Book a Complimentary Session

How to tell whether your AI is still working well

Four things to look at regularly once AI is running, and what to do when one of them moves.

Launch day is the start. Your business keeps changing, and AI that was right in spring can drift by autumn. A few plain measures, looked at regularly, tell you whether it's still pulling its weight.

Is it still fast?

Track the measure you cared about at the start, such as time to answer or requests cleared the same day. Compare it with your before figure as well as last week's.

Is it still right?

Have someone who knows the work read a sample of what AI handled. Count how many were right, partly right or wrong. A slow slide here matters more than any gain in speed.

What is reaching your people?

Look at the cases passed to your team. Are they the ones that need a person? Is the note useful? If the same kind of case keeps arriving, AI may be able to handle it, or the rules may need a tweak.

What does your team say?

Ask the people working alongside it what has improved and what annoys them. They'll often spot a problem before the numbers do.

What should you do when something moves?

  • Check whether something changed in the business: a new product, a price or a policy.
  • Read a handful of recent cases to see the problem first-hand.
  • Fix the cause, then watch the same measure to confirm it's better.

Questions for whoever runs your AI

  • How do you spot a fall in quality?
  • How do updates get made when our business changes?
  • How will we see what has improved?

For the first month, read measuring an AI pilot. For the bigger picture, see the questions to ask a managed AI provider.

Have a challenge like this?

Bring it to a Complimentary Business + AI Opportunity Session, 45 minutes with Brian Pinkus, our Founder & CEO.

Pick a time that suits you. It goes straight onto Brian's calendar.