Will AI eat my niche?

Someone typed your question into a chatbot last night. Not a version of your question. Yours. The one your business exists to answer.

They got a decent answer in four seconds, for nothing.

You have been turning that over for a while now. The question deserves a straight answer rather than reassurance. Part of what you sell is being commoditized. Part of it is becoming more valuable. Almost everything depends on which part you have been charging for.

What the customer was buying

Chegg is the case people point at, and it is worth looking at closely.

Students paid a monthly fee for access to worked answers. Textbook solutions, homework explanations, a library of questions somebody had already solved. A real business, built over 20 years, worth about $14 billion at its 2021 peak.

Then answers stopped being scarce.

In the quarter ending June 2026, Chegg reported revenue of $51.8 million, down 51% year over year. Its filings name the causes directly: Google's AI summaries and the spread of free generative AI tools. The company is still operating, profitably, at a fraction of its former size.

Chegg did not get worse at explaining calculus. The explanation stopped being worth a subscription. The same explanation arrived free, faster, and shaped to the exact question.

That is the pattern. Ask what the customer was buying, and the answer for Chegg was access to information that already existed somewhere.

The part a summary cannot reach

Be careful with the comfortable version of this argument. People say AI can only produce generic content, and that stopped being true a while ago. Give a good model your situation in detail and it will reason about your situation, not somebody else's.

The line sits somewhere else.

A model has not sat with 300 clients in your field. It did not watch which ones came apart in year two. It does not hold what you learned from the engagement that went badly. None of that is written down anywhere public. It cannot hear what a person tells you and notice what they left out.

And it is not answerable. Nobody is on the hook when the advice is wrong. No license, no liability, no reputation staked, no phone call in March when the decision gets tested.

That is what people have always paid experts for. Not the information. The application of it to their situation by someone who carries the consequence.

The test worth running on your own business

Three questions, and they are uncomfortable on purpose.

Could a diligent stranger reproduce your product by reading widely for a week? If the answer is close to yes, the market has already noticed.

Does your buyer's result depend on someone looking at their specific situation? Or does your product deliver the same thing to everyone who buys it?

When the advice is wrong, who pays for it? If the answer is nobody, you were selling information.

Answering those three tells you how much repositioning the business needs. For most expert-led businesses the answer is some rather than all. The parts under pressure are usually the parts you were least proud of anyway.

What repositioning looks like

The move is from the information to its application.

The course that taught a method becomes the program where you review the work the student produced. The report becomes the standing arrangement where you interpret what changes. The book stays a book, and stops being the product; it becomes the reason someone calls. The format questions that follow are their own subject.

Two things are worth saying about the shape of the business that comes out the other side.

It is usually smaller in unit volume and larger in price. Judgment applied to one situation does not scale the way a file that downloads scales. It also depends more on you personally, and that is the trade. The thing that cannot be copied is attached to a person.

Then there is the part that does scale, which is your reputation for having the judgment. That is what carries a name into rooms you are not in, and into answers you did not write.

Explanation is free now. Being answerable is not.

What you know is still worth what it was worth. What changed is that telling people what you know is no longer the business. It is the marketing.