Three predictions, one shape.

A prediction went viral this month. Over the previous fourteen months, the SaaS giants posted drawdowns of 25 to 50 percent off their highs — not bad companies, the author was clear about that, and not companies doing anything wrong. The math just changed. Salesforce took seventeen years to climb from $1B to $9B in revenue. Anthropic covered the same ground in ten months. People don't want more software, the prediction said. They want workflows that solve problems.

Here was the prediction for 2026. Fractional executives with fifteen years of experience. Agencies that run programs end-to-end and own both the tech and the execution. Super agents — one intelligent layer that understands the business. Three things to bet on.

He saw the shape. The shape is right.

It is also one shape, not three.

He saw three predictions. They are three faces of one object — humans on top, agency in the middle, workforce underneath.

Each of the three is recognizable on its own. Fractional executives have existed for a decade. End-to-end agencies have existed for longer. AI tools, depending on what you count, are five years or fifty years old. None of the three is the news.

The news is what happens when you stack them. The fractional executive sets the goals. The agency installs the workforce under their brand and runs the program end-to-end. The workforce does the work. Three layers. One operating model. The reason it reads like three independent predictions is that the discourse hasn't yet named the thing they make together when stacked.

The phrase the discourse did land on — service-as-software — names the software as the protagonist. That's most of why the framing wobbles. The software is no longer the protagonist. The software has been demoted to substrate. The protagonist is the workforce, and the workforce is paid for outcomes, not for seats. The agency installs it. The fractional exec sets its goals. The workforce shows up and does the work and earns or doesn't earn the next quarter.

The phrase the discourse landed on still names the software as the protagonist. The protagonist is the workforce.

This is also why the giants' drawdowns are not a bad-companies story. They are not bad companies. They are selling at the wrong layer of the stack. The customer is buying outcomes; they are still selling tools to operate. A 50 percent drawdown is the market re-pricing the layer at which value gets captured.

Historically, the customer who wanted all three layers — strategy from a senior, execution from an agency, software to operate at scale — had to assemble them separately, pay for each, and integrate them in their own time. The math made it impossible at SMB scale and barely possible at mid-market. Putting all three on the same engagement made the cost stack break. The shift the prediction is describing is the math finally working — one engagement, three layers, the agency carrying all three under one brand without the unit economics blowing up.

The giants aren't doing anything wrong. They are selling at the wrong layer.

The customer doesn't see three layers. The customer sees one thing show up, configured for them, doing the work, paid for what worked.

The customer sees one thing show up, configured for them, doing the work, paid for what worked.

Three predictions, one shape. The companies that win 2026 won't have the most software, or the most agencies, or the most fractional executives. They will have all three composed onto the same install — and a workforce underneath that makes the math work.

That's also the prediction, finally named.