Feedback on AI review flow

Hi team,

First, the AI-generated follow-up questions are excellent. The contextual analysis by sector is very strong and becomes even better with proper briefing.

However, I have an issue with the current review flow.

When a customer gives a high rating (e.g. 5 stars), the system still allows follow-up answers that include clearly negative wording (e.g. “waiting time was disappointing”). The AI then combines positive and negative feedback and suggests copying this directly into a public Google review.

This creates a problem:

A satisfied customer can unintentionally publish a mixed or negative review on public platforms.

In my view, this is not ideal for two reasons:

It distorts the initial high rating (NPS-style inconsistency).
It increases the risk of users posting unintended negative public reviews.

What I would suggest instead:

The AI should detect negative sentiment keywords in follow-up answers.
If negative sentiment is detected after a high rating, the flow should switch from “public review generation” to an internal feedback form (private feedback / support loop).
Ideally, there should be a final AI “review check” step before anything is sent to Google, to validate consistency between rating and text.

In short, the system is very powerful, but it needs a stronger gating logic before publishing public reviews.

Happy to discuss if needed.

Best regards,

Rudy Mence