Written for founders, CEOs, CFOs, and directors who actually have to decide something.
By Louis Lehot, Silicon Valley M&A lawyer
KEY TAKEAWAYS
Three public software companies are growing 30%+: Broadcom, Oracle, and Palantir. Their average valuation multiple is 21.0x revenue, while companies growing below 10% are trading at just 3.1x revenue.
Nearly 90% of public software companies are now growing below 20%, compared with 56% in November 2021. The high-growth category wasn’t repriced. It simply became much smaller.
Application software currently trades at 3.4x revenue compared with a five-year average of 7.0x. Vertical software trades at 4.4x. Companies relying on valuation benchmarks from 2024 may be using outdated assumptions.
The IPO window remains effectively closed for businesses that are not clearly tied to AI. M&A activity remains active but highly disciplined, and secondary transactions have become the most popular mechanism for providing partial liquidity.
If you run a software company or sit on a board, the question for the next meeting is no longer, “What is the market multiple?” Instead, it is, “Which market are we actually in, and what are we doing about it?”
According to Rob Bartlett’s May 2026 update from Jefferies’ technology investment banking team, the divide in the market is difficult to ignore. Only three public software companies are expected to grow more than 30% over the next twelve months, and they command a 21.0x revenue multiple. Companies growing below 10% trade at 3.1x.
The middle ground has largely disappeared.
This is not one market experiencing compression. It is two entirely different markets sharing the same Bloomberg terminal.
Since ChatGPT launched in November 2022, the companies in Bartlett’s AI Darlings group—Broadcom, Google, Meta, Microsoft, NVIDIA, Oracle, Palantir, and Amazon—have risen by 565%. AI Beneficiaries are up 160%. Cybersecurity companies have gained 94%. Vertical software is up 5%, while horizontal application software has declined 16%.
The shift is far from theoretical. Harvey has reached an $11 billion valuation while targeting Thomson Reuters and LexisNexis. Sierra and Decagon are AI-native customer service platforms challenging Salesforce and ServiceNow. Glean, valued at $7.2 billion, is attacking markets where Microsoft 365 and ServiceNow once commanded premium pricing.
Jake Saper of Emergence recently described the Cursor versus Claude Code dynamic as a warning sign for the broader industry. One product controls the developer workflow while the other merely executes tasks. Developers are increasingly choosing the latter. As agents begin performing the work, the traditional advantage created by having humans deeply embedded within software starts to weaken. Workflow stickiness, which has been a central argument in growth-stage software investing for years, now requires fresh justification.
Boards and leadership teams should be asking a difficult question: Is our growth story still credible in an AI-native world, or are we managing a melting ice cube?
A company growing 12% annually with a 22% free cash flow margin may appear healthy. However, if investors believe AI-native competitors could erode customer renewals within two years, enterprise value is already being discounted, whether management acknowledges it or not.
Founders are often the last people to recognize these shifts. Boards exist precisely for moments like these.
Two important facts frame the current environment.
First, diversified technology giants remain financially strong. Microsoft, Alphabet, Oracle, Broadcom, Salesforce, and their peers collectively trade around 5.5x revenue, slightly above their long-term averages and significantly above pre-pandemic levels. These companies possess strong acquisition currency and are deploying it selectively to acquire AI capabilities, distribution channels, or proprietary data that cannot easily be built internally.
Second, private equity firms continue to invest aggressively, though with greater discipline. Thoma Bravo recently closed a $24.3 billion software fund. Orlando Bravo has argued that markets are overestimating AI disruption while undervaluing domain-specific software businesses. Holden Spaht has been even more direct, suggesting that public markets are failing to distinguish between software that large language models can replace and software they cannot.
Recent take-private transactions involving Dayforce and Verint demonstrate that Thoma Bravo is backing those views with capital.
However, cautionary examples have also emerged. On April 22, Thoma Bravo handed Medallia back to its lenders. The $6.4 billion acquisition completed in 2021 resulted in roughly $5.1 billion of equity losses. It marked the second major SaaS equity wipeout in eighteen months following Vista’s transfer of Pluralsight in 2024.
The take-private strategy still works for the right companies purchased at the right price. It does not work universally.
For sellers, this means that there may be more potential acquirers in 2026 than there were in 2023. However, buyers are more disciplined, better informed, and operating with playbooks redesigned for the AI era.
Deals that once closed in ninety days may now require one hundred fifty days. Strategic buyers who previously paid for synergies are demanding evidence of AI defensibility before offering premiums above the value of recurring revenue.
This is not a bad market. It is simply a different one.
The biggest change visible in deal activity during 2026 is not traditional M&A or take-private transactions. It is structured secondary liquidity.
Carta reported that tender offer volume increased roughly 60% during 2025, and 2026 is expected to surpass that pace.
