What Happened to Medallia, In Plain Numbers
Thoma Bravo lost $5 billion. Not on paper. For real. On June 17, 2026, the private equity firm handed Medallia to a lender consortium led by Blackstone, Apollo, and KKR, wiping out its entire equity stake from the 2021 take-private Source.
That's the direct answer to the headline question. Medallia didn't die from bad customers or bad code. It died because it couldn't out-build AI-native competitors fast enough to service $2 billion in debt, and its owner refused to pay to fix it.
Orlando Bravo, the firm's co-founder, admitted publicly that Thoma Bravo overpaid. He said so on record, to the Financial Times, about his own deal Source. That kind of admission doesn't happen unless the math has gone from "difficult" to "impossible."
I've stood watch on submarines where one bad call sinks the boat. Medallia is that story with better lighting. This is not a SaaS company story. It's an operator story, and it applies to you whether you run a $400 million software platform or a $2 million services shop.
TL;DR:
- Thoma Bravo's $5 billion Medallia loss is the second-largest equity wipeout in private equity history, and the new owners immediately committed $150 million to accelerate AI, not preserve the old model
- SaaS valuations compressed from 6.2x revenue to roughly 3.3x in eighteen months while PE software deal activity cooled sharply, meaning passive ownership is now a liability, not a strategy
- Cross-stack AI joint ventures from Blackstone-Anthropic and OpenAI's Deployment Company show where capital is actually moving, and owner-operators without a system for AI adoption are on the wrong side of that line
The Deal That Should Scare Every Operator
Here's the sequence. Thoma Bravo took Medallia private in 2021 for $6.4 billion, financing roughly $2 billion of it with debt Source. Rates rose. Interest costs ballooned.
Growth slowed. The company kept borrowing to cover its own interest payments, and the debt load grew by more than $1 billion after the buyout closed.
Lenders asked Thoma Bravo to inject $500 million to cut the debt. Thoma Bravo said no. It chose to eat the largest loss in the firm's 20-year history instead of feeding new capital into a business it no longer believed could out-execute the market Source.
That's not a liquidity problem. That's a confidence problem. Confidence problems in capital allocation always trace back to the same root: nobody could show the board a system that would fix the growth rate.
The new owners moved differently within hours. Blackstone, Apollo, and KKR committed $150 million in fresh capital specifically to "accelerate AI-driven innovation," funding CEO Mark Bishof's existing $500 million product and AI investment plan Source.
Same asset. Same debt overhang. Completely different bet, because the new owners believe an AI-native rebuild can win where the old operating model couldn't.
That's the whole doctrine in one transaction. Passive ownership watched the boat take on water. Active ownership installed new watertight doors and turned the pumps on.
Systems Beat Slogans: The Valuation Math Doesn't Lie
Public SaaS valuations fell from a median 6.2x revenue at the end of 2024 to roughly 3.3x by March 2026, erasing close to $1 trillion in market value across the sector Source. Private software buyouts didn't escape it either: PitchBook-cited data shows software-focused buyouts totaled about $50 billion in the first five months of 2026, down from roughly $88 billion over the same window in 2025, the weakest start to a year since the pandemic Source.
Platform buyout value in software fell to about $16.24 billion through the first five months of 2026, a pace that would land 2026's full-year total near a quarter of 2025's $156 billion record. Platform deals dropped to just 41% of total software PE deal value, the lowest share in at least a decade Source.
Read that again slowly. The multiple didn't compress because these businesses stopped making money. Medallia's operations are, by every account, running fine day to day. The multiple compressed because buyers stopped trusting that legacy software companies could defend their moat against agentic competitors.
Multiple compression is the market pricing in founder dependency tax at the category level. It's the same math I use when I evaluate a business for acquisition: if the owner is the system, or the platform can't prove it's ahead of the AI curve, the multiple gets cut whether the P&L looks good or not.
Meanwhile capital hasn't disappeared. It's just gotten selective. Roughly $3.7 trillion in private equity dry powder is still sitting there, according to McKinsey figures cited across multiple reports, and it's chasing AI-native capability, proprietary data, and engineering talent instead of raw revenue growth Source.
That's not a retreat from software. It's a retreat from software that can't prove it has an engine room.
The Owner-Operator Frame, Applied to Your Business
I use the Owner-Operator Frame with every founder I coach: your business is either an asset you're compounding or a job you're trapped in. Medallia, at the balance sheet level, was still an asset. What broke was the system underneath it, the one that should have translated capital into AI capability fast enough to defend the multiple. Thoma Bravo held the wheel and steered nowhere.
You don't have $2 billion of debt. You probably don't have a board threatening a debt restructuring next quarter. But you have the same exposure in miniature: a business that either has a system for adopting AI into its actual workflow, or a business that's betting the owner will personally figure it out eventually, sometime, when things calm down.
They're not calming down. Cross-stack partnerships are now standard architecture, not experiments. In May 2026, Anthropic launched a $1.5 billion joint venture with Blackstone, Hellman & Friedman, and Goldman Sachs specifically to embed AI engineers inside mid-sized, PE-owned companies and rebuild their workflows around agents Source. Hours later, OpenAI countered with its own Deployment Company, backed by Bain Capital, TPG, Brookfield, and Advent, targeting up to $10 billion in commitments toward the same goal Source.
