In this episode of the RevOps Champions podcast, host Brendon Dennewill is joined by data and AI expert Andrew James to tackle this common feeling of paralysis among today's business leaders. They dive deep into why simply adopting more technology isn't the answer and explore the real source of competitive advantage in the AI era.
Andrew argues that the key to leapfrogging the competition lies not in external tools, but within your own company's data. The solution is to plumb your unique operational data—from your CRM, sales pipeline, and financials—directly into AI to make decisions with high conviction. This approach transforms AI from a confusing threat into a powerful ally, revealing the true constraints and leverage points within your business.
This episode is essential for RevOps professionals, B2B executives, and any leader feeling stuck, providing a clear framework to move from analysis paralysis to profitable, data-driven action.
What You'll Learn
Resources Mentioned
|
|
Andrew James is the Founder and CEO of Cerebro Analytics, helping advertisers make smarter, more profitable decisions by turning all their data into a single, clear dashboard. With nearly 20 years in direct-response and digital marketing, he created the S.E.E. Framework (Strategy, Execution, Eyeballs) to help brands reach the right audience profitably.
A lifelong entrepreneur and storyteller, Andrew is also the author of Connecting the Dots: How to Tell Stories That Sell Your Business or Idea in 5 Minutes or Less. He helps businesses scale with clarity, confidence, and data they can trust.
|
Brendon Dennewill: Have you ever felt or thought to yourself that you have all the tools and opportunities, and for some reason you feel like you should be further ahead than where you are? Today, I'm going to dig into this with Andrew James, my guest on the podcast. Andrew, welcome to the show.
Andrew James: Thank you. Glad to be here.
Brendon Dennewill: You and I have been working on and testing various ways to use AI to do the work of previously human-only sales, marketing, and business processes. What is the biggest challenge you're working on finding a solution for?
Andrew James: The biggest challenge we're solving is getting internal business data plumbed into AI so you can make decisions off of the hard numbers that matter to your business. LLMs and the internet as a whole have an amazing amount of data and information, and we all have access to that for free or twenty dollars a month times however many subscriptions we have. But they don't know things like the number of leads you got yesterday, or which ad they came from. They don't have the internal data about a conversation that happened in your CRM. They don't know which leads have been handed off to customer success but haven't been followed up with on the cadence they ought to be.
I started a long time ago as an early internet kid, the kind of young kids who were making money from their parents' basement. Early eBay, Google when it was five cents a click, all that. You started to figure out what worked and what didn't, and you tried to do more of what worked and less of what didn't. That's what led us to today, where we started pulling lots of data into people's universe to make better decisions.
Marketers are an interesting mix because they're really willing to pull in the latest and greatest and quickly adapt. They're not as engineering-minded, so things don't have to be perfect. But we've become more minded that way. I keep seeing this opportunity for people to pull all the data in, make better decisions from it, and go act on it.
Brendon Dennewill: I actually responded to someone's post on LinkedIn yesterday about this. They were talking about how we hear all the time, in our case it's HubSpot: "We're only using 30% of what HubSpot can do." My response was that I used to think that way too. I've completely flipped my thinking. HubSpot keeps adding functionality, like every other software tool does. That's the business they're in. They want to keep giving customers access to more tools to do more things.
But they're not suggesting you have to use every single one of them. I'm suggesting you should not be trying to use all of them. What you should be doing is figuring out what are the specific business challenges or opportunities you have, then going back to the platform to see which specific tools can serve those needs. I believe you can use 30% of HubSpot and drive incredible value.
So with that said, we have access to all these tools and opportunities, and people have this feeling that they should be further ahead. How do you think about that?
Andrew James: It came to me from the Spotify CEO. We thought a lot about how all of this plays together as we build, roll out, and buy products with AI. For the first time, AI will allow, or actually force, more business users and owners to build tools internal to their organization. They're basically building products to solve their own problems. That's the same thing you do when you build a workflow automation inside of HubSpot, or link it together with N8n, Zapier, or Make.
