The SaaS-pocalypse Is a Repricing, Not an Extinction

Koko: A trillion dollars. Gone. Seven trading days. First week of February 2026. And here's the thing — the market priced the wrong death.

Sam: That is a spectacular opening sentence for a piece. Walk me into it.

Koko: So the story the market told itself was: AI agents are coming, they will absorb software workflows, small teams will replicate platforms that took years to build, per-seat pricing is toast, therefore software companies are toast.

Sam: Which sounds... plausible?

Koko: It sounds plausible right up until you look at the actual spend numbers. Global SaaS spending is still going up. From around three hundred and eighteen billion in 2025 to a projected five hundred and twelve billion by 2028. That is not an extinction curve.

Sam: That is a growth curve.

Koko: Exactly. So what is actually dying? Not software. The seat. The per-user license. That is what is going the way of the billable hour.

Sam: And the piece has a very specific, very good name for this moment — which I think sets up everything that follows.

Koko: Right, so the piece is called The SaaS-pocalypse Is a Repricing, Not an Extinction — and that title is doing real work. Forrester actually coined the SaaS-pocalypse label, they tied it to the February sell-off, and then their own data immediately complicates the doom narrative.

Sam: Which is a fun position to be in. You name the apocalypse and then your numbers suggest it is more of a... aggressive renovation.

Koko: A very expensive renovation. The core argument is this: the unit of pricing is collapsing. The seat was always a proxy for value. You paid per user because humans did the work and headcount was a reasonable meter for value consumed.

Sam: And then an agent does the work.

Koko: And the meter reads zero while the value keeps flowing. The piece puts it really sharply — a finance function running source-to-pay matching through agents at a third of the human seats is consuming more workflow value on fewer licenses.

Sam: So the vendor is getting paid less for delivering more. That is an uncomfortable place to run a software company from.

Koko: Which is why the repricing is arriving from both sides simultaneously. Vendors are moving toward flex credits, seat-plus-usage hybrids, consumption tiers. And buyers are already pushing back on spend.

Sam: The comparison to the billable hour is the one that really got me. Outcome fees replacing hours in professional services, seats being replaced in software — the piece calls it one repricing event running industry-wide. Same logic, same quarter.

Koko: Hours meter human effort. Seats meter human access. Agents break both meters at once. I think that is the cleanest sentence in the whole piece.

Sam: Okay, let's get into the data, because there are a few numbers in here that I think land differently when you say them out loud.

Koko: Go.

Sam: UBS, in June, surveying enterprise IT leaders — roughly sixty percent are now throttling AI spend. Token pooling, usage caps, model downgrading. And part of that AI bill is reportedly being funded by trimming SaaS and external IT budgets.

Koko: So it is not that AI is adding to the software budget. AI is eating part of the software budget.

Sam: Cannibalized, reallocated, and re-metered simultaneously — that is the exact phrasing in the piece. And then there is the single data point that I think should go on a poster somewhere.

Koko: The thirty-five thousand dollar user.

Sam: UBS turned up a single user running thirty-five thousand dollars a month in inference. One person. And the piece's point is: multiply that across an enterprise and you have a cloud-circa-2019 problem — capable of surprising you at scale before you have the governance to handle it.

Koko: The FinOps comparison is really useful there. Cloud spend got chaotic, then enterprises built an entire discipline around taming it. The argument is you need that same discipline for tokens and agent consumption — and you need it before the surprise, not after.

Sam: Now, I do want to push on the growth numbers for a second, because three hundred and eighteen billion to five hundred and twelve billion sounds like everything is fine. But the piece is also saying those dollars are going to flow differently — to different vendors, under different contracts.

Koko: Right, aggregate spend going up does not mean every vendor survives. Forrester specifically expects horizontal point-solution vendors — the ones that exist to bridge gaps between systems — to be consumed or collapse. And vertical, data-rich vendors to come out stronger.

Sam: So the total is growing but the distribution is being reshuffled pretty aggressively.

Koko: And the renewal calendar is where that reshuffling actually happens.

