Nearly every week someone tells me they think they're falling behind on AI. It comes out the same way each time, usually near the end of a conversation about something else, in a slightly apologetic voice.
The anxiety is useless. The gap it's pointing at is often real. Here's how to find out which you're dealing with, in about twenty minutes, without reading another article about the future of work.
Separate the noise from the gap
A lot of money is currently being spent convincing you that you're late. That's a sales tactic and an old one. When someone tells you the window is closing, check whether they happen to sell windows.
Underneath that, though, something real is happening. A share of your competitors have quietly got faster at a few specific things. Not smarter. Faster. That's the actual gap, and unlike the anxiety, it's measurable.
The twenty-minute test
Stop reading about AI and go and look at your competitors. Pick three — the ones you actually lose work to.
Ring each of them as a customer with a real question. Then fill in their website contact form with the same question and start a stopwatch. Then, if they've got one, ask for a quote.
You're measuring the only things a customer notices: how long until someone responds, and whether the response is useful. Not whether they use AI. Whether they're faster than you.
When I've run this with clients the results split cleanly. Sometimes all three take two days to send a generic reply, and the client's anxiety evaporates on the spot because they reply in four hours and had no idea that was a competitive advantage. Other times one replies in eleven minutes with something specific, and the client goes quiet, because the vague dread just became a number they can't unsee.
That number is the whole thing. Everything else is atmosphere.
What being genuinely behind looks like
In my experience it almost never looks like missing technology. It looks like these.
You take more than a day to respond to an enquiry and your competitors don't. Speed of first response is the most consistent predictor of who wins a job in most service businesses, and always has been. AI matters here only because it made being fast much cheaper.
Your customers can't get an ordinary question answered without talking to a human. Every one of those is a cost to you and a friction for them, and some of them don't bother making contact at all.
You don't know your own numbers. How many enquiries last month, how many became jobs, where they came from. If you can't answer that, no tool will help, because you won't be able to tell whether it did anything.
The same information gets typed into two places by a person every week. That one compounds quietly and it's the easiest to fix.
None of those are technology problems. They're operational problems that technology now solves cheaply.
What being behind doesn't look like
Not having a chatbot. Not posting generated content. Not owning an AI strategy document. Not knowing what a language model is.
A business that has bought three subscriptions nobody opens is further behind than one that has bought none, because it has spent money, learned nothing, and now believes the box is ticked.
If the test says you're fine
Then you're fine today. Put it down, and run the same test in six months.
That's not complacency, and it isn't permanent either. The reason to re-run it is that the cost of being fast keeps dropping, so the competitor who was two days behind you can close that gap in a fortnight now if someone decides to. The test is cheap. Repeat it.
If the test says you're not
Then you've got something much better than anxiety: a specific, named gap.
A competitor answering in minutes while you answer in days is a problem with a handful of solutions, most cheaper than people expect and some of which aren't AI at all. That's the position you want to be in before you spend anything. Not "we should do something about AI", which is unbuyable and unmeasurable, but "we lose jobs because we're slow to quote, and here's what we're doing about it".
One of those sentences leads somewhere. The other leads to a subscription.