Slowing SaaS Revenue Growth and AI Fear & Adoption are Suppressing Valuations and Capital Velocity

The Software Community and possibly the entire world currently sits in AI purgatory. We watch the stock market indexes hit record highs on colossal expectations that frontier LLMs and AI startups will be the conquerors of all end markets, in many markets and subsectors where they haven't yet generated a dollar of revenue. The AI world is facing endless challenges and missed milestones. The conundrum is the enormous wealth creation that has transpired purely based on future performance of highly speculative bets. 50% of corporate earnings growth is coming from the AI infrastructure buildout. For instance, Samsung expects its semiconductor business to earn more this year than it has over the past 40 years combined. Do you think this might be unsustainable? This has the potential to be a bubble many times more than what we experienced with the dot-com, mortgage crisis or COVID bubbles. As great as AI is and will be, it has not yet come close to delivering on the hype that Silicon Valley and the media have showered upon it.

Virtually all public and private AI profits have come from the massive infrastructure buildout, not from successful use cases or measurable ROI for end customers. Up until recently, corporate and individual users have been using AI at a deep discount to its cost. Those discounts are ending and customers now face exploding AI bills that LLM providers will no longer subsidize. If companies across the world do not start to see real ROI soon from their massive AI experiments, there will be a major pullback on infrastructure spend and the global AI bubble will quickly deflate.

The good news is that the world has gone head over heels for engaging and experimenting with AI over the last couple of years, and AI needs the software world to convert the superpower of these LLMs into real work products and ROI. Gartner estimates that there will be ~$1.4T spent on AI infrastructure and another $450B spend on the application layer, with total 2026 AI spend at over $2.6T! Taking the optimist's perspective, when the business and consumer worlds commit so much mindshare and money to the deployment of AI in our everyday activities, there will eventually be success and significant adoption as those breakthroughs make their way through the system. The world currently spends roughly $3.3T on software and tech services. AI is currently reimagining and rebuilding the software landscape, but it is also expanding software's old frontier into much broader functionality, and with it, a much larger TAM. As AI shifts software from systems that organize work to systems that execute work, the software TAM expands beyond traditional enterprise IT budgets into labor and broader operational spend. Many of the AI-forward software incumbents will remain leaders in their areas of domain expertise and will share the bounties of this new TAM with the AI titans and AI startups. The lowly 11% growth rate that we are experiencing today for public software companies will increase when customers are truly benefiting from AI and ROI can be demonstrated.

SaaS incumbents and tech PEs have spent the last year in AI assessment/deployment mode. The proof will be in the pudding, which for most software incumbents and AI startups will take at least another year to prove out. The first phase of the journey is driving AI today through every part of a company's core operations, including development, marketing, finance, etc. That is the easy part. The second phase is weaving AI into incumbent software platforms. SaaS incumbents cannot and should not try to compete at the model layer. Models are becoming commodities. The top frontier models are increasingly similar, but software owns the workflows, data, integrations, permissions, and business logic that make AI useful inside a customer's business. The losers will try to become value-add token resellers. The winners will own every critical process around the model, letting AI access and act on that software layer.

SaaS incumbents need to decide how much of the AI environment they control versus outsource: whether to rely on rented frontier model APIs or self-hosted models, which party controls the layer where agents access the product, and whether proprietary data pipelines, customer workflows, and operational IP remain protected inside their own platform or are exposed to third-party model environments. This is the most sensitive and critical AI decision set where a SaaS incumbent chooses to maintain AI sovereignty or outsource its AI and expose its secret sauce to OpenAI and Anthropic. Successfully pioneering phase two of the journey will determine both the quality of your defensive moats and the strength and agility of your offensive positioning. We expect most SaaS incumbents will build their own AI environment around their proprietary purpose-built solutions and data pipeline to continue to provide the best solution in the market at the lowest cost.

The third phase of the journey is building, testing, and generating meaningful revenues through extending their product offerings to agentic workflows. These products/services are very early and unproven for most software incumbents. The stories are racking up of both incumbents and AI startups struggling to prove out the ROI for these agentic solutions. Moreover, for many SaaS companies, their end customers do not have the data infrastructure or controls today to properly leverage the latest AI capabilities – some end markets just aren't ready for AI. While there are always exceptions of success, the norm seems to be money-losing deployments with OK-to-low customer satisfaction. Only time will tell who the winners and losers will be in this new AI/Software frontier.

Forty years of over 25% hypergrowth for software has resulted in an industry much bigger, far more penetrated, and much slower growing. PE entry valuations were predicated on better growth, performance, and higher exit valuations – all of which have underperformed across the board and are pushing out exits. As dry powder hits an all-time high, it is harder than ever to find that fast-growing, quality company. When you do find it, everyone else is also at the party with more money than ever before, pushing valuations to the stratosphere for quality assets. It should be noted that these are the best of the best companies with AI-forward businesses and Rule of 40++ metrics.

