Pick one process your team complains about. Grab a pen. Answer these five questions honestly.
- How often does it happen? Weekly or more is a good sign. Once a quarter usually isn't worth it.
- How long does it take each time? Multiply by how often. That's your hours a month.
- Does it follow the same steps most of the time? If every case is completely different, AI helps less.
- Is the information it needs already somewhere digital? If it lives in one person's head, step one is writing it down.
- What happens if it's wrong? Low stakes: automate it end to end. High stakes: AI drafts, a person approves.
Reading the answers
If the first four point the right way and you're clear on the fifth, it's worth automating. If two or more don't, fix the process first, or leave it alone. Not everything should be automated, and that's fine.
A worked example
Sending a new deal out to investors, at EAS Advisors:
- Often? Every week.
- Long? Hours each time.
- Same steps? Yes: read the deal, find the right people, write, send, track.
- Digital? Yes, but spread across five places. Fixable.
- If it's wrong? A bad email to an investor matters. So AI drafts, a partner approves.
Verdict: worth it. It's now one of the jobs behind the 20+ hours a week they got back.
Want a second opinion? Our AI Roadmap runs this test across your whole team and ranks what to build first.