Who shouldn't
hire us.
Every vendor page says who they're for. This is the faster question. Six disqualifiers, what to do instead in each, and the narrow case that's left.
Six situations where hiring us is the wrong call. Not softened, not framed as a challenge we love. Wrong.
Every vendor page tells you who they're for. Almost none tell you who they aren't for, which is strange, because that's the faster question and it's the one you're actually trying to answer.
So this is the disqualification list. If you land in one of these, the honest recommendation is in the section with it. No call required.
The six
Someone in your building already built a working version
Not a plan. Not an interest. Not a good prompt they showed you once. Something running that a second person has used.
This is the clearest disqualifier on the list and it isn't close. When there's already a builder inside a company, the internal version wins, and it usually should. They know the business, they're already paid, and they'll still be there in month seven.
Do this instead. Fund the person who started. Give them the time formally rather than as a side project, put a review schedule on the output, and write down how it works so it survives them leaving. Then read the seven questions, which is the same argument at length.
Your process is genuinely standard
Payroll. Scheduling. Invoicing. Email marketing. Support ticketing.
If a software category exists and thousands of companies run the same shape of the process, a custom build is a slow way to spend money on something you can buy on Tuesday for a fixed monthly fee.
Do this instead. Buy the category tool and adapt your process to it. The custom work only pays when the process is genuinely yours, which is rarer than it feels from the inside.
You need breadth, not depth
Six workstreams at once. Paid media and content and lifecycle and design and analytics, all moving in parallel, all this quarter.
That's what an agency is built for and it isn't what one operator can do. We'd take the money and then be the bottleneck, and you'd find out in month two.
Do this instead. Hire an agency, and ask two questions: who is actually doing the work, and how much of their week do we get. Ask both again in month three.
You can't say what a month without it costs
A rough number is fine. An order of magnitude is fine. Two hours a week times a loaded hourly rate is fine.
But if no number exists at all, the project has no cost of delay attached, which means it loses to every urgent thing for as long as it exists. It'll get descoped, then paused, then quietly forgotten, and everyone will conclude the AI didn't work.
Do this instead. Spend an hour producing the number before you spend anything else. If the number turns out to be small, you've just saved a quarter. That's a real result.
Nobody will own it after we hand it over
Not the build. The Tuesday after it breaks.
Someone has to notice when the output stops being true, and that person has to exist before the work starts. An unowned system is the most expensive outcome available, because you pay for it twice. Once to make it, and again to work around it once nobody trusts it.
Do this instead. Name the owner first. If no name exists, fix that before hiring anyone. If the honest answer is that nobody has room, that's real information and it points at a hire rather than a project.
You want the number to move this month
Connecting systems takes weeks. Getting an answer engine to name your company takes longer than that, usually thirty to sixty days before anything moves and ninety before it compounds.
Anyone who tells you otherwise about either one is selling you a timeline, not a system.
Do this instead. If the month matters that much, spend it on paid acquisition, which is the only channel that turns on that fast. Come back to the operating work when you can afford a quarter.
Who this is for, since you read this far
The inverse of the list, and it's narrow on purpose.
A founder-led company between five and fifty people. A process that touches three or four systems that were never designed to talk to each other. Nobody internal with the room to own it. And a reason it has to still be running in six months rather than just be finished.
That's it. It excludes most companies and it's supposed to.
If that's you, the entry point is the Diagnostic. Two weeks, ends in a written scope and a fixed fee, and it's designed so that walking away with the scope and building it yourself is a legitimate outcome.
Two reasons, and the second one is the honest one.
First, a disqualification saves us both a call, and calls are the scarcest thing either of us has.
Second, we'd rather lose the deal in week zero than in month two. Every situation on this list is one where the work fails after the invoice clears. That's worse than not selling it, for us more than for you.
Keep reading
The 6-Point Check For AI Work Before It Leaves Your Company
Most AI failures don't look like errors. Here's how to grade output you didn't build, and turn it into an error rate you can actually act on.
Read the guide → Operations · 6 min readThe 30-Minute AI Inventory: What Is Running In Your Business And Who Owns It
Four questions, four groups, one specific order. Surfaces the undocumented setup carrying real volume before the person running it resigns.
Read the guide →None of the six
apply to you?
Thirty minutes, no deck. Bring one messy process and the list of systems it touches. If one of the six shows up on the call, we'll name it there.
Book a 30-minute call →