Per-Seat Pricing Is Dying and the Market Has Started Pricing the Funeral
Six go-to-market software companies lost $224B in 19 months. Every one ships AI, so AI is not the variable. Seat pricing is, and HubSpot proved it publicly.
Six go-to-market software companies lost about $224 billion in market value over 19 months, led by ZoomInfo down 76% and HubSpot down 71%. Every company involved ships AI, so AI adoption is not the distinguishing variable. Five of the six sell per-seat licences, a model that assumes headcount grows alongside delivered value.
Key takeaways
- Six GTM software companies lost roughly $224 billion in market value across 19 months, from ZoomInfo down 76% to Freshworks down 42%.
- Clay repriced up 900% from $500 million to $5 billion in the same window, and Sierra up 251% to $15.8 billion.
- Five of the six companies that fell sell seats, while Klaviyo prices on active profiles and usage and still fell 56%.
- HubSpot moved its Breeze customer and prospecting agents to outcome-based pricing on 14 April 2026.
- HubSpot stock closed down 19% the day after management walked through the pricing transition on its 7 May 2026 earnings call.
Reviewed and updated August 10, 2026
Per-Seat Pricing Is Dying and the Market Has Started Pricing the Funeral
Six go-to-market software companies lost about $224 billion of market value in 19 months. I went looking for the variable that explained it, and the answer was not the one everybody reaches for first.
Here is the board I built.
Repriced down
- ZoomInfo. $3.61B to $860M. Down 76%.
- Sprout Social. $1.76B to $455M. Down 74%.
- HubSpot. $35.97B to $10.49B. Down 71%.
- Salesforce. $320B to $134B. Down 58%.
- Klaviyo. $11.12B to $4.91B. Down 56%.
- Freshworks. $4.9B to $2.84B. Down 42%.
Repriced up
- Clay. $500M to $5B. Up 900%.
- Sierra. $4.5B to $15.8B. Up 251%.
- Glean. $4.6B to $7.2B. Up 57%.
- Zeta Global. $4.27B to $4.77B. Up 12%.
The test that rules out the obvious answer
The reflex explanation is that AI disrupted them. That explanation fails on its own evidence.
Every company on that list ships AI. ZoomInfo ships AI. HubSpot ships AI. Salesforce has shipped very little else for two years. So does Clay, so does Sierra, so does Glean. If AI adoption were the variable, it would not sort the list into two halves.
The variable is who captures the value AI creates.
Seat pricing assumes headcount grows alongside the value delivered. That assumption held for twenty years and it is the load-bearing beam under the entire modern software industry. Price per user, land a team, expand as the team grows, and your revenue tracks your customer's success automatically.
AI broke the assumption in the least convenient way possible. The work still gets done. The seat still does not get bought.
Five of the six companies that fell sell seats. The exception is Klaviyo, which prices by active profiles and usage.
The exception is the interesting part
Klaviyo still fell 56%, which means usage-based pricing is not armour. I want to be honest about that, because the neat version of this argument would quietly drop the inconvenient data point.
What the exception actually shows is that the pricing model is one variable among several. Growth deceleration, the rate environment, and multiple compression hit the whole software complex in this window, and no pricing model was immune to that. The pattern in the list is that seat-priced businesses were repriced harder and faster, not that usage-priced businesses were spared.
Correlation on a ten-company list is a hypothesis, not a proof. What turns it into something more than a hypothesis is that one of these companies ran the experiment in public.
HubSpot ran the experiment in public
On 14 April 2026 HubSpot moved its Breeze customer and prospecting agents to outcome-based pricing. Not a discount, not a bundle. A structural change to what the buyer pays for: the agent bills on a resolved conversation rather than on a licensed user.
After management walked through the transition on the 7 May earnings call, alongside mixed Q2 guidance, the stock closed down 19% the next day.
Read that sequence carefully, because it is the whole argument in miniature. A market leader publicly conceded that the seat is not the right unit, replaced it with an outcome, and the market marked the equity down on the transition. Wall Street priced the shift from seats to outcomes in the open, and it priced it as expensive.
That is what a repricing looks like from the inside. Not a collapse in demand for the software. A collapse in confidence about the unit the software is sold in.
What this means if you are on the buying side
You are on the other side of that trade, and the negotiating position is better than it has been in a decade. Five things I would put on the table at the next renewal.
Ask what happens to the price when your headcount falls. If you cut two SDRs and the invoice does not move, you are funding a model that has stopped tracking value. Get the true-down clause in writing, not the true-up.
Price the outcome where the vendor already offers one. Any vendor shipping agents has already built the metering to bill on resolutions, meetings, or enriched records. If they will not quote you that way, the reason is margin protection, not technical difficulty.
Audit seats against actual logins, then cut before you renew. In most stacks the gap between licensed and active is large enough to fund the replacement tooling on its own. The cold email pricing guide is a useful baseline for what the sending layer should actually cost.
