The Net-New ARR Collapse: Why "Land-and-Expand" Quietly Broke in 2026 — and the GTM Rewrite Happening Inside the Operating Model

For every dollar of new ARR the median B2B SaaS company booked last year, it spent roughly two dollars getting there.

Read that line again, because it sounds like a typo until you check the source. The CAC ratio — sales and marketing dollars in, new annual recurring revenue out — has roughly doubled across the B2B software market in five years. It crossed 1.0 some time around 2022. It quietly crossed 2.0 in the back half of 2025. By the time CFOs sat down to do 2026 planning, the math that has underwritten every venture-backed SaaS company for two decades — spend a dollar of CAC, get more than a dollar back inside the year — was no longer true for the median operator.

Most boards still haven't said it out loud.

For Chief Executive Officers, Chief Financial Officers, Chief Revenue Officers, Chief Marketing Officers, Heads of Customer Success, Revenue Operations Leaders, and B2B Founders building 2026 plans against an operating model where the "land" half of land-and-expand has structurally broken, the net-new ARR collapse is the most quietly consequential GTM shift of the last decade. It is not a discount cycle. It is not a pipeline coverage problem. It is a recalibration of how SaaS revenue actually gets built — and the companies waking up to it are reorganizing their commercial motion around a different center of gravity than the one their 2022 plans assumed.

The Numbers That Should Be on Every Board Deck

The shift is hiding in plain sight inside three numbers that move in the same direction.

Customer acquisition costs are up 40% to 60% across B2B SaaS since 2023, according to industry benchmarks compiled in early 2026. The median new-logo sales cycle has stretched from 107 days in the first half of 2022 to 134 days in 2026 — a 25% expansion that quietly torches every per-rep productivity assumption built into the 2024 comp plan. And the average B2B buying committee, which used to involve three to five people, now involves eight to twelve internal stakeholders — with Forrester putting the full picture closer to 22 stakeholders when you count the nine external influencers (analysts, peers, board members, consultants) the buyer leans on before signing.

Those three numbers, multiplied together, produce the CAC ratio that's now sitting on the median P&L: roughly $2.00 of sales and marketing investment for every $1.00 of new ARR. That is not a healthy SaaS business. That is a company subsidizing every new logo with the expectation that customer lifetime value will eventually pay back the gap — except customer lifetime value is also getting harder to extract, because the next number on the list is the one nobody wants to talk about.

Seventy-five percent of software companies reported declining retention in 2024. Median net revenue retention has slid from a 2021 peak near 125% to roughly 107% by late 2024, and sub-$10M ARR operators are sitting at a median NRR of 98% — meaning the average sub-scale SaaS company is now losing money on its installed base before a single new logo is added to the top of the funnel.

Stack those four data points and the picture is unambiguous. The land motion costs more. It closes slower. The product of those two — new ARR per dollar of CAC — has compressed by roughly half. And the expansion motion that used to absorb the inefficiency through 120%+ NRR is no longer doing the absorbing.

The land has broken. The expand has weakened. The model in between — the SaaS operating model that powered fifteen years of multiples expansion — is being rebuilt from scratch inside the operating reviews most companies are running right now.

Three Things Broke at Once — That's Why It Looks Like a Crisis

The temptation, looking at the numbers, is to call this a single problem with a single fix. Pipeline coverage is off, so add SDRs. Sales cycle is long, so cut deal stages. CAC is up, so reduce paid spend. Each of those interventions has been tried in the field across 2025, and each has produced the same result: marginal improvement on the input, no meaningful improvement on the output.

The reason is that what looks like one problem is actually three structural shifts arriving simultaneously, and each one degrades a different lever of the GTM engine.

The buyer behavior shift. Sixty-seven percent of B2B buyers now say they prefer a rep-free experience when evaluating software, per Gartner's March 2026 sales survey. Eighty to ninety percent of the research happens before a sales conversation. Fifty-one percent of buyers now begin product research inside an AI chatbot rather than a search engine. The funnel still has a top, but the buyer is no longer walking into it through the front door. They are arriving in a sales conversation having already formed an opinion, already shortlisted, and already priced the alternatives — which means the sales motion that worked when reps shaped the early journey now arrives too late to shape anything.

The committee complexity shift. Doubling the number of stakeholders on a buying committee does not double the deal cycle. It roughly triples it. Each additional internal stakeholder introduces new objections, new approval steps, new political dynamics, and — in 2026 — new AI-augmented procurement reviews that compress vendor margin before the deal even reaches legal. The "consensus crisis" inside enterprise buying committees is now the single biggest driver of the 134-day sales cycle.

