Software Valuations in Early 2026: Discipline, Darwinism, and M&A Readiness in the Age of the "AI Apocalypse"
Rob Bartlett's latest Jefferies monthly software valuation data reveals a market that isn't risk-on—it's ruthlessly selective. Here's what that means for founders, boards, and deal readiness.
By Louis Lehot
Key Takeaways
AI Darlings have surged 513% since ChatGPT’s launch; horizontal application software has fallen 14% over the same period—the widest performance gap in public software history.
Revenue growth is valued at 2.8× the importance of profitability in the current two-factor regression, but the market now demands that growth be credible, efficient, and repeatable.
Application software multiples have contracted 41% in the last twelve months (from 5.8× to 3.4× EV/NTM Revenue), now trading well below pre-pandemic averages of 7.8×.
~90% of public software companies now grow below 20%, compared to just 56% at the November 2021 peak—only two companies (Broadcom and Palantir) exceed 30% NTM growth.
The Jefferies GRAF metric has compressed 61% since 2021 (from 0.52× to 0.20×), meaning every unit of growth-adjusted performance buys far less valuation than it did four years ago.
M&A readiness is now valuation defense, not hygiene. Buyers are underwriting AI disruption risk, renewal quality, and “real” ARR—start preparing 12–24 months before you plan to sell.
Every market cycle has its tell. In 2021 it was “growth at any cost.” In 2023 it was “survive the rate shock.” In early 2026, the tell is judgment—the capital-allocators-with-spreadsheets kind. The latest Jefferies Software Valuation Update (March 2026), published by Rob Bartlett and the Jefferies Technology Investment Banking team, is a crisp reminder that public markets haven’t simply turned “risk-on.” They’ve turned selective.
I follow Rob’s monthly note closely because it’s one of the best external benchmarks available to the founders and boards I advise as outside counsel. In his March cover email, Rob frames what many operators felt in their bones: a continued sell-off and compression across verticals, growth tiers, and profitability buckets, followed by a sharp February rotation back toward growth. As Rob’s data shows, the market’s implied tradeoff now sits at roughly 2.8 points of growth for every 1 point of profitability. For the companies and boards I work with, that isn’t just market commentary; it’s a real-time scoring rubric that should inform your board deck, your capital allocation decisions, and—critically—your deal readiness posture.
1. “Software” Isn’t a Category Anymore—It’s a Sorting Mechanism
Rob’s performance data makes the point bluntly: the market is no longer paying for the word “SaaS.” It’s paying for business model durability, credible growth, and margin structure. Mega-cap and AI-exposed beneficiaries have held up materially better than broad horizontal application software, which faces sustained scrutiny.
The dispersion is staggering. Diversified mega-cap software tracked the NASDAQ Composite at +21%, while application software fell 28% and vertical software dropped 34%. If you're building in enterprise SaaS and you feel like you're being punished for someone else's sins—or hype—welcome to 2026.
The divergence since ChatGPT’s November 2022 launch tells the whole story. Rob’s data shows AI darlings are up over 500% while horizontal application software is now below where it started. The market has made its bet on which side of the AI divide matters.
2. The Rule of 40 Is Back—Except It Never Really Left
Rob’s work underscores that valuations remain anchored to Rule-of-40 thinking and a weighted variant of it—Jefferies’ GRAF framework (Growth-Adjusted “Rule of Forty”). The key regression finding from the Jefferies team: revenue growth still drives valuation far more than profitability, but the market now demands that growth be believable, efficient, and repeatable.
The two-factor regression explains 47% of the variation in software valuations—a meaningful improvement over revenue growth alone (39%) or the simple Rule of 40 (40%). The weighted Rule of 40 has remained the strongest single predictor of software multiples for over a year. Founders who internalize this framework will make better capital allocation decisions than those chasing vanity metrics.
3. Multiple Compression Is Deep and Broad
The scale of multiple compression across the software landscape—as Rob’s data illustrates—is striking. Every vertical and every growth bucket has contracted over the last twelve months, and most are now trading well below their five-year averages and even their pre-pandemic levels.
Two data points stand out. First, application software at 3.4× is now trading at less than half its pre-pandemic average of 7.8×. Second, the 30%+ growth bucket trades at 28.1×—but that “bucket” contains exactly two companies: Broadcom and Palantir. The rest of the software universe is in a very different valuation regime.
