Farrukh Butt

I ve been using AI coding tools Claude and Codex more often lately, and I keep running into the same problem:

1. I explain the task clearly

2. The agent writes a big chunk of code

3. The code looks correct at first glance

4. I start reviewing it and realise I don t fully know why some decisions were made

5. Now the fast part becomes slow because I have to reverse-engineer the changes

The issue isn t that the AI writes bad code every time. Sometimes it writes useful code. The harder part is staying confident that I understand what changed, why it changed, and what might break later.

I have been thinking about this recently. When I discover a new SaaS product, I don t only judge it by the design or features. I usually look at whether the homepage explains the problem clearly, if the pricing is simple, if there are real screenshots, and if the team/founder feels reachable.

Sometimes, a simple product with clear messaging feels more trustworthy than a very polished product with vague promises.

What do you personally check before trusting or trying a new SaaS product?

We launched a SaaS recently and noticed a cluster of accounts created within minutes of each other that never confirmed email or returned after signup. We want to keep our user table healthy without accidentally deleting real users.

Curious what other founders do in practice, more specifically:

• Do you auto-delete accounts after X days without email confirmation? If so, what timeframe worked for you?

• Do you keep soft-deleted rows for analytics or GDPR reasons? Any retention windows you recommend?

• If you have automated detection, what signals did you use (IP clusters; identical metadata; signup rate spikes)?