You bought ChatGPT licences. Nobody uses them. Here's why.
It's rarely the tool. It's that nobody showed the team which of their tasks it's for. How to turn unused AI licences into everyday habits.
Daine MawerFounder, Plaincraft
3 min read
The short answer
Unused AI licences are almost never a tool problem. People don't use ChatGPT or Copilot because nobody showed them which of their own tasks it helps with, what they're allowed to put into it, and what good output looks like. Fix those three things and usage follows.
The pattern I keep seeing
A business buys licences for the whole team. There's an email announcing it, maybe a lunchtime demo. For a week or two people try it: a poem for someone's birthday, a summary of a long email. Then usage drops off, and a few months later someone asks whether the subscription is worth renewing.
The people who kept using it usually worked out a use for it on their own. Everyone else didn't, and nobody followed up.
Why people stop
- They don't know what it's for. "Use AI to be more productive" isn't a task. Nobody wakes up wanting to be more productive in general.
- They're not sure what's allowed. Without clear rules, careful people stay away from it, which is the right instinct and the wrong outcome.
- The first answer was poor. Most people try once, get something generic, and conclude it doesn't work for their kind of job.
- It sits outside their tools. If it means opening another tab and copying text back and forth, it loses to habit.
What works instead
Start from the work. Sit with each team for an hour and list the tasks they repeat every week. Pick two or three that involve writing, reading documents or answering questions, then show people how to do those tasks with the tool, using their own documents.
Then write down the rules on a single page: what can go in, what can't, and who to ask. (There's a separate guide on writing an AI usage policy.)
A two-week reset
- Week 1Ask each team for the three tasks they repeat most. Pick one per team.
- Week 1Run a one-hour session per team on that task only, with their real examples.
- Week 2Share the one-page rules. Name one person per team to collect questions.
- Week 2Check back: who's using it, for what, and what's getting in the way?
If usage is still low after that, the licences may genuinely be the wrong fit, and that's worth knowing before you renew. Hands-on team training is built around exactly this.
Written by Daine Mawer
Founder of Plaincraft, an independent AI consultancy for South African SMEs, with 17 years of building websites and digital products behind it. AI is part of how I work every day, so I know where it earns its keep and where it doesn't.
More about me
