Surviving The Software Selloff

A New Private Equity Playbook for Software Value Creation in an AI-Driven Market 

The software multiple compression that began in 2022 is no longer a cyclical correction. With leading software names trading at a fraction of their 2021 revenue multiples and artificial intelligence (AI) reshaping how software is built, sold and consumed, private equity (PE) sponsors are confronting a more difficult question – which assets still have a credible path to their underwritten exit and which require a fundamentally different plan.

For years, software investors viewed multiple expansion as predictable and central to their investment thesis, underwriting assets based on durable recurring revenue and strong customer retention. That thesis is being stress-tested asset by asset. Some businesses are emerging stronger, more embedded and increasingly AI-leveraged, commanding premium multiples despite a more conservative market. Others, however, are discovering that the characteristics that justified peak pricing four years ago no longer carry the same weight. Differentiating between these two distinct company types and acting decisively on that diagnosis is now a critical task for sponsors and management teams.

Four Realities of the Structural Software Market Reset

At the product level, AI is replicating or streamlining discrete workflows, particularly where functionality is rules-based or not deeply embedded. AI-native tools offer faster or lower-cost alternatives, increasing pressure on more interchangeable software categories. This dynamic is reflected in public markets, where share price volatility is increasingly tied to perceived AI adoption. Customer behavior is also shifting, with greater scrutiny on software spend and pressure on traditional pricing models. The predictability that historically supported leverage and flexible capital structures is becoming less reliable and causing significant ripple effects across the software market.

Reality #1 – Valuation Multiples Have Reset, but Quality Still Commands a Premium

The result is significant valuation compression. Public SaaS revenue multiples have declined materially from 2021 peaks, as investors apply greater scrutiny to downside scenarios and operating performance. Median public SaaS EV-to-revenue (EV / R) multiples have reset from approximately 18x in late 2021 to roughly 6x to 7x entering 2026.¹ The iShares Expanded Tech-Software Sector ETF has declined approximately 30% from its September 2025 peak, reflecting increased volatility across public software equities.²

Reality #2 – Dispersion Separates Durable Platforms from At-Risk Assets

At the same time, dispersion has widened significantly. Top quartile public SaaS companies are trading at approximately 13x to 14x EV / R, while bottom quartile assets trade closer to 1x to 2x.³ This spread signals a more selective market where quality and durability drive outcomes.

This valuation pressure is compounded for companies that borrowed from private credit providers under peak-market assumptions. Many software loan books include assets acquired at peak valuations with leverage ratios of 6x to 7x EBITDA or 3x to 4x ARR, underwritten on assumptions of high customer retention and stable growth which no longer reliably hold. Even modest changes in retention, pricing or growth can tighten EBITDA cushions, reduce covenant headroom and limit operational flexibility.

Reality #3 – Extended Hold Periods are Intensifying Exit Pressure

Pressure is most acute for assets held beyond their original investment horizon. Continuation vehicles (CV), net asset value (NAV) loan facilities and GP-led secondaries have extended periods across the industry, with many assets approaching a sixth year of ownership or beyond without a clear path to exit. PE firms are managing a growing backlog of unsold assets, with recent reports estimating approximately 30,000 portfolio companies representing roughly $3.6 trillion in unrealized value and materially extended hold periods across the industry.⁴ Waiting for the market to recover is no longer a viable strategy.

Reality #4 – Subsector Performance Is Diverging as AI Reshapes Demand

These dynamics are not affecting all companies equally. Cybersecurity businesses are experiencing tailwinds as increased automation drives complexity and demand for protection, while contact center management and horizontal productivity tools face steeper headwinds given their functions are more easily replicated. The gap between durable platforms and more commoditized solutions continues to widen.

This is a market reset, not a dislocation. Capital remains available and high-quality assets continue to attract interest. However, the definition of software quality has evolved. Sponsors need a more disciplined framework for assessing whether an asset can defend relevance, pricing and growth in an AI-driven market. Three factors now shape that assessment – mission criticality, AI maturity and pricing architecture.

A Three-Part Framework for Evaluating AI-Ready Software Assets

Part 1 – Determine Mission Criticality

Software that is deeply embedded in core workflows can be costly and disruptive to replace, even if alternatives appear less expensive. Enterprise resource planning (ERP) systems, electronic health records, vertical software, payments infrastructure and supply chain platforms tend to share this profile, where functionality is tightly integrated with daily operations. Deep data moats, high switching costs and specialized builds signal durability and lower attrition risk. Sellers should support these attributes with evidence of integration depth, data exclusivity and customer-specific builds rather than relying on marketing claims.

Assets lacking these characteristics, particularly those that sit at the edge of workflows, rely on generic functionality or lack proprietary data, are more exposed to displacement. To mitigate this risk, companies should aim to deepen integration, expand into adjacent workflows and build differentiated data assets. For example, a scheduling tool that integrates payments, captures customer history and automates follow-ups becomes more embedded in daily operations than a standalone calendar product. Greater specialization and tighter alignment to customer outcomes can further reinforce resilience.

While mission criticality remains necessary, it is no longer sufficient to drive value. Assets that score well on traditional embedded software metrics, whether it is high net retention, strong customer concentration in regulated verticals or deep workflow integration, still cannot grow back into the multiples at which they were acquired. Mission criticality protects against churn and prices in a floor, but it does not on its own generate the growth profile today’s buyers require for premium valuations.

