It happened. Customers don’t hate talking to AI anymore.

Yeah AI is a great way to scale your support team, but if your org isn't obsessed with ✨delighting the customer✨, your AI agents will fail. Here's why:

A year or two ago, putting a bot in front of customers felt risky. You gave it a goofy “clippy” personality but people were still smashing the equivalent of a the phone # to get to human who might be warm and understanding.

Today, "AI-style" IS often associated with warm and understanding. I recently saw a study that about 1/3 of U.S. adults regularly use ChatGPT (58% of under-30s), so AI chatting now feels feel normal.. maybe even preferred? AIO search is also on the rise - many customers will skip YOUR bot and do their own research with THEIR AI tool.

But momentum isn’t inevitability - plenty of “agentic AI” projects are getting shelved for murky ROI or immature tech.

What works

🤖 AI front door, human fast-lane. Let AI handle instant answers/triage; keep a clear “talk to a person” with tight handoff SLAs. You want to get the best of both worlds for the person who needs it.

🤖 Scope for certainty. Start with policies, billing, membership management, availability, pricing... and places with one source of truth. Hallucinations are real, ambiguity is your enemy, and confidence gating is hard. No one wants to be the customer on the other end of that.

🤖 Build in routing by intent or sentiment. If you don’t understand these triggers for your customers, do NOT try AI first. AI will make things worse before they get better.

🤖 Build with talented teams who obsess over edge cases, content integrity, and most importantly are first focused on absolutely ✨delighting the customer✨

Because at the end of the day, if ✨delighting the customer✨ isn't your main north star, your AI is going to p*ss people off, and you will be sitting around scratching your head as to why its not working.

Previous
Previous

Be a legend to your team.