around quickly and cheaply, of defining your metrics. All of that is possible, but unless you think about it from a systems perspective, you're not going to get there.

You have to step away from the business and ask, how are we operating, and what's the next step we can take to be more efficient, more effective, to produce better outcomes? For example, I work really closely with our marketing team. Each week we're producing marketing content, running campaigns, recruiting affiliates, co-marketing with partners, doing all of that. But we also take time to ask what the repetitive things we're doing are, and how we create a system that lets us do that much more efficiently, a system so somebody else can handle a piece of it, so I'm freed up to do something else.

So we're dedicating some of the marketing team's time to say, instead of creating the tenth YouTube short, write a skill that takes our latest podcast, identifies the ten shorts it could become, and writes the script and edits them. Or take all the written material about our next product launch and come up with a video script for our launch webinar. All of that can be done so much more efficiently with AI, but unless you step back and ask what you can create skills for, or automate, you're never going to make progress. It needs to start there, not with the process, not with the tool.

Brendon Dennewill: Yes, not with the tool.

Peter Caputa: Exactly. Well, I think there's a management layer that comes with this too. We actually ran into a problem where we now have Claude skills we've written that contradict each other, because they do similar things. So somebody needs to keep an eye on the whole system. The team that built the system and wrote the skills knows what's happening and can detect it, but they have to take the time to revise and review what they did, to make sure we're not giving the system contradictory instructions.

Brendon Dennewill: So you have to review your agents and skills the same way you'd review the work of a human.

The Future: Visionaries, Operators, and Self-Correcting Businesses

Brendon Dennewill: Pete, as we start to wrap up, looking ahead, how do you think AI will change the way leaders interact with data, make decisions, and run their businesses over the next couple of years?

Peter Caputa: I think the potential is to achieve what most executives and business owners have always wanted: to set the vision and let it run. I don't know if you use the Entrepreneurial Operating System, the EOS model, which talks about setting your ten-year vision and figuring out your annual and quarterly rocks, or initiatives, to prioritize. There's a role in EOS for the visionary: all they do is set the vision, and then the operator executes. The visionary focuses on talking to customers and figuring out where the market's going, guiding the ship.

In reality, what I find is that most companies that implement EOS, or systems like it, never quite get there. They either quit, or the owner says, I'm going to be the visionary, and you go run the business, stepping away in a way that doesn't quite work. But that's the potential I think AI offers: building a system where both people in the company and agents can act autonomously and innovate autonomously toward the goals the executive team sets. It's not so much about how you build a system that lets the business owner or CEO make decisions, but how you build a system that enables as many people as possible in your org to make smart decisions, while being accountable to them, recording them, understanding them, so AI can interpret whether those decisions were any good.

That's the future: getting to the point where entrepreneurs can be entrepreneurs, operators can be operators, and the system works and course-corrects itself to some degree.

Brendon Dennewill: That sounds very appealing, because most business owners, entrepreneurs, and executives love the idea of not just a self-managing business, but a self-multiplying business.

Peter Caputa: Totally, yes.

Brendon Dennewill: Which comes back to the four pillars of revenue operations, which, in your case and many other businesses that have gone the same route, makes a lot of sense. You no longer have separate revenue operations, business operations, people operations, and finance operations; it's all just operations, because technology and AI have made everything completely cross-functional. We see that even in our work. We're not just working with a CTO, a CMO, or a sales leader; ultimately it's the COO, the operator, making the cross-functional decision, because it's a cross-functional benefit and a cross-functional solution.

Getting Started: Document and Revisit Your Processes

Brendon Dennewill: So what do you think is a good first step to get to that self-multiplying company?

Peter Caputa: What I constantly do, and have done since early days at HubSpot, I learned this from Mark Roberge, my first boss at HubSpot as VP of Sales and later CRO, is force your team to write down their processes. That does two things: one, you get the process written down, which is step one to building a system, to being consistent and able to measure it. Two, you empower the team to build and own the process, because the people doing the work should ultimately own it. They'll see the opportunity to improve it, or at least some of them will.