Refuse multi-year lock-in on a per-seat basis. A three-year seat contract signed in 2026 is a bet that your team size in 2029 will look like your team size today. Nobody running a revenue team believes that right now.
Compare the seat cost to the outcome cost directly. A per-seat data stack and an outcome-priced service are comparable if you do the arithmetic. The breakdown in SDR cost versus an automation stack and the numbers in what a lead generation agency costs give you both sides of that comparison, and outsourced SDR versus in-house covers the third option.
What I am not claiming
I am not claiming per-seat software stops existing. Seats will survive wherever the seat genuinely is the value, which mostly means tools a specific human uses to do creative or judgment work.
I am not claiming these six companies are bad businesses. Several of them are extremely profitable and will be here long after this cycle.
The claim is narrower. The unit of pricing has stopped tracking the unit of value in go-to-market software specifically, because go-to-market work is the work AI absorbs first. The market noticed before most buyers did, which is the usual order of these things.
If your vendor still bills you per seat, you are funding a model the market is walking away from. Buy outcomes.
The same repricing logic explains why the most valuable AI companies founded after ChatGPT skew towards vendors selling finished work, why the GTM software valuation ranking sorts the way it does, and why the Clay, Apollo and ZoomInfo comparison keeps landing on cost per usable record rather than cost per user. If you are evaluating agents specifically, what an AI SDR actually is is the pricing question in disguise.
Frequently Asked Questions
Why did six go-to-market software companies lose $224 billion in 19 months?
Their revenue is priced per seat, and AI broke the assumption that headcount grows alongside delivered value. The work still gets done while the additional seat does not get bought, so the market repriced the growth expectations built into those multiples. Every company on the list ships AI, which rules out AI adoption as the distinguishing variable.
Does usage-based pricing protect a software company?
Not on its own. Klaviyo prices by active profiles and usage and still fell 56% in the same window, alongside broad multiple compression across software. The pattern is that seat-priced businesses were repriced harder and faster, not that usage-priced businesses were untouched.
What did HubSpot change about Breeze pricing?
On 14 April 2026 HubSpot moved its Breeze customer and prospecting agents to outcome-based pricing, billing on a resolved conversation rather than a licensed user. After management discussed the transition on the 7 May earnings call alongside mixed Q2 guidance, the stock closed down 19% the following day.
What should I renegotiate at my next software renewal?
Ask what happens to the invoice when your headcount falls, and get the true-down in writing. Audit licensed seats against actual logins before renewing. Refuse multi-year per-seat lock-in. Where the vendor already ships agents, ask to be quoted on resolutions or meetings, because the metering to do that already exists.
RevenueFlow builds AI-native pipeline systems and you pay per qualified meeting, not a retainer. No paying for activity. You only pay when we book you a qualified sales meeting. See if you qualify.
Market capitalisation figures reflect publicly reported values over the 19 months to mid-2026 and move daily. Not investment advice.
Frequently asked questions.
Frequently asked questions- Why did six go-to-market software companies lose $224 billion in 19 months?
- Their revenue is priced per seat, and AI broke the assumption that headcount grows alongside delivered value. The work still gets done while the additional seat does not get bought, so the market repriced the growth expectations built into those multiples. Every company on the list ships AI, which rules out AI adoption as the distinguishing variable.
- Does usage-based pricing protect a software company?
- Not on its own. Klaviyo prices by active profiles and usage and still fell 56% in the same window, alongside broad multiple compression across software. The pattern is that seat-priced businesses were repriced harder and faster, not that usage-priced businesses were spared entirely.
- What did HubSpot change about Breeze pricing?
- On 14 April 2026 HubSpot moved its Breeze customer and prospecting agents to outcome-based pricing, billing on a resolved conversation rather than a licensed user. After management discussed the transition on the 7 May earnings call alongside mixed Q2 guidance, the stock closed down 19% the following day.
- What should I renegotiate at my next software renewal?
- Ask what happens to the invoice when your headcount falls, and get the true-down in writing. Audit licensed seats against actual logins before renewing. Refuse multi-year per-seat lock-in. Where the vendor already ships agents, ask to be quoted on resolutions or meetings, because the metering to bill that way already exists.
About the author.
Ben Carden is CRO at RevenueFlow, which builds and operates outbound revenue engines for B2B companies. Previously at Gartner Enterprise. Studied at London School of Economics.
Ben Carden · CRO
Connect on LinkedIn →Explore more.
Ready to scale your outreach?
We build GTM engines that book real meetings. See the receipts.
Related articles.
ZoomInfo Renamed Its Ticker to GTM. Fourteen Months Later It Trades Under $1B.
ZoomInfo took the ticker GTM in May 2025. By late July 2026 it traded under $1B. Here is what the GTM software category is actually worth, with dates attached.
The Paris AI Map: 23 Companies, $35B, and 10 of Them Inside 2km
23 AI companies, more than $35B of reported value, and 10 of them inside one 2km circle. Here is the Paris AI map and why the density is the real asset.