The pricing model shift. Sixty percent of B2B SaaS companies have moved away from pure per-seat subscription into consumption, hybrid, or outcome-based pricing. That move is the right one for value alignment, but it carries a brutal secondary effect on net-new ARR forecasting: a "land" deal in a consumption model often books a fraction of the value at signature compared to the equivalent seat-based deal three years ago. The revenue is now backloaded into usage that the customer has to grow into — which is great for NRR over time, and terrible for new-ARR run-rate measured at the close of any individual quarter.

Each of these three shifts independently degrades the land economics. Stacked together, they compound. And because the three shifts originated in three different parts of the operating model (the buyer, the deal, the contract), most companies are trying to solve them with three different teams — which is precisely why none of the point interventions are working.

The Quiet Pivot: 40% of New ARR Now Comes From Inside

Underneath the public conversation about pipeline coverage and SDR productivity, a much quieter restructure has been happening on the revenue side of the operating model.

Approximately 40% of new ARR booked by the average B2B SaaS company in 2026 now comes from existing customers — through expansion, upgrades, cross-sell, and increased consumption — up from roughly 25% two years ago. That is a 15-point swing in the source of new revenue inside 24 months. It is one of the largest internal mix shifts the industry has experienced since the move from perpetual licensing to subscription.

What that number is telling you, if you read it correctly, is that the median SaaS company has already stopped acquiring its growth from net-new logos. It just hasn't restructured its commercial org to match. The expansion motion is producing 40% of the revenue while still being staffed, comped, and reported as if it were a customer-success function rather than a revenue function. The land motion is producing 60% of new ARR while consuming the majority of sales and marketing spend — and producing it at a $2.00 CAC ratio.

The companies pulling ahead in 2026 are the ones that have noticed the gap and started reallocating.

What the Operating Model Looks Like When You Solve for This

The rewrite, where it's happening cleanly, has four moving parts. None of them is exotic. The exotic part is that all four are being changed at the same time, which is why the companies doing it carefully are seeing 15-to-20 point margin improvements on their GTM spend while the companies treating it as a series of point fixes are still missing plan.

Reallocate the rep mix. The most visible move is structural: take a meaningful slice of new-logo AE headcount and convert it into expansion AE headcount. The expansion rep covers a smaller book, owns a quota that is heavily weighted to multi-product attach and consumption growth, and works the installed base with the same rigor a new-logo AE used to bring to outbound. Companies that have made this move in 2025 are reporting expansion ARR per rep that runs two to three times what their new-logo AEs produce per quota dollar — which, once you do the math on CAC ratios, is the only operating move that mechanically restores GTM efficiency at the speed the market is demanding.

Re-price the marketing motion against committee reality. If the committee is now 22 stakeholders and the buyer is 80% through the journey before a rep engages, the role of marketing inverts. Demand generation as a "lead production function" was always a useful fiction. In 2026 it is actively misleading. Marketing's job in this operating model is to be present and credible across the entire buying committee — including the seven or eight non-buyer influencers — through brand, third-party validation, original research, customer advocacy, and (increasingly) presence inside the AI surfaces where 51% of research now starts. The MQL is not the artifact. The committee's collective trust posture is the artifact.

Treat the renewal as a forecast input, not a service event. When 40% of new ARR comes from inside, the renewal conversation is the most leveraged commercial event of the customer lifecycle. The companies doing this well in 2026 are running renewals with the same forecast discipline they apply to new-logo deals — pipeline reviews, MEDDIC scoring, executive sponsor calls, and a deal desk inside the customer success org. The renewal stops being a customer-success activity. It becomes a revenue activity that customer success co-owns.

Re-architect the data model so installed-base signal is treated as pipeline. In most companies the CRM was built around new-logo opportunities and the CSP (customer success platform) was built around health scores. The operating reality of 2026 is that the highest-quality pipeline most companies have access to is sitting inside their installed base — usage anomalies, integration depth changes, executive turnover, contract-window approach, ticket sentiment, expansion conversation history. Surfacing that signal as pipeline (with stages, owners, and forecast categories) is the data-layer change that makes the rest of the operating rewrite executable. Without it, the expansion org is flying blind on the same metrics CS used in 2022.

The "Land Reps vs. Expand Reps" Restructure

The single most counterintuitive move inside the rewrite is the one that breaks the most established reporting line: separating the land AE role from the expand AE role and comping them on different mechanics.

The argument against it has always been that the buyer relationship lives with the AE who signed the original deal. The argument for it, in 2026, is that the two motions are now genuinely different jobs.

A land rep operates in a 134-day cycle, against a 22-person committee, with most of the early journey invisible. They live in outbound, signal-based prospecting, multi-thread orchestration, executive alignment, and increasingly long procurement and legal reviews. The skill set is closer to consultative enterprise sales than the high-velocity inbound motion most SaaS companies built in the 2010s.

