Intelligence-First

Every so often the ground under business shifts, and a generation of companies that thought they were modern wake up to find they’re the incumbents about to be eaten. It has happened before, it’s happening again, and most of the companies it’s about to happen to can’t see it, because they’re standing on the last victory.

Here’s the pattern. A couple of decades ago there were traditional companies, and then there were digital-first companies. The digital-first ones didn’t just put the old business on a website. They rebuilt the business around the assumption that everything was digital. The tired example is the taxi: before Uber and Lyft there was a company almost nobody remembers called Sidecar, doing the same thing first. The point isn’t which logo won. The point is that “we have an app and a fleet” beat “we have a fleet and a dispatcher,” and a whole industry that thought it was fine got reorganized around a new default. The survivors were the ones who stopped treating digital as a channel and started treating it as the foundation.

We’re at the same kind of hinge again, and I’ve started calling the next default intelligence-first. Here’s the uncomfortable version of the thesis, the one that makes the room go quiet:

“You think you’re digital-first. You’re already outdated.”

Being digital was the last war. The companies that do the reorganizing this time will be the ones that build intelligence into how they actually operate, not the ones that bought a chatbot.

The froth is not the thing

Yes, there’s a bubble, and yes it will hurt some of the loudest names. A bust doesn’t un-invent a capability, though. The pioneers usually take the arrows and the well-capitalized second-movers buy the pieces cheap, the way it went with railroads, telecom, and the early internet, and the way it will probably go with a few of today’s front-runners. None of that changes the direction. When the froth clears, the companies still standing will be the ones that used this window to change how they work, not the ones that used it to raise money or issue a press release. So set the hype aside and look at the substance.

Intelligence-first is not a shopping trip

The most common mistake I see is a company deciding to “do AI” by buying tools. A law firm stands up an AI team out of its IT department, buys a research assistant and a drafting tool with nice interfaces, and calls it a strategy. But those exact tools will be available to every competitor within the year. Whatever edge they carry evaporates the moment everyone has the same thing, which is soon. Buying the generic tool is table stakes, not differentiation.

Intelligence-first isn’t something you buy. It’s something you rebuild, and it shows up in four places at once.

The first is operations: the intelligence gets woven into how the work actually gets done, not bolted onto the side as a feature nobody trusts.

The second is software. For thirty years, building custom software was so expensive and slow that most companies never seriously considered it, so they bent themselves around whatever they could buy off the shelf. That barrier just collapsed, and the first people the tools made cheap were developers like me. Now a company can build small, purpose-built tools that encode exactly how it wants to work, quickly and inexpensively. The work shifts to the parts that were always the point: deciding what to build, gathering the requirements, exercising judgment.

The third is teams. Every team, even the ones with no “creative” in the title, is now expected to think. The finance department stops being the place where numbers go to be filed and becomes a place that asks what else the data could say and what small tool would answer it. More people move up into judgment and away from drudgery, and the org chart quietly reorganizes around ideas and ownership.

The fourth, and the most important, is the moat. The single most undervalued asset in almost every company is the thing that never makes the balance sheet: how it actually operates. The accumulated, specific, hard-won way it does the work. Companies gesture at this vaguely (“our service is better than theirs”) and then never invest in it, never encode it, never sharpen it. In an intelligence-first world that internal way of working is the whole game, because it’s the one thing a competitor can’t buy off the same shelf you did. The winners will treat how they operate as a product they are constantly improving.

The obstacle nobody wants to name

If the direction is so clear, why is it a hard sell? Two reasons, and the second is the real one.

The first is that companies are genuinely bad at valuing their own operations. Efficiency gains are hard to quantify, so getting better at how you work gets filed under “soft,” and soft things don’t get funded. Worse, most companies don’t actually know how they operate today. They have a clean slide in a deck, and underneath it is a person quietly copying data out of one system into a spreadsheet because someone in accounting never got a seat in the real one. You can’t improve a process you’ve never honestly described, and most never have.

The second reason is the one people won’t say out loud: intelligence brings accountability, and a lot of leaders don’t want it. When you have clean data and accurate reporting, and the projections turn out to be right, the person who overrode them is suddenly, undeniably on the hook. Good numbers, looked at honestly and on a schedule, make hiding very hard. I’ve watched executives push back on technology that would help them the moment they realized it would also expose them. That resistance isn’t a technology problem. It’s a human one, and it’s the quiet reason a lot of these initiatives die in the boardroom while everyone blames the software.

The bet

I won’t pretend the timeline is tomorrow. Plenty of the companies I talk to aren’t ready to hear that the modern, digital operation they’re so proud of is already the past tense, and some of them will take years to come around. But the direction doesn’t bend. The companies that get genuinely intelligent about how they work, that encode their own way of operating and keep honing it, are going to move faster, see their own business more clearly, and quietly steamroll the ones still congratulating themselves for having an app.

Being digital was the last war, and most of the field is still fighting it. The next one goes to whoever builds the intelligence in. That’s the whole thesis, and nearly everything else I write here is a footnote to it.