So I think it's worth saying: we're all builders. I came back to this idea from one simple question about a product, and I think it's good for the audience. Any time somebody gets done using your product, or service, or having an interaction with you, what's the way you want them to feel?
Compare how you feel when you finish using something like Spotify versus something like Twitter, or a CRM. The core objective of anything we're building with AI is this: you should finish the session feeling ahead. The simplest version of that is time savings. I'm ahead of where I would have been had I done this manually. But expand your thinking a little, and that translates to ahead of my competition, ahead on the timeline, ahead in my career.
The only way I know to help people leapfrog ahead, with or without technology, is to elevate their thinking, ask a better question, set a better goal, and look at the core driver of what's going to get them there. What's the leverage point? What's the constraint you can alleviate?
Your HubSpot 30% example: if I can use 1% of HubSpot and it gets me closer to my goals five times faster, do I really care about the other 99%? No. So first, find and identify the right problem. What's the one metric that, if we changed it, would materially impact the speed at which we move in our business? Then you go into how to attack that problem, break it down into components, alleviate the constraint, and get there faster.
Andrew James: I believe that if you can make enough right decisions in the right order, you can have almost any outcome you want. That's a fundamental underpinning of what we've done. We try to translate everything into decisions. But if you have all the business data and the power of a gazillion super genius computers, plus an intelligent human directing it, you can ingest all the data and ask: where actually should we focus? Where is this business constrained?
If we make it practical in a RevOps or sales and marketing automation scenario, it might be looking at the various stages of a pipeline and figuring out where deals stall. Where are we spending more days than we should? Could we shrink the cashflow cycle from 41 days down to 30? What's the material impact of that?
Just by using the data to focus on the right problem, you're already ahead. Every time we do a deeper fundamental analysis, we almost always uncover something more exciting that could get you further ahead.
I had a consult with a therapist this week who was asking about scaling their practice. We looked at all the numbers and I said, "Why don't you just raise your prices?" It was that simple. He was about to embark on a 12-month journey to do something that had a 50/50 shot of success, plus a lot of cost and risk upfront. When you looked at all the data, you could intuit with math that he could do a little better with a much simpler move.
Here's another real example from the data. A new customer who does epoxy coatings, recoating garage floors. We were looking at their sales cycle: 41 days from lead to cash. But we found almost as much accounts receivable that wasn't on a collection cadence as their entire sales pipeline. We could go figure out how to increase top-line revenue, or we could just go collect the rest of the money faster. There's no cadence sitting there. Let's collect the money quicker.
There are a lot of ways to use data to first find the leverage point. Let's not talk about tools. Let's not talk about which decision to make. Let's find the core bottleneck to the business: the thing that, if this number changed, would move the business forward the fastest.
My framework for this is what I call X Factor Optimization. The key is finding it and boiling it down to a number with a dollar sign next to it. It's almost always cost per X or revenue per Y. If you can translate your business problem into a cost per X or revenue per Y that needs to move from here to there, it gets rid of not 70% but 99% of the things you need to think about and worry about.
Brendon Dennewill: One of my mentors shared with me five or six years ago: "Brendon, you do realize that 90% of leadership is making decisions." That really stuck with me because I'd never thought about it through that lens before. Now listening to you, it sounds like maybe more than 90% of leadership is making decisions.
Where a lot of leadership teams are currently getting stuck, because of the opportunity on one hand and the threat of not jumping on that opportunity that AI brings, is in trying to figure out which LLM to use or how to structure it for their business. They're not starting in the right place. They need to start from a strategic business perspective: what are our opportunities, what are the metrics we'll use to measure them, what are the dollar signs? Start there, get the leadership team in a room, figure out the opportunities and challenges, and then selectively bring in the data.
Some of what you and I have been working on recently are things we haven't been doing for the last three or four years, simply because it would take too long with humans. We have very smart people here at Denamico and we didn't see the value proposition, even though the value was there. For example, auditing a CRM. That would take 20 to 50 hours depending on complexity. Our clients didn't want to pay $5,000 to $15,000 for a CRM audit, even though they probably had hundreds of thousands or millions of dollars hidden in there. They just didn't see the value. And many still don't, because we're only just starting to see it ourselves.