While we grind away in this AI purgatory, public SaaS companies under $10B EV are trading at 2.4x and 10x 2026 calendar revenues and EBITDA, respectively. The private markets continue to trade at a huge premium to public markets, with AGC's median revenue multiple of 9x and EBITDA multiples of 20x+ for 2026 platform deals.

Dry powder for PE Tech funds surpassed $500B for the first time ever in June. PE funds have been investing twice as much as they have exited for years while raising funds at record speeds. More recently, investing has become more difficult, creating a massive slowdown in capital deployment which has ballooned the dry powder kegs. Supply of quality companies is low because PE GPs do not want to bring their companies to market if they can't hit the current marks to market which is not possible for many of the 4,600 companies they currently hold. Given "SaaSpocalypse" and concerns of AI risks, many of these companies could not be sold for their holding values. The companies we bring to market today with AI-forward stories and strong fundamentals are capturing the highest values ever.

Slowing SaaS revenue growth, AI fear & the prioritization of AI adoption are suppressing valuations and capital velocity. The global tech PE community has only exited 79 companies through the first half of this year, which puts them at a run rate of 158 exits. With 4,600 majority-owned companies in the portfolios of the tech PEs, at this pace, it would take 28 years to sell these companies. On the capital deployment side, PEs have invested in 79 companies which is down 56% from 2025. They have done 407 add-ons which is down 24% from last year. Excluding hardware, overall tech and AI deal value are at an annual run rate of $175B which is down 44% from last year. To put it mildly, it is not a lot of fun to be an AI and tech investment banker nowadays ☺.

If the 160 public SaaS companies are growing at 11% with a 36% on the Rule of 40, then it is fair to say that roughly half the private SaaS companies are growing less than 11% and probably operating below 36% on the Rule of 40. A large chunk of these PE- and founder-backed software companies do not have the profile for a standalone platform investment by the PE funds. These lower- and mid-tier companies will likely be the first to transact when the dam breaks and deals start to flow again. From this group of assets, the typical platform deal will be 8-20% growth and high EBITDA margins, with AI defensibility and offense with a more focused acquisition gameplan. When the dam breaks, we could see a relative flood of companies hitting the market. Average holding periods are currently at 5 years for the 4,600 PE controlled software and tech services assets. Add to that the minority-, VC- and founder-controlled assets and the numbers get quite large.

The day has finally come where strategic buyers are acquiring more AI & software companies than the PEs are. In 2025, the PEs acquired 1,472 companies whereas the strategics only acquired 583 – 28% of the pie. Over the last ten years, as the tech PE world has rocketed to $3.2T in AUM, they have also taken over as the dominant acquirer of software. In the first half of '26, with AI becoming a strategic priority, the crushing blow of the SaaSpocalypse, slowdown in software growth, the strategics raced ahead of the PE buyers with an annual run rate of ~1500 deals (62%) versus the PEs with ~900 (38%). The PEs, with $500B currently in their wallets, will be back in force in the not-too-distant future.

It is somewhere between hard and impossible to predict when optimism and greed will overtake fear and we will begin to sell and merge this massive backlog of AI and tech companies. What is certain, is that day will come and when it does, we will see a significantly larger volume of dealmaking than we have seen in the last four years. As the supply of companies coming to market rises to meaningful levels, it won't be just the highest quality private companies, it will include more of the middle of the road assets. PE and strategic investors will have a lot more opportunities to choose from and the private valuations will move closer to the public company benchmarks – eventually settling into the new normal for software valuations. In this new world we see four flavors of deals: i. lesser growth add-ons going for 0-3x revs; ii. low growth platforms at 3-6x; iii. high growth platforms at 4-12x; and, iv. high octane deals to strategics at 8x+.

After a tough first quarter, AGC Partners has clawed back into the game, closing 8 deals in the second quarter. We currently have 10 deals under LOI and 20 deals in preparation stage. Our Boston conference on September 17th already has 135 CEOs confirmed – 3 times the pace compared to last summer. Although the fog, uncertainty, and fear around AI have eased, it continues to be a fierce headwind for GPs and founders to get liquidity, commit capital, and make deals. In the face of these continued challenges AGC Partners is creatively threading the needle, powering through to close deals again, and pitching at an accelerated pace.

AGC is currently brainstorming with founders, CEOs, and GPs on how and when to optimize the path forward, whether it be no action today or full speed ahead on a deal. Our Partners' deep AI and subsector knowledge, highly active live deal experience, and strong PE and strategic relationships empower our Partners to be exceptional advisors for CEOs trying to chart a course through these challenging waters. Whether the technology is AI, Cyber, GRC, or one of 50 verticals we cover, the AGC advisory team is deep in knowledge and relationships, ready to go 24/7, and able to power through those bumps in the road to bring about great outcomes.