The way we sell technology
is breaking. Quietly.

Most technology and services today are still bought the way they were bought ten years ago. Per seat. Per user. Per hour. And more recently, per unit of consumption.

And for a long time, that has made sense. Usage has been a fair proxy for value. If more of your people log in every day, you are probably getting more out of the product. If the managed service consumes more hours, more work is getting done. The whole engine of the SaaS era, land the deal, drive adoption, expand the seats, renew the contract, was built on one quiet assumption: usage and value move together.

That assumption is breaking. Futurum's 2026 survey of enterprise software buyers found fewer than one in five still prefer classic per-user pricing, but it's still how the vast majority of software is contracted for. Preference has moved. Contracting hasn't.

Why the old model held for so long

Billing for time, activities, and access wasn't lazy. It was rational. Measuring a customer's actual business outcome is hard, slow, and contestable. So both sides settled for proxies everyone could count: seats, hours, tickets, uptime. Procurement liked it. Finance liked it. It worked.

AI is doing two things to this model at the same time, and they compound.

First, it severs usage from value. When AI does the work, the best result for the customer is less usage. Fewer seats. Fewer interactions. Fewer tickets, because the issues got prevented rather than resolved. A vendor paid per seat now makes less money when the customer succeeds. Run that forward and the model is going in reverse. So what are vendors doing right now? Adding AI features they can charge for, to make up the delta from the reduced usage. That buys time. It doesn't fix the model.

Second, and this is the deeper shift, AI removes the measurement excuse. Instrumented workflows, connected data, and agents operating inside the customer's business processes make the actual outcome visible in ways it never was. Billing for time and activities was acceptable when it was hard to measure outcomes. That excuse is going away.

This movie has run before

The technology industry likes to believe its problems are new. This one isn't.

I spent fifteen years selling and delivering BPO services. Charging by the FTE and by the hour. And I watched that industry get force-marched through this exact transition, two decades ahead of tech. Some clients got tired of paying for effort. They demanded results. Per-transaction pricing came first, then gain-share, then genuine outcome contracts with agreed baselines and auditable governance. I've written about one of those deals, and one we walked away from.

Not because BPO providers were visionaries. Because their model commoditized and buyers gave them no choice.

And here's the part worth sitting with. Twenty years on, most BPO revenue is still priced by the FTE and the seat. The industry that got force-marched first still hasn't finished the walk. What happened instead is that outcome deals became a growing part of the mix, and the industry got good at the ones it did.

The lessons from that forced march, the outcome deals priced too low, the baselines nobody agreed upfront, the attribution fights, are sitting there waiting for anyone in tech willing to go look. It's also one of the reasons those same BPO companies are now moving fast to eat their own lunch, bundling AI that completes tasks and resolves issues into contracts that are truly outcome-based.

The natives aren't waiting

While incumbents debate, a new generation of AI-native companies has built the shift directly into how they sell. They prove value on the customer's own data before a contract is signed. They walk away from deals where the value case doesn't hold. They sell results, not access.

Customers who have experienced that motion once bring the expectation into every other vendor conversation. Why would I pay you to find out if this works, when the last company showed me before I spent a cent?

Incumbents can't copy this by renaming their discovery phase. It's a different economic model, with the cost of proof sitting on the other side of the contract. That's a topic I'll come back to later.

It still moves at the buyer's pace

None of this happens because vendors decide it should. It happens as fast as buyers are ready to move, and most aren't yet. Procurement is built to compare unit prices. Finance wants a number it can forecast. The people who would have to agree a baseline often don't own the data to set one.

What the AI-natives are doing is shortening that wait. They are not persuading the market one deal at a time, they are moving what a buyer thinks is normal. That is a catalyst, and it will pull this forward faster than it would have gone.

But it won't flip. You can be entirely ready to sell an outcome and still spend two years waiting for a customer who can buy one. That is not a reason to wait. It is a reason to be ready early, because readiness is the thing you can't assemble in the six weeks after a buyer finally asks.

There's one more thing, and it doesn't go on a readiness checklist. An outcome contract asks both sides to agree a baseline, agree attribution, and then not fight about either for three years. That takes a level of trust most vendor relationships don't have. Every outcome deal I've seen work was built on years of relationship before anyone proposed one.

The AI-natives don't have those years. What they have instead is visibility. When value is measurable from day one, on the customer's own data, there is much less to take on faith. That's the part worth stealing. Trust built on relationship takes years. Trust built on evidence takes weeks, and it doesn't leave when your champion does.

Which leads me to the word itself

So now "outcomes" is on every slide, in every pitch, in every pricing conversation. And that's exactly the moment a word is most at risk of meaning nothing.

Some of what's being sold as outcome-based genuinely is. A lot of it is output pricing in new clothing, or SLA contracts wearing a badge they didn't earn. The customers being asked to sign these deals deserve a way to tell the difference. So do the providers building them in good faith.

Which raises the question this whole series hangs on. What actually counts as an outcome, and what doesn't?

That's the next piece.

Martin Dove is the founder of It's the Outcome, and has spent the last decade helping technology and services companies make the shift to outcome-based client engagement.