Read what that actually means. The largest capital pools on earth just built permanent infrastructure to force AI adoption into portfolio companies whether the operator asked for it or not. If you're not building that system yourself, someone with a checkbook is going to build it for you, and they'll take a bigger slice of your equity to do it.
The 90-Day Bottleneck Audit Is Cheaper Than a Recapitalization
I ran the 90-Day Bottleneck Audit with a client last year who thought his agency was healthy because revenue was flat, not falling. Flat felt safe. We mapped every workflow that touched a client deliverable and found four bottlenecks that were 100% dependent on him personally, none of them documented, none of them AI-assisted.
He wasn't losing money. He was building founder dependency tax without knowing the invoice was coming due at his next valuation event.
That's the compressed version of what happened to Medallia at a thousand times the scale. Flat growth, high debt service, no visible fix, no system anyone could point to. The market didn't wait for the fix. It repriced first and asked questions never.
Small businesses are moving on this even if enterprise software wasn't fast enough. 77% of U.S. small and midsize businesses reported regular AI use as of January 2026, up from 48% just eighteen months earlier, according to Intuit QuickBooks' 2026 AI Impact Report Source. But adoption isn't the same as a system.
Only 14% of small businesses report AI is fully integrated into core operations, per Goldman Sachs' small business research Source. That gap between "we use AI sometimes" and "AI runs inside our actual system" is the exact gap that turned Medallia's balance sheet into a case study.
Doctrine Connection: Systems Beat Slogans
Medallia had every slogan a board could want. "AI-driven innovation." "Customer-focused product investment." "$500 million innovation commitment." None of it saved the equity because slogans aren't systems.
A system is documented, repeatable, and doesn't depend on any single person's memory or mood on a Tuesday. Thoma Bravo had capital and slogans. It didn't have a watertight system for converting AI investment into growth fast enough to outrun its own interest payments.
Your business doesn't need a press release about AI. It needs a documented process for where AI plugs into your actual bottlenecks, verified against real output, not optimism. That's the whole difference between an operator who gets acquired at a premium and one who gets handed to the lenders.
The Math You Should Be Running This Week
If your business went up for sale tomorrow, could a buyer see a system, or would they see you? That single question is the informal audit private equity ran on every software company this year, and it's why deal count fell even as the largest deals got larger. PE activity across software specifically dropped to about $19.9 billion across 86 transactions in a recent quarter, a decline of more than 40% from the prior quarter, even as headline M&A totals stayed inflated by outlier megadeals like SpaceX's $250 billion acquisition of xAI Source.
Strip the outlier and the real story is contraction, selectivity, and a hard preference for businesses that can prove operator-independent systems. That preference isn't going away. It's the new floor.
Frequently Asked Questions
Q: Is the Medallia situation just a private equity problem, not something that affects small business owners? No. The mechanism that killed Medallia's valuation, an operating model too slow to prove AI-driven growth, applies at every size. A four-person agency with no documented AI workflow carries the same risk in miniature: lower sellable value, higher founder dependency, and a widening gap versus AI-native competitors who move faster with less overhead.
Q: What does "AI readiness" actually mean if I'm not a software company? It means your core workflows, the ones that touch revenue, delivery, and customer retention, have at least one AI-assisted layer that's documented and repeatable without you personally running it. It's not about having a chatbot. It's about removing yourself as the single point of failure in the process AI could handle.
Q: Why did Thoma Bravo's new owners commit fresh capital instead of just cutting costs? Because cutting costs doesn't fix a growth problem, and the new owners believe growth requires AI-native rebuilding, not austerity. The $150 million commitment funds engineering and product work aimed at closing the competitive gap, not shrinking the company into stability Source.
Q: How do I know if my business has the same exposure Medallia had? Run a bottleneck audit. List every process that depends entirely on you or one irreplaceable employee, with no documentation and no AI assist. Every item on that list is a discount on your eventual sale price, whether you ever sell or not.
Q: Is this the right time to invest in AI systems, given how uncertain the market looks? Uncertainty is exactly why the investment matters now. The businesses absorbing the AI joint-venture wave, Blackstone-Anthropic, OpenAI's Deployment Company, are targeting mid-sized companies specifically because that's where the compounding advantage is largest and least defended. Waiting for certainty means waiting for someone else to take the seat.
Your Move This Week
Pick one workflow in your business that depends entirely on you and has zero documentation. Write the steps down like you're handing the business to a stranger tomorrow. Then find the one step in that process an AI tool could take off your desk within 30 days.
That's not a strategy deck. That's the first entry in your own engine room manual, and it's worth more to your eventual buyer than any slogan you could put in a press release.
*Jeff Barnes is the founder of demg.ai and Digital Evolution Marketing Group. This article is educational and does not constitute business, legal, or financial advice. All claims are sourced where possible. Results vary by business, market, and execution.*