Letting content go stale
If you change your pricing or policies and never let the bot re-crawl, it'll confidently give outdated answers — which is arguably worse than giving no answer, since a visitor has no reason to doubt a confident-sounding response. Make sure your platform re-crawls automatically on some schedule, and re-trigger it manually after major content changes rather than assuming it'll catch up on its own timeline.
This mistake compounds silently: a small pricing error might go unnoticed for weeks, quietly quoting the wrong number to every visitor who asks, until someone finally flags it.
No path to a human
Trapping a frustrated visitor in a loop with no way to reach a person is the fastest way to turn a helpful tool into a source of real complaints. Always have a visible way out — email, phone, or a handoff — and don't bury it three menus deep where a frustrated visitor won't find it.
This mistake is especially damaging because it happens exactly when a visitor is already annoyed — the chatbot failed to answer, and now they can't even find a way around it. That's the worst possible moment for a dead end.
Turning it on and never checking transcripts
A chatbot isn't 'set and forget.' Reviewing real conversations occasionally is how you catch bad answers before they've been repeated to fifty visitors — the businesses that get the least value from a chatbot are almost always the ones that installed it once and never looked at what it was actually saying afterward.
This doesn't need to be a heavy process. Even a quick monthly skim of recent transcripts catches the majority of issues before they become a pattern.
Overpromising in the welcome message
If your greeting implies the bot can do things it can't — process a return, check an order status, override a policy — visitors will be frustrated when it can't follow through on what it seemed to promise. Keep the framing honest about what it's actually good at: answering questions from your site, not acting as a full self-service portal unless it's actually integrated to do so.
Ignoring how it looks on mobile
A widget that looks fine on a desktop browser but covers half the screen or breaks the keyboard layout on a phone is a mistake that's easy to make and easy to miss, if most of your own testing happens on a laptop. Given that most site traffic for many businesses is mobile, this is worth checking directly on an actual phone, not just assuming it translates.
