Is my business problem actually an AI problem?
Probably not. Before you pick an AI tool, name the commercial problem. A weak message, an unclear offer or no activity will not be fixed by a better stack.
Probably not. Before you choose an AI tool, finish the sentence "My actual business problem is..." If the answer is that the right people have never heard of you, or nobody has been asked to buy, a better stack will not fix it.
AI is everywhere, so ordinary business problems keep getting framed as AI problems. Which agent should I build, should I run a local model, which tool is better, what should I automate. Those can be valid questions. They are often downstream questions.
Why does everything start to look like an AI problem?
Because AI has become the lens for everything. It is a sophisticated way to avoid asking what is actually broken.
Asking how AI can get you customers feels productive. You get to research tools, compare options and build something. Asking why you are not getting customers is less fun, because the answer might be that nobody has been told, or that you have not asked.
What should I ask before I ask about AI?
Swap the question. Instead of asking how AI can get you customers, ask why you are not getting customers.
The honest answers are usually plain ones.
- Maybe the right people have never heard of you.
- Maybe they do not trust you.
- Maybe the offer is weak.
- Maybe nobody is being asked to buy.
The same split runs through how to get clients without sales calls. If the right people have never heard of you, no amount of following up changes anything. If they know you and nothing happens, they have not seen how you think.
If they trust you and the money never moves, the offer is unclear, the path to working with you has dead ends, or you never asked. Most people assume it is the first one, so they chase more reach and more activity.
None of those is an AI question. Answer the commercial one first, then decide where AI helps.
Can a hobby pass for a business priority?
Yes, and I know because I do plenty of this myself. Local models, harnesses, agents and new tools can be genuinely fun.
The trap is telling yourself a hobby project is the highest-priority commercial work. It feels like progress because it takes effort and it is clever.
You can have an AI hobby and an AI-enabled business. Just do not confuse the two.
When does AI actually help?
Once the business question is clear, AI is excellent. I use it to turn stories into posts, analyse reactions, build tools, run repeatable outreach, do research, repurpose content and update systems.
What it does is accelerate a direction. It does not magically prove the direction is commercially right. If the message is weak, you get a weak message produced faster.
Which tool does what, and what to hand over, is a separate decision. I cover that in AI-powered client acquisition, and the question of timing in when to automate a business process.
Can AI help me work out the strategy?
Absolutely. Brainstorm with it, research with it, simulate and challenge your assumptions with it. It is a good thinking partner.
Just do not confuse an AI-generated answer with market evidence. A model can tell you what sounds plausible. It cannot tell you whether the people you want to work with will trust you and buy, and only they can answer that.
| Where I am stuck | The non-AI question first |
|---|---|
| Which agent or model to use | What is my actual business problem |
| Not getting enquiries | Do the right people know I exist |
| People know me and nothing happens | Have they seen how I think |
| Interest but no sales | Is the offer clear, and have I asked |
How do I check which kind of problem I have?
Write the sentence. "My actual business problem is..." and finish it.
The answer should be commercial or operational. If it starts with "I need an agent that", you have described a tool, not a problem. Go back a step.
I use ridiculous amounts of AI, and that has made me more convinced that sometimes the right answer is not another AI tool. Solve the business problem. Then bring in the machine.
KnownTrustedBought
Join my daily email