Part 2 – Validate AI Maturity

Assets that clearly demonstrate how they benefit from AI adoption through quantifiable KPIs will command premium valuations. The bifurcation is now stark. AI-native SaaS companies are trading at 25 to 30x EV / R in public markets, while traditional SaaS sits at 5 to 7x.⁵ While many companies may position themselves as AI-powered, this alone does not indicate meaningful product evolution. Businesses that command premium valuations will provide specific, defensible explanations of how AI is embedded within their offering and how it drives measurable outcomes such as time savings, cost efficiencies, revenue expansion or EBITDA improvement. For example, a customer support platform that deploys AI to automate ticket resolution and demonstrates a 40% reduction in response times alongside a 25% decrease in support costs presents a far more credible value proposition. A platform that has rebranded existing automation as AI does not.

The gap between AI claims and AI reality is now a primary diligence workstream. We routinely see management presentations claiming material AI-driven cost savings or revenue uplift that often compress significantly under scrutiny. Sellers that demonstrate measurable economic outcomes from AI will command premium multiples, while those that cannot may face structured outcomes, earnouts or unsuccessful processes.

Beyond identifying present day use cases, AI maturity requires a clear view of forward-looking risk exposure. Companies should evaluate how continued disruption could impact pricing, differentiation and customer acquisition, including specific scenarios such as changes in per-seat pricing models, accelerated churn from AI-native competitors or shrinking software budgets to accommodate AI deployment. Modeling and quantifying these specific scenarios can clarify an asset’s full financial exposure so management can respond strategically through deeper integration, vertical specialization or investment in proprietary data assets.

Part 3 – Assess Pricing Architecture

AI is reshaping how customers interact with software and how it is priced. As automation reduces the need for manual user input, engagement is shifting from consistent seat-based usage toward more intermittent or outcome-driven interaction. For example, a marketing team that previously required multiple users to build and optimize campaigns may now rely on a single AI-driven tool to generate, test and deploy campaigns with minimal human involvement, reducing the need for multiple paid seats. This shift often creates misalignment with traditional SaaS models, driving pricing scrutiny and customer pushback that can erode revenue quality and pressure EBITDA.

In response, software companies are adopting alternative pricing models including consumption-based pricing tied to usage volumes, outcome-based models linked to measurable results and tiered or bundled offerings that package AI-driven capabilities at a premium. Some companies are also deploying AI as add-on or usage-based overlays to capture incremental value without disrupting the core offering. Each of these models creates a real strategic tension that sponsors and management teams must navigate carefully. Consumption-based and outcome-based pricing better align with how value is delivered in an AI-enabled environment, but they can reduce near-term revenue visibility. A predictable seat-based ARR base often presents better in a CIM than a consumption model whose run rate is harder to forecast, even if the consumption model is more durable over the long term. Companies approaching a process within 12 to 18 months face a genuine trade-off. Sponsors and management teams must decide whether to protect ARR optics ahead of an exit or evolve pricing for longer-term durability and accept increased disclosure complexity.

Regardless of the model, leadership must clearly articulate how pricing reflects both customer value and long-term growth strategy. Companies that proactively evolve their pricing architecture will be better positioned to maintain alignment, defend margins and present a more durable investment profile.

Applying the Framework to Portfolio Strategy and Exit Planning

Once sponsors understand where an asset falls across mission criticality, AI maturity and pricing durability, the strategic path becomes clearer. Sponsors are increasingly facing portfolio situations where standard approaches do not apply. Three patterns are emerging:

1. Some assets are mission critical and have incorporated AI but were acquired at multiples the current market will not support. In these cases, sponsors are pursuing GP-led secondaries, partial recapitalizations and structured exits, focusing on capital recycling rather than maximizing valuation.
2. Other assets face more fundamental risk, where underlying workflows are being absorbed into AI-native platforms. In these situations, consolidation or a strategic sale may provide a more viable path than waiting for an independent outcome.
3. In cases where the path forward is unclear, operational diligence can stress-test the durability of the business before additional capital is committed.

The buyer universe has also evolved. Strategic acquirers, particularly large software platforms, remain disciplined and focus on assets that fill platform gaps or accelerate AI capabilities. Sponsor buyers are increasingly selective, prioritizing assets with clear AI narratives and conservative leverage profiles. Take-private activity remains active as public market dislocations create opportunities. Positioning should align with the most credible buyer before launching a process.

Traditional Value Drivers Still Need to Support the Exit Narrative

Even as the software industry undergoes structural change, certain performance benchmarks remain relevant. The Rule of 40 continues to serve as a useful indicator of quality, with companies that meet or exceed this threshold still commanding premium valuations. Addressing operational or financial exposures proactively also allows sponsors to shape the narrative rather than leaving buyers to identify these issues during diligence.

At the same time, valuation expectations must be grounded in current market conditions. Peak 2021 benchmarks are unlikely to return in the near term and anchoring exit assumptions to them can undermine credibility. Sponsors seeking strong outcomes should position assets against today’s operating and valuation landscape, ensuring buyers assess performance relative to realistic and relevant comparables.

Footnotes

1. SaaS Capital Index; Aventis Advisors SaaS Valuation Multiples 2015 to 2026; Software Equity Group 2026 Annual SaaS Report
2. PitchBook; public market data for iShares Expanded Tech-Software Sector ETF (IGV)
3. SaaS Capital Index; Software Equity Group 2026 Annual SaaS Report
4. The Wall Street Journal, “Private Equity Confronts Swollen Investment Backlogs With Dealmaking Stuck,” June 2025
5. PE International AI M&A Trends 2026; PitchBook