A writer in Entrepreneur argues that a small business can start working the AI way in one quarter, with no new software and no outside help. The idea is to pick one process that brings in money and redesign how it runs. The author says a young company has an edge because it has few old habits to undo.
Why the author says now
The piece leans on a June survey by Genpact and HFS Research of more than 2,000 senior executives. Of those, 85% said messy processes, unusable data, aging systems and unprepared people limit what they get from AI. The survey found only 33% of company data is ready for AI. It also estimated manual or inefficient work eats about 40% of a typical employee's week.
The author also cites a McKinsey survey from this year. Nearly three-quarters of top AI performers have fundamentally redesigned their workflows, versus roughly one-quarter of the rest. The author concludes the gap is redesign, not spending.
The plan
Start with one process tied to revenue, such as quote to cash, customer onboarding or returns. Have the people who run it describe the real path, including the hidden spreadsheet and the approval stuck in one inbox. Then mark each step as production (drafting, gathering, comparing, summarizing) or judgment (deciding and taking responsibility). By day 30, the goal is a one-page map a new hire could follow on day one.
- Trace where the information for each step lives, then name one owner and one definition for key fields like customer, price and status.
- Give each judgment step a name, a definition of a good outcome and a rule for when a human reviews before anything reaches a customer.
- Question the remaining approvals. The author says many survive only out of habit once data is visible to everyone.
How to know it worked
The author says to drop metrics that count tools or licenses and track the business instead. Measure cycle time, cost to serve and customer turnaround on day 61, then again on day 90. The difference shows whether anything changed. The author claims, from experience, that one fully redesigned process returns more than the first 10 pilots combined.
What comes after day 90
The author suggests repeating the quarter on the next process, since the second map takes half the time. Only then should a business buy tools, because it can say what a tool must do and measure the result. New vendors should also pass an export test, meaning the company can pull its data out, before a contract is signed.
The closing test is simple: if you hired 10 people on Monday, would they know by Tuesday how the work happens? This is one contributor's opinion, and the article offers no results from small businesses that followed the plan.
The takeawayBefore buying any AI tool, write down how one revenue process really runs, then measure it on day 61 and day 90.