4. The Growth Distribution Has Fundamentally Shifted
One of the most underappreciated charts in Rob’s report shows how the composition of the software universe has changed. At the November 2021 peak, 44% of public software companies were growing above 20%. Today that number is just 13%.
This isn’t just a multiple problem—it’s a growth problem. Nearly half the public software universe is now expected to grow below 10%. The profitability picture, however, has improved materially: 60% of software companies now generate 20%+ FCF margins, up from just 34% in late 2021. The industry has structurally re-profiled toward efficiency.
5. The “AI Apocalypse” Narrative Is Real—Because Incumbents Are Writing It
Here’s the part that keeps surfacing in board conversations I’m part of: fear that AI will vaporize enterprise SaaS moats, compress pricing, and turn yesterday’s category winners into tomorrow’s “nice feature.” Rob’s data doesn’t say “apocalypse,” but it shows a market that rewards clear AI leverage while punishing undifferentiated horizontal software.
The practical implication for my clients: Your product roadmap is now a diligence item. Buyers and public investors are asking the same question: is AI a tailwind to your unit economics, or a headwind to your differentiation? If you don’t have a crisp answer, the market will supply one—usually in the form of a lower multiple—and a buyer’s counsel will use it as leverage at the negotiating table.
Rob’s data makes the scoring visible. Companies in the “AI Darlings” index (AMZN, AVGO, GOOGL, META, MSFT, NVDA, ORCL, PLTR) are up 30% over the last twelve months. The “AI Beneficiaries” basket (ANET, DDOG, DELL, HPE, NOW, NTAP, PSTG) is up 8%. Horizontal application software? Down 36%. The market has rendered its verdict on who benefits from AI and who gets disrupted by it—and as I tell the boards I advise, those same signals are now showing up in M&A diligence.
6. M&A Readiness Is No Longer “Good Hygiene”—It’s Valuation Defense
This is where Rob’s valuation benchmarking connects to the work I do every day as a lawyer advising on M&A transactions. In recent work with Vitaly Golomb, we’ve made the point that M&A is not a lottery ticket—it’s a brutal, high-stakes process that rewards the prepared and punishes the complacent. Our advice is deliberately direct: start 12–24 months before you plan to sell. Clean up the cap table. Fix IP assignments. Upgrade your financial reporting. Build a killer data room. Get your house in legal order before the buyer’s counsel starts pulling at threads.
In March 2026, that readiness isn’t just about speed—it’s about surviving harder diligence in a market that’s skeptical of enterprise SaaS narratives. From my seat at the negotiating table, I see buyers underwriting AI disruption risk, renewal quality, customer concentration, security posture, and “real” ARR (not vibes). The legal and structural diligence has intensified in direct proportion to the multiple compression Rob’s data reveals.
The companies that win outcomes will be the ones that can answer, quickly and credibly:
What’s our durable wedge? What moat survives the AI disruption cycle?
What is AI doing to our margins and retention—today, not someday? Concrete metrics, not roadmap slides.
Where are the landmines? Contracts, IP ownership, data rights, customer concentration.
What to Expect Through Mid-2026
Based on the trends in Rob’s data, I expect continued bifurcation: premium outcomes for category leaders and AI-levered platforms, and tougher price discovery for everyone else—especially enterprise SaaS businesses that can’t demonstrate differentiated AI positioning alongside disciplined economics.
Translation for founders: treat readiness as an operating principle, not a transaction checklist. For board members: insist on a readiness cadence now—because the market is already scoring you, and buyers’ legal and financial diligence will reflect these compressed multiples in the terms they offer.
And if you want a clean, consistent external benchmark to keep you honest, Rob Bartlett’s monthly Jefferies note is one of the better “reality anchors” making the rounds in boardrooms and on deal teams today. I recommend it to every client.
Disclosure: All market data and charts referenced in this post are sourced from the Jefferies Monthly Software Market Valuation and Performance Update, March 2026, published by Rob Bartlett and the Jefferies Technology Investment Banking team (Capital IQ data as of 2/27/2026). Infographics were created by the author based on Jefferies’ published data. This post reflects the author’s perspective as a legal advisor and does not constitute investment advice or banking analysis.
About the Author: Louis Lehot is a corporate and securities lawyer who advises founders, boards, and investors on M&A transactions, capital markets, and corporate governance in the technology sector.