But we all know it's just a matter of time before everyone realizes they can suddenly start finding incredibly valuable things. Things that will completely change what they do, how they do it, how quickly and inexpensively they do it, and what the impact is to their business in real dollar terms, whether it's revenue, profitability, or all the other things that make it very attractive to work for or with a company.
So with that said, let's talk about a mistake that you made that taught you the most about measuring real profitability.
Andrew James: The biggest mistake I've made is being the shoemaker with no shoes.
Brendon Dennewill: I have no idea what you're talking about.
Andrew James: Every CRM implementation expert I've ever talked to, when we asked to look at their CRM, they almost always tell me the same thing most agencies say: "Our ads and landing pages and sales funnels and processes are not as good as our clients', because we spend all of our time fulfilling for our clients." And the ones where that's not true? They're pretty bad at fulfilling for their clients.
So for me personally, the biggest mistake is when I deviate from this: measure what matters and align all decision-making around that.
It's not just for leadership to make decisions. That's the other huge mistake. We think about this from the frame of: people at the top will have all the data and therefore be able to make decisions, and everybody below them just executes. But you look at the biggest companies in the world, and it's literally the exact opposite of how they function. Google famously uses OKRs, objectives and key results. They don't even know what 80% of their people are working on directly half the time because those people are choosing some of their own objectives.
When you think about decision making, we silo it into this idea of somebody at the top making a decision. It's a lot more about pushing it down. We've got to build infrastructure such that even machines can make effective decisions within the right safeguards and rule set to move the business forward. At a high level, sure, there's a business constraint. But if we go down into marketing, sales, or customer success, there are going to be decisions for people in those departments to make. And that's something they can push against and improve upon.
Of course, there are times when there's data some people in the business shouldn't have. But I'm a pretty big believer that if we feed the right systems, tools, and people data that's both available to them and relevant to them, and they look at the scoreboard, they can optimize against it. Humans are great at this. It's one of the things we're best at. The wisdom of the crowd always beats the intelligence of the one player at the top. The market collectively always beats the individual.
So to me, the biggest mistake is deviating from picking the thing to focus on, not measuring it for ourselves and using it for ourselves, and thinking about decisions as a top-down-only exercise rather than giving every user visibility into the piece of the business they affect, with their own individual metrics they're trying to move.
Brendon Dennewill: What I'm seeing is so many leadership teams, whether it's a $50 million business or a $500 million business, are not making the decision to equip their people with the tools and the data to do the things you just described. They're stuck and afraid to make the changes to move into this AI era. They know they're going to be left behind, but they still don't know what the first step is.
We know the statistics: 100% of business leaders who are serious about growing believe AI will be an absolutely necessary part of their future. But only 5% of them are actually doing something about it. That's what I'm trying to figure out how to help those leadership teams overcome. Start simple: whether it's marketing, sales, or customer service, look at the obstacles currently blocking revenue, profitability, efficiency, or productivity, and just start building use cases for that.
With Cerebroplum, we can pull data from pretty much any source they use, whether it's their CRM or accounting software, into one place and help them make decisions. But the message I'm trying to get out is: how do we help more of that 95% get beyond where they're stuck?
Andrew James: There's a well-received book, Morgan Housel's Psychology of Money, and I think his fundamental answer to your question is: conviction. People want conviction more than they want accuracy.
Brendon Dennewill: Which comes back to what you were saying about decision making.
Andrew James: Exactly. When I was coaching team members on how to lead up the chain of command, the idea was: how can we give people above us decisions, not projects? Or better: the trade just hasn't been made clear enough to them.
Because what business owner, if you said, "If you put $10,000 a month here, the forecasted ROI is $100,000," and they have conviction that it will yield that result, wouldn't make the decision? It's not a difficult decision anymore. I think it's a matter of conviction, which is why I go back to the data. Often it's a trust factor, not just whether the data is accurate. It's getting to sufficient conviction to go further.
