We decline roughly one in five prospective clients — not because their businesses are bad, but because automation would fail or underdeliver given where they are. This diagnostic is the internal checklist we use in scoping calls, published so you can run it yourself for free.
Score one point for each yes. Eight or higher: automation will very likely pay for itself quickly. Five to seven: automate selectively, starting with your highest-scoring process. Below five: fix the fundamentals first — automation amplifies whatever process quality you already have.
The ten questions
1. Is there a task your team performs more than 50 times a week in roughly the same way? 2. Can you name its current cost — hours, error rate, or missed revenue? 3. Does the knowledge needed for the task exist in writing (policies, scripts, templates) rather than only in someone's head? 4. Do your core systems (CRM, calendar, helpdesk, accounting) have APIs or common integrations? 5. Is someone on your team able to spend two hours a week with an implementation partner for a few weeks?
6. Can you tolerate a supervised pilot where a human reviews the AI's work before full autonomy? 7. Is the process's volume growing faster than you want to hire? 8. Are there measurable customer-experience gaps — response times, missed calls, backlogs? 9. Does leadership agree on what success looks like in numbers? 10. Is your data (customer records, documents, policies) accessible enough that a new employee could learn the job from it?
Question 10 is the sleeper. Ask yourself: if you hired a sharp human tomorrow, could they learn this job from your documentation and systems? An AI employee learns the same way. If the answer is no, the first automation project is often fixing exactly that — and it pays dividends with every future hire, human or AI.