An expand rep operates in a 30-to-60-day cycle inside an existing relationship, with full data visibility, a champion already in place, and a usage curve that telegraphs intent weeks before the conversation happens. The skill set is closer to strategic account management with a quota — and the per-rep productivity, when the role is staffed and resourced correctly, is structurally higher than the land equivalent.

When companies try to make one role do both jobs, the rep optimizes for whichever is easier to hit at quota time. In 2026, that's almost always expansion — which means the land function quietly atrophies, and the company ends up with the worst version of the new operating model: under-invested net-new, over-leaned-on expansion, and a forecast that depends on a finite installed base to keep producing.

The separation, where it's happening, is being done deliberately. Different quotas. Different comp accelerators. Different pipeline reviews. Different forecast lines. Different leaders. A few public examples have started showing up in earnings calls — companies announcing the appointment of a Chief Customer Revenue Officer or a VP of Installed Base Growth as a peer to the Chief Sales Officer rather than a report inside customer success. That title change is the surface signal of a much deeper org-chart rewrite.

The CFO View: Re-Pricing the GTM Engine

Sitting inside the finance org, the net-new ARR collapse looks different than it does inside sales.

The CFO's question is not "how do we close more new logos." The CFO's question is "what is the marginal return on the next dollar of GTM spend, and where in the operating model should it be spent?" In 2024 that question still had "new logo acquisition" as the dominant answer for most operators. In 2026 it almost never does.

The 2026 framework most efficient operators are converging on looks roughly like this. Best-in-class CAC payback under 12 months. Median acceptable CAC payback by segment: 8–12 months for SMB, 14–18 months for mid-market, 18–24 months for enterprise. Burn multiple under 1.5x. Gross margins above 75%. Rule of 40 firmly above the line, with the quality of the 40 — the mix between growth and profitability — increasingly scrutinized by both public-market analysts and private investors. Companies that consistently beat the Rule of 40 are commanding roughly 2x the revenue multiple of less-efficient peers, and the spread is widening, not narrowing.

What that framework does, when applied honestly to most B2B SaaS P&Ls, is force the reallocation. Every dollar of CAC has to be benchmarked against the marginal expansion dollar — and at current CAC ratios, the expansion dollar wins on almost every comparison. Once that math becomes visible inside the operating plan, the rep mix, marketing budget, and org chart move with it. The CFO is rarely the loudest voice in the room when the rewrite begins, but they are almost always the proximate cause.

Five Questions Every Revenue Org Should Answer Before Q4 Planning

If the rewrite is the right read of the market — and the underlying data argues that it is — then there are five questions a revenue org should put on the table before the 2027 planning cycle begins.

• What is our actual CAC ratio for net-new ARR right now, broken out by segment and channel? If the consolidated number is above 1.5, the rest of the conversation is mostly noise until the underlying segment mix is understood.
• What percentage of our new ARR in the last four quarters came from existing customers — and how is that mix trending? If the number is 30%+ and rising, the operating model is already shifting whether or not the org chart has caught up.
• Do we have a separate expansion AE role with its own quota, comp plan, pipeline, and forecast — or is expansion still being treated as a customer success activity that occasionally produces revenue? This is the single highest-leverage operating decision in front of most revenue leaders.
• Where does signal from the installed base currently live in our data stack, and is it visible to revenue leadership as pipeline? If installed-base intent signals (usage, integration depth, executive activity, sentiment, contract proximity) are not surfaced in the forecast review, the expansion motion is being run blind.
• Are we resourcing marketing against the 22-person buying committee, or are we still optimizing for MQL volume? The committee math has changed. Most marketing budgets haven't.

The companies that walk into 2027 with clean answers to those five questions will be the companies operating on the new model. The companies that don't will be the ones explaining to their boards, again, why CAC is up, cycles are long, and the forecast missed.

The 2026 Picture

The land-and-expand operating model isn't dead. It has been rebalanced — by buyer behavior, by committee complexity, by pricing-model evolution, and by the unforgiving arithmetic of a $2.00 CAC ratio. The companies that read 2025 as a temporary headwind and are planning for a return to the 2021 environment are mispricing the next 24 months. The companies that read it as a structural shift and are rebuilding their commercial motion around a 40-percent-from-inside, expansion-led, signal-driven, committee-aware operating model are the ones that will set the benchmark for what efficient B2B SaaS growth looks like for the rest of the decade.

The land has broken. The expand is being rebuilt around it. The companies that move first will be the ones writing the playbook the rest of the market eventually copies.