The other thing equally important: there's a volume of low-accuracy, low-conviction-required, obvious small decisions that just aren't getting made, and people can make them much faster and more of them with AI. The simplest example: a subject line for an email, even a follow-up email to one person. There is so much lift in subject lines. And how many times is an email being sent where someone doesn't take a second to write three of them?
When I go create a broadcast email to a list of 100,000 people, one of the core things we make people do before they write the emails is write 10 subject lines, like a marketer would. Those kinds of little decisions: they're not even presenting themselves with a trade to look at two options.
So to ladder back up to it: why are the 95% not moving forward? Because they're asking, "What should I do?" or "What should our AI strategy be?" And they're divorcing it from "What should be our strategy to exist, continue existing, grow, and be more important? Then where does AI fit there?" Data fits everywhere. I think of it as strategy first, then data, then: how can AI help us do those things better?
The reason they're not acting is the conviction isn't high enough. Usually what drives conviction is either pain or overwhelming data-driven evidence that you'd almost be stupid not to act on it. The question you have to ask yourself is: how do I get convinced enough before everybody else is feeling the pain and doing the thing anyway? How do I boil this down into trades? It's not a data problem. It's not an accuracy problem. It's a getting-to-a-clear-trade problem in the decision-maker's mind at every level.
Brendon Dennewill: Understanding the consequences of not making a decision matters enormously. I definitely was not a quick decision-maker for a very long time, and I learned it very late in my career. I'm still working on getting better at it. One of the ways I've been thinking about it recently is: just make a decision and then make it work. Don't wait for a way to make a decision you think will work. Make the decision and then make it work.
This has been one of the things I've learned over the last number of months as we've been leveraging AI more. For me, it's been a complete revelation of something I'd never thought about or thought was possible, taking this engineering approach to business in general. I don't have an engineering bone in my body, but this whole idea of: test something, see what happens, see what you learn. You're either winning or learning, and the only way you can learn is by taking action.
Thanks to AI, I can start testing things I never had even thought about before. I'm testing dozens of things every day. Through that process, I'm learning more. And I'm sure millions of other people are having the same experience. I'm trying to understand what is holding leadership teams back and how I can help them.
Andrew James: You can't steer a parked car. A lot of our advertising success was based on the fact that we'd test a thousand ads versus ten. We built frameworks to test in small batches early, get feedback, and iterate through those loops.
But I think there's a bigger answer here, and this may be self-serving on my part. One of the reasons these leaders are frozen is that the output they're getting is wrong or not quite there. They're promised a level of magic, they see it demonstrated, but then their first attempt doesn't yield the same kind of magic applicable to them. Something we do when we train customers: it's not about a magic prompt. It's actually all about the follow-up, going back and forth three, four, or five times. Just think out loud with it and you get somewhere much faster.
But the other side is part of why we took the thesis we did: if we could plumb their real data into AI, they would get to utility much faster, because there are things you can't ask the LLM that are the most relevant. What's going to change your business more: the number of leads you're getting and the cost per lead, or some external fact the LLM already knows about? If the LLM knows it for everybody, it's not an advantage for you anymore. But you are sitting on this gold mine of data.
AI has been here; we just used to call it machine learning. What happened is there was a user experience leap forward in terms of how easy and magical it was to get something back from ChatGPT. Then everybody experienced that output and said, "Wow, this is really cool." Some people went all in on prompt engineering. Some people used it to look up a cooking recipe when it was easier than Google. Then they hit a zone of disillusionment: "Well, that's cooler than the other thing, but it's not good enough."
There's almost always with new technology a long winter, a hype cycle where people get really excited, try a bunch of stuff, then find it's not as easy. A bunch of people fall off. There's a trough of, "It's not as magical as I thought. This is still going to take some work." Then the rest of the market says: "There's potential here. This is the new way of doing things." They keep working at it and get an outsized payoff.
The stick for those frozen leaders: other people are going to figure this out. You don't want to reach the pain of people beating you because they went through the trough after the hype faster than you. I say this even to the plumbers and trades people who are not directly disrupted by AI businesses. We're probably pretty far from robots that go fix toilets. But what's here now is: your competitor is using AI to generate leads, close sales, serve customers, and follow up effectively. Your competitor, also in a less-than-easy-to-disrupt business, can use AI to out-compete you. Do you want to get left behind?
