At a glance

  • I named a retailer in the prompt.
  • I'm asking my various AI assistants to help me with a lot of tasks, but there's one thing that's quite annoying: advanced bot protection on websites.
  • Yes.

I'm asking my various AI assistants to help me with a lot of tasks, but there's one thing that's quite annoying: advanced bot protection on websites.

I can understand why it's in place. But I'm trying to get things done, especially in the context of research, pricing, and everything-but-the-click-to-order.

For instance, I asked one of my agents to check out the most modern, best priced iPad under £180 on sale at one leading UK vendor of refurbished and recycled devices. My daughter had broken hers and it was sadly at least £160 to get the screen repaired.

Nothing would work to query this particular website, because they've deployed some big, beautiful, amazing bot protection capabilities that got in the way of the agent doing its research.

Eventually the agent suggested trying a competitor's website. I agreed. It did the research successfully, no bot protection, and guess where I ordered from?

Yes. I rewarded the competitor. This is despite me instructing my AI agent to visit the original site by name.

It's a fascinating issue. I understand the frustrations and the serious reasons that require many sites to deploy ultra-advanced bot prevention tools. But there is a big but here.

My agent is working for me.

For me. So when you're making it incredibly difficult for the AI agent to browse and interact with your website, you're making it really difficult for me. Me, the customer, who specified your brand by name in the prompt.

I am most certainly bleeding edge here. But I wonder how long before some marketing and sales optimisation teams start thinking about customer agents rather than just customers.