I just spent an hour with ChatGPT planning my business strategy, and I am definitely a genius. Which probably means I’m about to make a terrible decision.
Here’s the thing I’ve noticed about business owners using AI for strategic planning: they come away from these sessions feeling incredibly validated. Everything suddenly makes sense. Their plans look professional, thorough, and perfectly logical.
And that’s precisely why they should be worried.
Not because AI is stupid, but because they’ve accidentally trapped themselves in what I call the “Business Echo Chamber of One.” And if you’re using AI for business strategy (which you probably are), you’re likely stuck in there too.
The Seductive Lie of Frictionless Advice
Think about the last time you asked a business mentor for advice. They probably asked annoying questions. Challenged your assumptions. Maybe even said something like, “Hang on, didn’t you try something similar last year and it didn’t work?”
That friction is uncomfortable, but it’s valuable. Your accountant pushes back on your financial projections. Your operations manager questions unrealistic timelines. Your business partner asks if you’re solving the right problem.
AI doesn’t do any of that.
I’ve watched business owners describe their “difficult clients” to AI, and it immediately starts helping them manage difficult clients. It doesn’t ask whether boundaries have been unclear or if service delivery has been inconsistent. It just takes the “difficult client” framing as fact and builds a strategy around it.
This feels amazing in the moment. Finally, an advisor that doesn’t question your judgment! No uncomfortable conversations about whether you’re seeing things clearly. Just sophisticated-sounding solutions that validate exactly what you were already thinking.
Which is precisely why it’s dangerous.
How Business Owners Build Perfect Wrong Strategies
Here’s the pattern I see over and over:
What they tell AI: “Our sales process takes too long. Prospects want multiple meetings, custom proposals take forever, and we’re losing deals to faster competitors.”
What AI gives them: Complete “productised” strategies. Standardised packages. Fixed pricing. Streamlined proposals. Even copy for sales pages.
What they don’t tell AI: That their conversion rate is actually decent. That the real bottleneck might be personal discomfort with follow-up calls. That most lost deals aren’t about speed—they’re about budget timing or poor qualification.
The result: Brilliant strategies for solving the wrong problems.
This is how the echo chamber works. AI can only see what you show it. Your stress, your assumptions, your selective memory—that becomes its entire reality. It takes your incomplete story and makes it sound complete.
You give it frustration with time-consuming processes, and it hands back strategies to avoid those processes. Perfect execution of flawed thinking.
The Stuff AI Can’t See
The thing is, there’s so much context that never makes it into conversations with AI:
The external stuff: Like how three competitors just raised prices, making your “slow” proposals suddenly look reasonable. Or that economic uncertainty has everyone making longer decisions.
The internal stuff: Like how your team’s been working overtime for months and might not have capacity for process overhauls. Or that your “difficult” client actually gives you the most referrals.
The personal stuff: Like how you always get excited about efficiency projects when stressed about something else. Or that you have a pattern of changing systems instead of addressing skill gaps.
Human advisors who know you spot these patterns. They ask questions that make you uncomfortable. AI just builds beautiful plans on top of your blind spots.
The Pattern You Probably Recognise
Here’s what this looks like in practice:
A marketing consultant gets fed up with writing custom proposals for every prospect. Asks AI to help “productise” services. Gets back a gorgeous strategy: three-tiered packages, standardised pricing, streamlined onboarding.
Six months later? The packages aren’t selling. Prospects want customisation that the new model can’t provide. They’d solved the wrong problem.
The real issue wasn’t custom proposals—it was discomfort with sales conversations and poor follow-up. Conversion rates on interested prospects were terrible, but they’d focused on the part of the process they could systematise instead of the part that required personal growth.
Or the business owner who asks AI to help with “employee retention issues.” Gets a comprehensive strategy covering compensation, benefits, work-life balance, and career development paths. Implements it all.
Retention doesn’t improve.
The real problem was micromanagement, which had increased as the business grew. The team felt trusted to deliver but not to decide. No amount of benefits packaging was going to fix a management style issue.
These aren’t isolated cases. This is the pattern when you use AI to validate thinking instead of challenge it.
So What Do You Actually Do?
Look, I’m not saying don’t use AI for business stuff. It’s brilliant in execution. Just terrible at forming ideas where so much context is missing.
Here’s what I’ve learned works:
Use it as a research assistant: “Find me data on industry trends” or “What are common pricing models for professional services?”
Use it for scenario planning: “If this strategy fails, what would the warning signs look like?”
Use it to stress-test plans: But only plans you’ve developed somewhere else first.
The killer technique: Before implementing any AI strategy, try this prompt: “Act like a sceptical business advisor who thinks this plan is flawed. What are the three biggest problems you can see? Be brutally honest.”
This forces it to work against your confirmation bias instead of feeding it.
The Real Test
Here’s the thing that’s changed how I think about this: before you ask AI for strategic advice, ask yourself one question:
“Have I talked to another human about this problem?”
Not just vented about it. Actually discussed it with someone who might disagree with you.
Because AI will always make your thinking sound smarter, it’ll take your half-formed frustrations and turn them into professional-looking strategies. It’ll validate your assumptions and polish your biases until they shine.
But sometimes the most valuable business advice doesn’t start with “Here’s the plan.” Sometimes it begins with two words no AI will ever say to you:
“You’re wrong.”
Your AI assistant is brilliant at helping you execute the strategy you’ve formed elsewhere. Let it research, plan, and implement. But when it comes to deciding where your business should go next? Talk to humans first.
Trust me on this one. The hour you spend getting uncomfortable feedback from real people will save you months of perfectly executing the wrong plan.
Even if it doesn’t feel as good as having AI tell you you’re a genius.