The fastest way to not get overwhelmed is to pinpoint exactly: where do we need to get further ahead? How do we bring our data to the problem, bring AI against that, and attack it?
Brendon Dennewill: Let's come back to something you started talking about earlier: a single metric. How would you recommend leadership teams identify the single metric that unlocks profitable scaling for them?
Andrew James: Let me define it first, because that's really important to how the system I use works. I think about an X Factor Decision Metric. This is at the top level, but you could apply this to any point A-to-B that somebody's optimizing. I wrote an entire book about it called X Factor Optimization, which I've pulled off the market to retool it for AI.
The X Factor Decision Metric is some number with a dollar sign next to it that must be changed for us to increase the speed toward our goals. When you can put a dollar sign next to the metric, that's what clarifies thinking up and down the chain across the board.
Elon Musk is famous for doing this. He looked at what a rocket cost, looked at the raw materials, and said: "We need to make the cost per rocket this number." The materials cost something like 2% of the final cost of the rocket. He calls it the idiot index. So he was saying: "We should be able to get from a billion dollars down to 100 million per launch."
In a normal business dependent on a CRM, it's probably something like cost per lead, cost per acquisition, or dollars per X. Our real secret sauce within X Factor Optimization is there's some number a little bit after cost per lead or cost per acquisition but before lifetime value that is likely the core key driver of that business. In businesses with subscriptions, one example is: cost per new subscriber, backing out all the cost of acquisition and all the initial revenue in the first 30 days. How far ahead or behind are we? That's early enough to measure frequently, but deep enough into the customer journey for them to do something about it regularly to move the needle.
So first: X Factor Decision Metric, a number with a dollar sign, usually a cost per or revenue per.
Then what you act on every day is a Needle Move Metric: what percentage has the most outsized impact on that number? If your X Factor is cost per acquisition, your needle move metric might be conversion rate.
Just give people: what's your dollar sign? What's your percent sign? What's your number sign? Use the number sign to elevate your thinking: "We're getting 300 leads right now. How would we go to 3,000?" Well, you can't go to 3,000 because what dollar sign is holding us back? Is the cost per lead really too high?
Very often, if you ask enough of these questions, you'll find that they're just not measuring. That's the best thing I can find for somebody: "Did you know these numbers actually work and you could just turn up the ad budget? You could add five more people to the outbound team and go harder?" They just think the $500 per lead doesn't work. If that's really the case, then it's just a matter of breaking it down further.
The magic of all this is: plug the data into AI and ask these questions. Ask it, "Where am I constrained?" You can't ask a normal model that because it doesn't have your business data.
Brendon Dennewill: Which is what Cerebroplum does. If people just want to experiment to get more conviction, they can export data from any tool they use and feed it to the LLM. To get initial conviction and traction: pull any data set out of any tool you use, analyze it, and ask it, "My goal is this. Help me figure out, in dollar signs, a cost per or revenue per to measure our efficiency here, and tell me what I should focus on."
Brendon Dennewill: As you're talking about that, it makes sense and it would be helpful to a lot of people because we're having these conversations every single day with many different types of businesses. The obvious differences: some need 10,000 customers who pay them $10,000, or 50 customers who pay them a million dollars. Both extremes have come up in just the last week for me.
The company doing million-dollar deals currently has six salespeople. Without looking at the data, the obvious move is: add three or four more salespeople. But in looking at their data, we realized they could actually hit their 30% revenue growth goal without adding any salespeople. They just have to invest that money somewhere else.
We give them the data and say: you could actually do this without hiring any additional salespeople, or instead of hiring four new salespeople, hire one and invest a fraction of that in marketing, efficiency, data, and tools. Because everything else then drops to the bottom line. They were chasing revenue, but once we present the fact that they can have X revenue with Y profit, or lower revenue with higher profit, every business leadership team has to decide that for themselves because their business is not the same as everybody else's.