That's the first step, exactly what I do with my teams: hey, you guys did something cool over there, where's the process for that? We haven't written it down yet, okay, you've done it a few times now, it's time to sit down and write a first draft. We don't have to stick with it, but let's think about how we build a process for that. Before, when I pushed that, it was largely so we could measure the outcome and the effort required. Now it's because we can automate big portions of that process.

When you start thinking about doing all these processes in isolation, what reveals itself is a system that can be optimized. For example, somebody in marketing is responsible for all of our social publishing, and somebody else handles our product marketing launches, which usually involve building a landing page, maybe a webinar, an email to customers, and so on. Those two processes were completely separate. Once we wrote them down, we could daisy-chain them: we're doing a product launch, let's do our normal email, landing page, blog post, but then let's tack on all the social posts, and now that we have all that context gathered and copy written, because we have skills that tailor content to the channel and format, it's done quicker.

That's one simple example. We also do partner co-marketing. Imagine taking all that context and attaching partners at the end: hey, you have first access to this new feature, we want you to be part of the launch, we're going to promote the use case you've been pushing to the market, and we'd like you to promote the launch too. So now partner co-marketing is attached to the end of that process, and that's just marketing. Add sales enablement, add customer communications, and you build a really efficient system. But it starts with asking everybody to sit down and ask, what do you do on a repetitive basis, is that documented, and do you have a system to measure it?

Brendon Dennewill: And I think the only thing I'd add is, how often do you go back and revise that process? Because one of the things we see, in the franchise space and many other industries we've worked in, is that mindset of what got us here will get us to where we're going. Going back to your question, have you ever written down that process? Yeah, we wrote it down five years ago. So when you had thirty units five years ago, and now you have one hundred thirty, don't you think that process might have changed? If you don't review the process because the process has changed, I can promise you, unless you haven't grown, the process probably hasn't changed much either.

Peter Caputa: Yeah, I think as people do things at more volume, they realize they can make it more efficient, or they get feedback that some percentage of the time something's off and they need to work that into the process. Processes should get more robust and be revisited. I think the beauty of AI is that you're forced to write down the processes, because otherwise you can't instruct AI on what to do or how to do it.

But we actually ran into a problem with that: we now have skills we've written, Claude skills, that contradict each other because they do similar things. So there's a management layer on top of it; somebody needs to keep an eye on the whole system. The team that built the system and wrote the skills knows what's happening and can detect it, but they have to take the time to revise and review, to make sure we're not giving the system contradictory instructions.

Closing Advice: Don't Sit on the Sidelines

Brendon Dennewill: Pete, thanks so much for joining me. Any last word of advice?

Peter Caputa: I think there are too many people sitting on the sidelines right now. I went to school for chemical engineering, graduated in ninety-eight, and quickly realized the internet was going to be the biggest thing in my lifetime, the majority of my career. So I went back to learn how to write code, started a web startup, and got myself a job doing that stuff. I was, whatever, twenty-one, so I was very flexible and didn't have any responsibilities. I'm not saying everybody has to make that big a pivot in their career.

However, I'd challenge everybody to go out there and learn something new every week they can do with technology, with AI. Of course I'm biased here, but I'd encourage everybody to think about how they can automate data analysis, do more in-depth data analysis than they're used to, by leveraging AI. There's a massive opportunity to make more data-informed decisions if you structure your data correctly, aggregate it correctly, define it well, and use a proper system to do actual data analysis with AI. I'd challenge everybody to try it. It's a bigger change than the internet, from my perspective. I lived through that change and saw how much it changed society. I believe this is an even bigger change.

Brendon Dennewill: Absolutely.

Peter Caputa: So don't sit on the sidelines, get aggressive with learning this stuff. Hands up.

Brendon Dennewill: Good advice. Pete, thanks again so much. Take care.

Peter Caputa: Yeah, thank you.

Subscribe