Ran an AI agent on my membership page for a month. Conversion flat.

Case study

Author-reported results.

Starting situation

I'm in Canada with a small email list of ops folks, maybe 900 subscribers, and I'm trying to turn free readers into paid members. Open rates are okay but the join link gets clicks, not purchases. So I added an AI chat agent to the membership page, hoping it would handle objections and push people over the line.

What changed

I spent a weekend writing prompts. The agent greeted visitors, answered questions about what's inside the membership, and after three messages offered a join link. I tracked how many chats started and whether any became members.

Result

Over two weeks I saw maybe twenty-five chats. The agent answered most questions correctly, but almost everyone left before the join link. No new members from the chat. The one join I did get came from a direct reply to my email, where I answered personally.

Limitations

Small sample and my offer might just be weak. My real mistake was treating the agent as a closer instead of a qualifier. I can't conclude agents are useless, but for me the agent felt like a very polite receptionist that forgot to ask for the sale.