Scott A ten-minute exercise that almost no business does and almost every business should.
How it works
Open a blank page. Two columns.
Column 1: What AI tools are we paying for?
Every subscription. Every licence. The Copilot rollout. The ChatGPT Team account. The thing your marketing person signed up for last quarter. The free trial that quietly converted. Everything.
Column 2: What are we actually using each one for?
One sentence per tool. What it actually does in your business this week. Not what it could do. Not what the sales deck said. What it is doing.
Where the value lives
The audit is in the gap.
- Tools with no sentence next to them are not earning their licence. They are costing you money every month for zero return.
- Tools where the sentence is vague ("we use it for various things") are the ones where nobody has been trained properly. The tool is capable, but the team is guessing.
- Tools where the sentence is sharp ("the sales team uses it to draft follow-up emails after discovery calls") are the ones quietly producing value.
The three decisions
Do this once. Three decisions usually fall out of it:
- Something to cancel. The tool nobody uses. The duplicate subscription. The enterprise tier when everyone is on the free version anyway.
- Something to train on. The tool people have tried but abandoned because the output was not good enough. That is not a tool problem. It is a training problem.
- Something to do more of. The tool one person is using well that nobody else knows about. Get them to show someone else.
Example: a 20-person professional services firm
| Tool | What it is doing this week |
|---|---|
| Microsoft 365 Copilot (20 seats) | 4 people use it for email summaries. 16 have never opened it. |
| ChatGPT Team (5 seats) | Marketing uses it for social media drafts. Nobody else has logged in since February. |
| Jasper (1 seat) | Free trial converted. Nobody remembers signing up. |
| Otter.ai (3 seats) | Used by one partner for meeting transcripts. Works well. |
Cancel: Jasper. Train: Copilot (16 people paying for a tool they do not use). Do more of: Otter.ai (show it to the other partners).
Ten minutes. Three decisions. That is the whole exercise.