Episode 340 | The Profit Answer Man | Jeff MacPherson
You just spent real money on an AI tool because everyone told you it would save you time, and a month later your team is more confused than before you started. Meetings take longer, not shorter. Nobody’s sure which version of the process is the “real” one anymore. If that’s where you’re sitting right now, you are not bad at technology. You just ran into something almost nobody warns you about before you buy.
The Problem
AI does not fix a broken business. It accelerates whatever is already happening inside it, good or bad. That’s the core claim Jeff MacPherson, co-founder of Cloud37, made on a recent episode of Profit Answer Man, and it lines up exactly with what Rocky sees across founder-led businesses doing $2M to $20M in revenue: the technology was never the missing piece. The process underneath it was. Jeff put it plainly, when you plug new technology into a business that’s already running on chaos, all it does is amplify the chaos. It doesn’t matter how big the business is either. Rocky has watched this play out the same way in nine and ten figure companies as he has in $3M shops. A bigger balance sheet doesn’t buy you a cleaner process. It just gives the mess more places to hide, for longer, until it can’t hide anymore and it shows up as a client complaint, a missed deadline, or a number that doesn’t reconcile at month end. By the time an owner notices, the AI tool has usually already been blamed and quietly shelved, when the actual problem was never sitting inside the software at all.
The Wrong Path
Of course you bought the AI tool. Every advisor, every LinkedIn post, every competitor’s case study said it was the answer, and the fear of falling behind is real. Buying the tool wasn’t the mistake. Expecting the tool to organize a process that was never organized to begin with, that’s where things went sideways, and it’s a completely understandable place to end up. Nobody hands you a warning label that says: this tool will do exactly what your current process does, just faster, mess included. You found that out the hard way, which is the same way almost everyone finds it out, usually after the invoice for the software has already cleared and a few frustrating weeks have gone by with your team quietly blaming each other instead of the process nobody ever wrote down.
The Insights
Standardize the Input Before You Trust the Output
Jeff’s argument is simple and easy to skip past the first time you hear it. AI is only predicting the next likely output based on the information it’s given. If the information going in isn’t standardized, and the process it’s supposed to follow isn’t standardized, the tool has nothing consistent to predict from. That’s where hallucination and inconsistency actually come from, not some flaw baked into the model. In practice, that means before you automate a sales call, an onboarding step, or a client deliverable, you need to define exactly what information goes in, what the process does with it at each stage, and what the finished output is supposed to look like every single time, not just the one time it worked well. Skip that step and you don’t get a faster process. You get a faster version of the mess you already had, delivered with more confidence than it deserves. The version of this most owners actually try first is the reverse order: feed the tool a vague instruction, get back a rambling fifty page document, and conclude the tool is the problem. Start from the finished output instead. Show the system exactly what “done right” looks like before you ever ask it to produce anything, the same way you’d train a new hire by handing them a real example before turning them loose.
Raise the Floor, Not Just the Ceiling
Every business has A players, the people who move fast and don’t have to think about the fundamentals anymore because the knowledge lives in their gut. The instinct is to hand AI to those people first, to make your fastest person even faster. Jeff’s reframe is the opposite bet: the bigger win is training a system around what your A players already know, then handing that system to your B players so they can execute at an A level too. He points to Henry Ford as the model. Ford didn’t make his best mechanics faster. He broke the process into steps simple enough that workers who weren’t specialists could still build a car at a specialist’s quality. That’s the shift sitting in front of most owners right now, and most of them are aiming AI at the wrong end of the team, the end that was already fine.
The 95% Infrastructure Gap Is Real, and It Has Nothing to Do With Technology
Jeff estimates that roughly 95% of businesses have no real infrastructure for AI to plug into. Not because they haven’t bought the tools. Because the underlying processes, the ones that were fine when one experienced person was doing the thinking in real time, were never written down clearly enough for a system to follow without that person standing over its shoulder. If your business runs on tribal knowledge and good instincts held by two or three key people, that’s not a weakness AI is going to fix for you. It’s the exact thing that has to get fixed first, in plain language, on paper, before any automation touches it. Otherwise AI just inherits the same ambiguity your team has been quietly working around for years, except now it makes decisions about that ambiguity at a speed nobody can catch in time. The businesses Jeff has seen succeed with this are the ones willing to slow down for a few weeks and write the thing down properly, even when it feels like a step backward while competitors appear to be racing ahead.
