Alice Hayes

I tested 10 prompts like "best tool for X" in ChatGPT, Perplexity and Gemini. Our product came up in 2 of 30 answers, and a competitor with less traffic showed up in 14.

Curious how others approach this:

• Do you track AI mentions at all, or is it too early to matter?

• Has anything you changed (docs, reviews, comparison pages) actually moved the needle?

• Is "AI search visibility" a real channel yet, or mostly hype?

The more AI customer support projects I work on, the more I feel support is only the starting point. Most teams begin with fairly obvious use cases: answering FAQs, handling tickets, updating orders, or triggering refunds.

That makes sense. These problems are easy to define and the value is easy to measure.

But once the AI is connected to customer data, CRM, billing, product usage, and other systems, its role can become much broader.

We segmented 9,000 accounts by login frequency and compared to 12-month retention.

Daily logins: 41% retained
Weekly logins: 68% retained
Monthly logins: 79% retained
Quarterly logins: 71% retained

The users who logged in least were the most retained, up to a point. The daily users were often doing manual work that the product should have automated. The monthly users had set up automations, integrations, or notifications and only came back when something needed a human decision.

We had been treating login frequency as a health metric and using it to trigger "we miss you" emails. We were emailing our best users to tell them they were not using the product enough, when their lack of logins was the evidence that the product was working.