Andrew James: There's also a fair question of why not just do both. Why not add the six salespeople and make the process more efficient? That's how the people who are going to beat you are going to do it. They're going to do more volume and get it right through repetition. You could always ask: why not?
But the bigger conversation is also: how is this going to change? There are things you might have only done deeper into a customer engagement for money that now, with AI, can move to the front. That's a lot of what I'm thinking about for many businesses: what step in the process that's five or ten steps deep could actually be the beginning of the customer journey, which is a complete game changer as to how they interact with you and why.
LendingTree is a simple, dated example that seems trite now, but it disrupted an industry: "Let's underwrite before we apply. Let's collect the information and then feed it to a whole bunch of different banks." Almost everybody has something laborious they might be able to do with AI that could actually become a lead magnet or the start of their customer journey and completely change the game.
When I need to get down to brass tacks, I think about three buckets: cost to market, cost to sell, cost to fulfill. If you say, "Is there something in that fulfillment bucket we can do as our marketing on the front end?" it changes the numbers on all three. You stop thinking about percentage improvements and start thinking about reinvention. And I think that's the bigger thing we're facing: the risk of not reinventing.
Brendon Dennewill: Absolutely. And we've learned that ourselves. The changes and the possibilities are cross-functional in a way that's entirely different from how we used to think about it. We used to think of cross-functional as the horizontal: marketing, sales, service, ops, finance. What we've been working on for the last seven or eight years is how do we bring cross-functional customer teams into one system, tracking aligned data? That was a huge advancement, which is what we call revenue operations. One RevOps team aligning marketing ops, sales ops, and service ops together.
Now with AI, we're realizing we can uncover so much value already in the marketing, sales, and service process that makes things completely different and better for the delivery side of the business. Because we already have so much data before we even start doing the work, which is more efficient for everybody and eventually will bring prices down and value up for everyone.
Brendon Dennewill: Andrew, as we wrap up: is there one framework or piece of advice you can leave our audience with as we close out 2025 and get ready for what looks to be a very exciting 2026?
Andrew James: Find your profit prompt.
To take this back to AI: you're looking for something you can ask AI on a regular basis that yields a profit for you. If the AI has your data, however you get it in, and whichever model or models you choose, you need recurring data and a recurring action you can take as a result of that data that makes you more money or saves you cost. Examples are usually either on the buy side or the sell side: a decision to buy more at this price or not, who to follow up with, which ad to spend more on, which leads to prioritize.
And I give people this simple two-by-two matrix. To increase profit, we all agree you either have to increase revenue or decrease cost. That's one half of my two-by-two. Then I break it down into new and existing.
The simplest way to leave someone here is: where am I going to be more profitable? By increasing revenue on the new side, or increasing revenue on the existing side. By decreasing cost on the new side, or decreasing cost on the existing side. Just two-by-two. Up revenue, down cost, new, existing.
Pick a quadrant and load some data and ask some questions. You should be able to find a prompt you can run on a regular basis to figure out where the opportunity is. If people want to take the easy path, we will happily load all their data and give them several example prompts to help them find the clue and see if there's something there. Because every time we do an AI-enabled audit, especially on CRMs, we almost always find orders of magnitude more value than this could cost in leads that have stalled, been forgotten, or represent small revenue opportunities that add up fast.
Find your profit prompt. Look under new and existing. Look for increases in revenue and decreases in cost. Pick a quadrant and start throwing darts. You'll probably make more money in less time than you thought, and get further ahead faster.
Brendon Dennewill: Awesome. That's great advice. Andrew, thank you so much for joining me today. I always learn so much from your analytical way of thinking and your strong conviction about what works. One of the other things I've learned from you is that you're not really concerned about what doesn't work, because you just focus on trying things you think will work, keep learning from them, and eventually they do work. I believe a lot of our audience can learn from that. I know I have. Thanks again.
Andrew James: Thank you. Success comes from our winners.
Brendon Dennewill: That's right. Thanks again, Andrew.
Andrew James: Thanks.