The real cost of "I'll get to AI eventually"

"Eventually" is a reasonable plan for a lot of things in business. New software, a website redesign, a CRM overhaul — these things can usually wait a quarter or two without much consequence.

AI adoption might be different, and not for the reasons you'd expect. It's not that the technology will disappear if you wait. It's that the businesses around you — including, statistically, your competitors — aren't waiting, and the gap between "adopted" and "not yet" tends to compound rather than stay flat.

Canada's productivity problem, in one number

Zoom out for a second. Between 2015 and 2025, labour productivity in Canada grew by just 3%. In the United States, over the same period, it grew by 18% — roughly six times faster.

The OECD has gone as far as warning that, without significant change, Canada is on track to rank last among advanced economies for GDP-per-capita growth over the next four decades.

That's a big, abstract, economy-wide number. But it's built from small, concrete, business-level decisions — millions of them, made by individual owners deciding whether this is the year they finally connect their booking system to their CRM, or whether that's a "next year" problem again.

Why "the gap" compounds

Here's the part that matters at the level of an individual business. The same survey that found Canada's productivity numbers also broke down returns by how digitally mature a business is — and the gap between the least and most mature businesses isn't small:

Now think about what that means over time. A business earning $2.40 on every dollar of technology spend has more margin to reinvest — in marketing, in hiring, in better service — than a business earning $1.40 on the same spend. That extra margin compounds. Next year, the gap is bigger. The year after, bigger again.

This is the mechanism behind "the gap is widening." It's not that AI adoption is a one-time event that either happens or doesn't. It's that each year of delay is a year of compounding at the lower rate, while competitors who've moved compound at the higher one.

This isn't a "move fast or get left behind" scare tactic

It's worth being precise about what this data does and doesn't say. It doesn't say every business needs to become an "AI company" or that waiting six months will sink anyone. Plenty of businesses operate successfully at every stage of this spectrum, and the survey itself notes that not every business needs to become fully digitalized to succeed.

What it does say is something quieter and arguably more useful: the businesses getting the most value aren't doing something exotic — they're just further along on a path that's available to everyone, and the cost of staying where you are isn't zero. It's the difference between $1.40 and $1.70, or $1.70 and $2.40, applied to whatever you're already spending on technology and tools.

"Eventually" is a fine plan — for some things

If AI adoption feels like a someday project, that's understandable. Most "someday" projects are big, vague, and easy to defer because there's no clear starting point.

The reframe that tends to help: this doesn't have to be a project at all. It can be one specific thing — the missed call, the lead that didn't get followed up, the review that never got asked for — fixed once, and then it just runs. "Eventually" becomes "this month," because the scope is small enough to actually finish.