Sam: Here is where I want to push back a little, or at least stress-test the optimism. The piece says buyers hold unusual leverage right now — vendors caught mid-transition need reference customers for their new pricing models. And I buy that in principle.

Koko: But?

Sam: But the same vendors who are mid-transition are also the ones who have you on a renewal cycle they designed. The window is twelve to twenty-four months, the piece says. How many large enterprises can actually move that fast?

Koko: I think the piece is honest that it requires someone in the CFO seat who is paying attention to the contract mechanics, not just the technology story. The CIO gets the headline, the CFO has to do the actual work.

Sam: And there is one rule that I think the piece states almost as a non-negotiable. Do not sign a multi-year, seat-committed renewal in 2026 without an agent-conversion clause.

Koko: Because you would be locking in the old meter at the precise moment the meter is being replaced. That is not a small mistake — that is potentially years of overpaying for access that is no longer how the value flows.

Sam: My counterpoint is that agent-conversion clauses are not yet a standard thing vendors hand you. You have to ask for it. You have to know to ask for it.

Koko: Which is exactly why the piece argues this belongs to the CFO and not just the procurement team on autopilot. Someone in the room has to understand what they are actually negotiating.

Sam: Fair. And I think the rationalization angle is genuinely interesting — the idea that agent deployment is the audit that seat counts never triggered. You map which workflows agents absorb and suddenly you can see which point solutions only exist to fill gaps the agents now cover.

Koko: The average enterprise runs hundreds of applications that do not talk to each other. Agent deployment surfaces that redundancy in a way that periodic IT reviews somehow never managed to.

Sam: Probably because the periodic IT review did not have a commercial forcing function. This one does.

Koko: Okay, so what does a smart executive actually take away and do? Because I want to be concrete here.

Sam: Three things, the way I read the piece. First: do not confuse the softness in software valuations for a buying opportunity on seats. The seat itself is the problem. If a vendor is offering you a great deal on more seats, that is probably not a great deal.

Koko: Second: the negotiation window is real and it is time-limited. Forrester's guidance is renegotiate enterprise SaaS as agents deploy, consolidate to strategic vendors, separate true differentiators from utilities an agent can absorb. That is active work, not watching and waiting.

Sam: And third — and this is the one I think gets underweighted — get the FinOps function ready for software before the consumption billing surprises you. Cloud spend caught a lot of enterprises off-guard. Tokens and agent consumption are next in line. Build the governance now.

Koko: The piece frames the buy-side winners really precisely. Not the companies that slash software spend — the numbers say that option is not really available. The winners are the ones that re-contract around the new unit of value before those meters get standardized without them.

Sam: Workflows completed. Outcomes delivered. Agent-hours consumed. Those are the meters that matter going forward.

Koko: And if you wait for the vendor to price it first, you have already lost the negotiation.

Sam: Which is a slightly uncomfortable thing for a CFO to hear in a quarter where the instinct might be to hold spending and see what shakes out.

Koko: Right. Caution costs money here in a way it does not in a normal repricing cycle.

Sam: So the piece ends with a line I keep coming back to. The model was never the moat, and neither was the seat. Value moved to what you own and can meter honestly — the workflow, the outcome, the governed capability. Software is just the latest proxy to find that out.

Koko: Which is the whole argument in three sentences. And I think the reason The SaaS-pocalypse Is a Repricing, Not an Extinction lands is that it refuses to be either a doom piece or a reassurance piece. It is a mechanics piece. Here is what is actually changing, here is who it belongs to, here is the window.

Sam: The question I would leave listeners with is this: in your current SaaS estate, do you know which vendors are pricing by seat and which are already moving to consumption or outcome models? Because that map is the starting point for everything the piece is describing.

Koko: And most enterprises, if they are honest, do not have that map. Which is probably where the work starts.

Sam: The full piece — every number sourced, every claim grounded — is on Koko Knows. It is called The SaaS-pocalypse Is a Repricing, Not an Extinction. Worth reading slowly. Maybe with your head of procurement.

Koko: And maybe before your next renewal lands on your desk.