Large company-led tender programs are becoming increasingly common. OpenAI completed a $6.6 billion tender offer at a $500 billion valuation. Stripe is conducting another transaction at a valuation exceeding $140 billion. Anthropic is reportedly preparing one at $350 billion, while SpaceX continues its semiannual programs.
These are not fundraising rounds. They are liquidity events designed to allow employees and early investors to sell shares while companies remain private.
Pricing and oversubscription have become critically important. Notion’s $270 million tender earlier this year experienced demand that exceeded available capacity, forcing prorated allocations and prompting a public apology from the CEO.
General partner-led secondaries and continuation vehicles have also accelerated. Funds raised between 2017 and 2019 are reaching stages where investors expect distributions, yet many portfolio companies remain strong despite a slow exit market.
Continuation vehicles allow managers to transfer high-quality assets into new funds, return capital to investors seeking liquidity, and preserve upside for those who remain invested.
Secondaries are also moving down-market. Harvey combined a $75 million tender with a $150 million primary round at an $8 billion valuation. Gamma completed a $44 million tender benefiting approximately thirty employees.
Secondaries are no longer limited to the largest technology companies.
Boards considering secondary transactions should keep four principles in mind.
First, involve legal counsel early. Securities regulations, fiduciary duties, valuation methodologies, and information rights all intersect. Pricing decisions are especially sensitive.
Second, a secondary transaction is not a fundraising round. While it changes ownership dynamics significantly, it does not inject primary capital into the company.
Third, the secondary price establishes a reference point for future valuations, including 409A analyses, primary financings, and eventual exits. Pricing methodology should be carefully documented.
Fourth, the most successful secondary programs are executed before liquidity pressures become urgent. Boards that delay often negotiate from a position of weakness.
During the era of 7x revenue multiples, operational sloppiness could often be corrected during negotiations. In today’s 3.4x market, poor preparation can kill deals or reduce valuations significantly.
Regardless of whether the goal is a sale, take-private transaction, secondary liquidity program, or strategic review, the preparation is largely the same.
Clean up the cap table. Resolve outstanding SAFEs, convertible notes, acceleration triggers, and unexpected equity grants before they become problems.
Conduct a thorough intellectual property assignment review. Buyers in the AI era closely examine code ownership and provenance, including work performed by contractors, former founders, offshore developers, and engineers using AI coding tools.
Maintain three years of audited financial statements. Reviewed statements are no longer sufficient for competitive processes involving meaningful transactions.
Be transparent about customer concentration and renewal trends. If one customer accounts for a significant portion of annual recurring revenue, address it proactively rather than waiting for diligence to reveal the issue.
Most importantly, develop a credible AI narrative. Buyers care less about future plans and more about what has already been built, how customers use those capabilities, and whether retention data demonstrates real value.
None of these tasks are glamorous, but they are increasingly separating companies that achieve successful exits from those whose deals collapse after the letter of intent.
The capital remains available. Strategic acquirers are active. Private equity firms continue to invest where the economics make sense. Secondary liquidity markets are deeper and more sophisticated than ever before.
What has disappeared is the era when companies could point to a 7x revenue comparable and assume growth would compensate for every weakness.
Medallia’s lenders learned that lesson. Pluralsight’s stakeholders learned it as well. Many boards overseeing horizontal software companies have watched valuation multiples decline sharply while hoping the cycle would reverse.
The best valuation opportunity may have been in 2021.
The second-best opportunity is the one preserved through decisions made today.
If a transaction is likely within the next twenty-four months, preparation should begin at the next board meeting, not the one after that.
Call before the out-of-office replies begin. I answer the phone.
Data Source: Rob Bartlett and the Jefferies Technology Investment Banking team, Monthly Software Market Valuation and Performance Update, May 2026. Figures are based on data as of April 30, 2026. Examples referenced are drawn from publicly reported activity.
Stay updated by subscribing to our newsletter
Attorney advertising – prior results do not guarantee a similar outcome.
Opinions expressed here are my own and not those of my law firm.
Please read and acknowledge the following before using this AI Assistant:
• This AI Assistant provides information for general informational purposes only and does not constitute legal, tax, financial, medical, or other professional advice.
• Responses are generated by artificial intelligence and may be incomplete, inaccurate, or outdated. Important information should be independently verified.
• Use of this AI Assistant does not create an attorney-client relationship with Louis Lehot or any affiliated law firm.
• Do not submit confidential, sensitive, or privileged information through this chat.
• Decisions should not be based solely on AI-generated responses. Please consult a qualified professional for advice specific to your situation.
• This website and AI Assistant may be considered attorney advertising in some jurisdictions.
• Past performance is not a guarantee of future outcomes. AI-generated responses are not a guarantee of future outcomes or accuracy.
• By selecting “I Agree”, you acknowledge that you have read, understood, and accepted this disclaimer.