# Everyone’s adding AI features. Few users asked.

By Venus Paneliya, full-stack AI engineer. Category: AI engineering. Originally posted on LinkedIn.

Everyone’s adding AI features.

Nobody’s asking whether their users actually need them.

I’ve had four conversations with founders this month that followed the same pattern.

“We need to add AI to the product.”

“Why?”

“Our investors are asking about our AI strategy.”

Not "our users are asking for it." Not "we identified a workflow that AI improves." Not "we tested it and it reduced time-to-value."

Investors are asking. So founders are building.

This is how you end up with:

- A chatbot on your homepage that nobody uses.
- An "AI-powered" search that returns worse results than a basic filter.
- A recommendation engine that suggests the same three items to every user because the training data was too thin.

The best AI features I’ve built started with a user problem, not an investor question.

One founder said, "Our users spend 45 minutes filling out forms manually."

We built an AI form generator. Usage was immediate. Value was obvious. Manual editing dropped 60%.

Another founder said, "Our support team answers the same 20 questions repeatedly."

We restructured the product to surface those answers before users needed to ask. Support queries dropped 40%.

No chatbot. No "AI-powered" label. Just better architecture.

The pattern is simple.

- Start with the user pain point.
- Evaluate whether AI solves it better than a simpler approach.
- If yes, build the AI feature.
- If no, build the simpler solution.

"AI" on your pitch deck means nothing if your users don’t experience the value.

Have you added AI features because users asked, or because investors expected it?

Canonical: https://venuspaneliya.com/writing/adding-ai-features-few-users-asked/