Invest1 distinct publisher2 min readPublished
The Vertice index puts software inflation at 12 to 16.4% against 2.7% in the G7, and four years of seat-price rises have left vendors choosing among three successor models.
The Investor · Invest desk
Compiled by The InvestorSomething wrong?How this is made
Software is inflating at roughly six times the broader rate [17], and a buyer will swallow one year of that; four in a row is the problem. Cost per employee climbed from about $7,900 in 2023 to roughly $9,100 by the end of 2025, a rise of about 15% [6][18], and per Zylo's data 79% of IT leaders hit a price increase at their last renewal, 78% saw unexpected AI or consumption charges, and 61% cut planned projects to absorb the overrun [8]. None of that extra spend bought more software; it came from price, packaging and tier moves inside contracts that already existed [c7b], on top of four straight years in which seat pricing carried the revenue burden and used up its own headroom [3].
The deeper break is in the unit. Seat pricing charges for headcount, which tracked how much work was getting done well enough for twenty years [15]. That link is coming apart. A support team clearing three times the volume with the same forty people pays the same invoice; a team that falls to twenty-five because agents took the L1 queue pays less for delivering more [15]. The vendor that built the automation shrank its own bill.
That is why the fifth annual increase lands differently from the first, and why CIOs are now arriving at renewals asking for cuts rather than defending a number [9]. A model whose bill moves in the wrong direction as the customer gets more done cannot survive that conversation, and the successor has to price something other than headcount.
Ranked by verification strength, evidence, and original report placement.
Seat pricing has carried the entire revenue burden of B2B for four years, vendors leaned on it hard, and those repeated increases used up all the headroom.
After four years of increases the fifth ask lands very differently, and CIOs are now uniformly asking for cuts at renewals.
Publicis Sapient has publicly said it is cutting its traditional SaaS licenses by roughly half, Adobe included, and substituting AI tools; the author expects Global 2000 procurement teams to build the same deck.
Seat pricing prices headcount, a good proxy for how much work was being done for twenty years, and that link is breaking: a support team handling 3x the volume with the same 40 people is billed the same, and a team that shrinks from 40 to 25 because agents do the L1 work is billed less for delivering more, so the vendor shrank its own invoice.
The author's framing is that if you sell per seat you are the funding source, and your renewal is now competing with a token bill.
Henry Schuck, CEO of ZoomInfo, says that after talking to top consultants and hundreds of his own customers, nobody, expert or customer, knows where B2B pricing is going; the author adds that three distinct successor models have emerged, unequal in viability and in how hard they are to run.
Follow any of these and your For You feed starts watching them — no settings page required.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Named third-party datasets, one publisher, no independent verification
The cost side is unusually well quantified for a single article: two named indices with disclosed scale (Vertice's tens of billions in processed spend; Zylo's 40M+ licenses and $75B+ spend), a dated 141-CIO survey, and three additional named research houses on AI budget funding. But everything reaches the reader through one publisher's summary, none of the underlying reports are reproduced or independently checked in the cluster, and the central editorial conclusion that 'just seats' is finished is authorial synthesis plus survey expectations rather than measured vendor repricing.
Buyer-side behavior documented; vendor repricing still mostly intent
Adoption is asymmetric. Buyer behavior is observed and quantified: 79% of IT leaders took a renewal price increase, 78% saw unexpected AI or consumption charges, 61% cancelled planned projects, 54% run consolidation programs, and one named Fortune-scale buyer (Publicis Sapient) has committed to halving traditional SaaS licenses. The migration of vendors to consumption or outcome pricing, which is the story's thesis, is evidenced only by CIO expectations (46% expect usage or outcome pricing to spread, 29% expect seats to decline) and by the observation that many vendors shipped consumption pricing without spend controls; no count of converted vendors or realized revenue mix shift appears.
Cost data solid; 'seat model is dying' runs ahead of it
The pricing-pressure evidence is real and specific, but the framing overshoots it in three ways: total SaaS spend is still growing 8% on flat portfolios, which is the opposite of a collapsing model; the article itself concedes seats are not going to zero and that successor models keep a platform access fee; and the strongest structural claim (per-seat vendors are now the funding source, so their model is finished) rests on illustrative scenarios plus CIO expectations rather than observed vendor revenue-model change. The acknowledged RBC counter-finding is also given a single sentence against the reallocation thesis, and the Global 2000 replication of Publicis Sapient's cut is prediction, not observation.
Vendor-facing outlet relaying data from cost-management and CEO sources
Every quantitative input carries a commercial interest that points the same way. Vertice and Zylo sell SaaS spend management and negotiation services, so a high software-inflation reading and widespread unexpected AI charges support their market narrative; Redpoint is a venture investor publishing CIO research that frames a pricing-model transition; the framing prompt comes from a vendor CEO (ZoomInfo's Henry Schuck) whose company sells data lookups, a naturally meterable unit. SaaStr is a vendor-facing media and community business whose audience buys pricing-strategy guidance, and the piece repeatedly cites its own prior coverage. None of this is disclosed as an interest in the article, though the sources are named.
Reliable on buyer cost, weaker on vendor transition
Confidence is moderate. The quantitative buyer-cost picture is internally consistent, dated, and attributed to named datasets with disclosed scale, and the article surfaces its own contradictory source, which raises trust. It is lowered by the single-publisher cluster, the absence of primary reports or independent corroboration, the commercial alignment of nearly all cited data providers, and the fact that the load-bearing conclusion about vendor pricing models is argument and expectation rather than observed change.
invest
BigCommerce turned down $1.5B for second place in ecommerce. It is now worth about $200M.1 distinct publisher
product
Nebius funds $4.5bn of AI capacity on terms that pay lenders mostly in stock2 distinct publishers
build
SMIC's first $3 billion quarter comes with a wafer price increase attached1 distinct publisher
build
Your .ai viewer is a pdf.js problem, and its worst bugs never throw1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 25, 2026