Essay

When the go-to-market playbook stops working

May 2026 . 8 min read

The earliest signal that a software company is hitting its growth ceiling is not in the product. It is in the go-to-market motion. Win rates compress, cycles lengthen, expansion stalls, and the inbound machine quietly stops returning what it used to. The dashboards smooth the signal because the dashboards were designed for a different era. By the time the board sees the break in the headline numbers, the operating model has been wrong for twelve months. The decision-bearing question is not whether the go-to-market playbook needs tuning. It is whether the playbook itself has stopped working, and whether the company is in the first quarters of the second growth curve.

The playbook the first growth curve rewarded

Classic B2B SaaS rewarded a specific motion. A repeatable inbound funnel feeding an outbound layer feeding a controlled sales process. Two to three economic buyers per deal. A discovery-to-close cycle measured in 30 to 60 days for mid-market and 90 to 120 days for enterprise. Net retention above 110 percent. Win rates in the high twenties to low thirties. Payback in 18 months. That motion compounded because each input scaled cleanly: more reps, more pipeline, more bookings. The forecast model was linear, and the board pattern-matched on linearity.

That is the motion most of the public SaaS index was built on. It is also the motion that 2026 data shows is no longer producing the same outputs.

Five early signals, all read late

The break shows up first in five places. None of them is loud enough to trigger a board response in isolation. Together, they are unambiguous.

One. The sales cycle elongates. The median B2B SaaS sales cycle has reached 84 days, up 22 percent since 2022 (Prospeo, 2026). The damage is heaviest at the top of the deal range: median cycles now run 90 days plus for deals above 100,000 dollars ACV and 170 days plus for enterprise contracts above 250,000 dollars (Growthspree, 2026). Most CROs report cycle elongation as a discounting and pipeline-coverage problem. It is actually a buyer-behaviour problem and a packaging problem.

Two. Win rates compress. The SaaS and technology median win rate now sits at 22 percent (Martal, 2026). The first-curve dashboard was calibrated to 25 to 30 percent. Three to eight points of win-rate compression, sustained, halves bookings efficiency. It rarely registers as a structural break because no single quarter looks alarming.

Three. Buying committees expand. The average B2B buying group has grown from 5.4 in 2020 to 6.8 in 2024, and Forrester and 6sense data put the median committee for deals above 50,000 dollars at 11.2 stakeholders, up from 9.7 in 2024 (Corporate Visions, 2026; Martal, 2026). Each additional stakeholder is more internal selling that the champion has to do without the rep in the room. Sales productivity drops invisibly: same activity, lower yield.

Four. Procurement load increases. SOC 2, GDPR, AI governance, and vendor risk reviews now add two to four weeks to mid-market and enterprise cycles, and the reviews are getting longer rather than shorter (Corporate Visions, 2026). What used to be a closing tax has become a stage of the cycle. Champion fatigue is the most underweighted forecast risk in software today.

Five. Expansion stalls. The expansion line was always the most leveraged input to net retention. AI-era buyers are tracking realised usage and resolved outcomes rather than seat counts, and they are repricing accordingly. Per-seat pricing collapsed from 21 percent of SaaS to 15 in the twelve months to early 2026, while hybrid models reached 41 percent adoption (Monetizely, 2026). Any company whose expansion motion still depends on seat creep is selling a unit of value the buyer has stopped paying for.

Why the dashboard reads the break late

The instrumentation behind the first-curve playbook smooths the signal in three ways. Pipeline coverage masks win-rate compression because more pipeline arrives to compensate. Bookings totals mask cycle elongation because deals close eventually. Net retention masks expansion deceleration because reported NRR is a trailing twelve-month figure that holds up well even as new-period expansion thins. By the time the headline numbers move, three to four quarters of operating reality has already changed.

That delay is the explanation for the gap between the operating teams that know the playbook has broken and the boards that are still waiting for evidence.

The evidence was always there. It was sitting one layer below the dashboard.

What changed in the market, not the company

The cause is not internal execution. It is buyer behaviour, and buyer behaviour has changed because the underlying software value chain has changed. Three structural shifts compound.

First, AI agents have shifted what buyers are willing to pay for. They are paying for resolved tasks, not access. Eighty-three percent of AI-native SaaS now bills on usage rather than seats, and Zendesk and Intercom have moved to outcome-based pricing for AI-driven resolutions (Stormy AI, 2026). A vendor priced on seats is selling a unit of value the market has demoted.

Second, the AI investment cycle inside customer organisations has absorbed buying attention. McKinsey reports 5.8x average ROI on production AI within 14 months, but only 29 percent of executives see significant returns, and 79 percent of organisations report adoption challenges (McKinsey, 2025; Writer, 2026). That gap means more internal selling, more steering-committee oversight, and longer purchase decisions for everything that is not the AI bet itself.

Third, gross margin has reset. ICONIQ data puts the average AI product gross margin at 52 percent against the SaaS benchmark of 80, and inference is consuming roughly 23 percent of revenue at scaling-stage AI B2B companies (Monetizely, 2026). That changes what every customer expects to pay and what every competitor can sustainably underwrite. The pricing pressure does not arrive evenly. It arrives at renewal.

The lazy reading and why it fails

The comfortable diagnosis inside many software exec teams is that the go-to-market team needs better enablement, sharper messaging, and more pipeline. That is the first-curve answer to a second-curve problem, and it makes the situation worse. Adding pipeline at a 22 percent win rate consumes capital without changing the underlying yield. Sharper messaging cannot move a buyer who is now buying a different unit of value. The classic SaaS playbook was rational under classic SaaS economics. It is no longer enough.

A sharper version: SaaS playbook isn’t enough. The motion that compounded the first curve does not compound the second. The seat-based pricing, the access-led expansion, the discovery-driven cycle, the linear forecast were artefacts of a margin profile and a buyer expectation that no longer hold.

The diagnostic. Five questions for the next operating review

Run the diagnostic at the next operating review. Pull twelve quarters of data. Read the answers honestly.

One. What has happened to the median sales cycle, by deal band, over twelve quarters. If it is up by more than ten percent, the issue is structural, not execution.

Two. What is the year-over-year change in win rate by deal band. If win rates have compressed by three points or more without a recognised product change, the buyer is buying something the product is not yet pricing.

Three. What is the average committee size and the average procurement duration. If both have grown, the go-to-market motion needs more champion enablement, not more rep enablement.

Four. What is the new-period expansion run rate, distinct from trailing twelve-month NRR. Trailing NRR is comforting. New-period expansion is the real signal.

Five. What share of new ARR is outcome-priced or usage-priced rather than seat-priced. If it is under 20 percent, the pricing model is exposing the company at every renewal.

The board agenda for the second growth curve

Three choices follow.

Stop. Stop calibrating sales capacity off first-curve win rates. Stop reporting NRR without the new-period expansion line. Stop treating cycle elongation as a discounting problem.

Test. Test outcome-based pricing on the segment with the highest measured AI usage. Test a champion-led buying motion on enterprise. Test a renewal pricing model that meters resolved work.

Redesign. Redesign the forecast around outcomes per account, not seats per account. Redesign the comp plan to reward expansion that survives the next renewal. Redesign the operating review around buyer-behaviour signals: cycle length, committee size, procurement duration, outcome-priced share.

The board question that should replace “how is the pipeline” is simpler and harder: what is the yield of our go-to-market motion, and which line of the operating model do we need to redesign first.

Sources

go-to-market, SaaS, pricing, boards, AI-native

Back to writing