What Actually Keeps a Client Once AI Can Do the Deliverable
Here’s the part that matters most for anyone worried AI is coming for their business model. Jeff describes it as an iceberg. The tip, the part everyone sees, is the deliverable itself, and AI is closing the gap on deliverables fast. But the bottom two thirds of the iceberg is what actually keeps a client: accountability, a real relationship, a sense of community, access to a network AI cannot replicate. A client who is only paying for the deliverable will leave the moment someone cheaper, or some tool, can produce the same thing. A client who is paying for being known, held accountable, and connected to something bigger than the transaction stays, because that’s the part no model can fake, no matter how good it gets at the tip of the iceberg. If your business’s whole pitch is the deliverable, this is the moment to figure out what your version of that bottom two thirds actually is, before a competitor with a faster AI setup finds out first and takes the client who was only ever loyal to the output.
Rocky’s Perspective
Here’s what I see when an owner tells me they’re “doing AI” now. Usually what happened is they bought a subscription and pointed it at a process nobody had actually mapped out first. The tool isn’t lying to them and it isn’t broken. It’s doing exactly what it was built to do, predict the next likely step based on what it was given, and what it was given was inconsistent from the start. I’ve watched this exact pattern play out across businesses of every size, because a spreadsheet behaves and people don’t, and neither does a business process that only lives in someone’s head instead of on paper. This is the same discipline behind every Profit First conversation I have that has nothing to do with AI at all. You cannot get clean cash flow, clean pricing, or a clean profit number out of a business that has never standardized what “clean” even means for that business. AI didn’t create that problem. It just made it impossible to ignore any longer, which, honestly, might be the most useful thing it’s done for a lot of the owners I talk to, because it finally forces the conversation about process that most fractional CFO and small business finance work has been trying to have for years without a tool loud enough to make the owner listen.
The One Thing
Before you plug AI into anything else in your business, pick one process, just one, and write down exactly what information goes in, what the process does with it step by step, and what the finished output has to look like every single time. Don’t automate it yet. Just get it standardized on paper first, the way you’d document a pricing formula or a cash flow forecast before handing it to someone new. Once that document exists, automating it becomes straightforward, almost boring. Skip it, and you’re just teaching a very fast system to repeat your current mess with more confidence than it’s ever deserved.
You didn’t do anything wrong by trying AI before your processes were ready. Almost everyone does it in that order, and the tool companies aren’t exactly rushing to tell you otherwise, because a confused customer still pays the subscription fee. The fix isn’t throwing out what you bought. It’s going back one step, standardizing what the tool is supposed to run on, and then letting the technology do what it’s actually good at once it finally has something clean to work with. The owners who get this right aren’t the ones with the biggest tech budget. They’re the ones willing to admit the process was never really written down in the first place, and to fix that before buying anything else.
About Jeff MacPherson
Jeff MacPherson is a three-time tech founder and co-founder of Cloud37. He built the first Instagram DM automation platform, processing more than 1 billion messages for clients including Tony Robbins, Nelk Boys, and Atlantic Records before Facebook shut it down and later released competing functionality. He later became a top creator on Crypto.com, launching an NFT project that generated $2.5 million in under 14 days with zero ad spend. Today, Jeff helps coaches, consultants, and service-based businesses transform their expertise into AI-powered Expert Studios that create scalable client experiences, recurring revenue, and lasting business value.
Links
Website: http://www.cloud37.ai
Facebook: https://www.facebook.com/jeffmacp/
Instagram: https://www.instagram.com/jeff.macpherson/
LinkedIn: https://www.linkedin.com/in/jeffmacai/
Profit Blueprint Calculator I Profit Comes First: https://lp.profitcomesfirst.com/profitblueprintcalc-page
Watch the full episode on YouTube: https://www.youtube.com/@profitanswerman
Sign up to be notified when the next cohort of the Profit First Experience Course is available!
Free Copy of the Profit Blueprint Book: https://lp.profitcomesfirst.com/landing-page-page
Monthly Newsletter signup: https://lp.profitcomesfirst.com/newsletter-signup
Relay Bank (affiliate link): https://relayfi.com/?referralcode=profitcomesfirst
Profit Answer Man Facebook group: https://www.facebook.com/groups/profitanswerman/
My podcast about living a richer more meaningful life: http://richersoul.com/
Music provided by Junan from Junan Podcast
Any financial advice is for educational purposes only and you should consult with an expert for your specific needs.