Most founders we talk to are not behind on AI because they are lazy. They are behind because every article says something different, every tool demo looks amazing, and nothing tells them what to do first on Monday.
Here is the framework we use in every audit. It fits on a napkin.
Ask three questions about every piece of work
Take any recurring task in your company and ask:
- Does it happen at least weekly? If not, automating it will never pay back the build cost. Skip it, whatever the demo looked like.
- Can you write down what "done correctly" means? If two of your people would disagree about whether an output is right, an AI will not settle it. Fix the definition first, automate second.
- What happens if it is wrong once? If the answer is "mild annoyance", automate it fully. If the answer is "we lose a customer", automate it with a human approval step. If the answer is "we get sued", leave it with a human and give them better tools.
That third question is the one everybody skips, and it is the one that decides whether your team trusts the system or quietly stops using it.
The order that works
Start with work that is frequent, well-defined, and cheap to get wrong once. In most companies of ten to a hundred people that means the same short list: meeting notes and follow-ups, inbound triage, weekly reporting, and document data entry.
Do not start with the clever thing. The clever thing is usually rare, fuzzy, and expensive to get wrong, which is the exact opposite of what you want first. You want an early win the whole team can see, because the second automation is ten times easier to land when the first one is quietly working every day.
Why we give this away
This framework is the first hour of the free opportunity audit we run. The rest of the audit is applying it to your actual workflows and pricing the automations worth building. If you can run it yourself from this article, good. That is a company that will be a better client later, or a better operator without us.