Medallia: PIKing Up the Pieces

Welcome to the 210th Pari Passu newsletter.

This is one we have been waiting to write. Medallia is one of the most consequential restructurings in recent memory and one of the clearest case studies of the excesses embedded in the 2021 software vintage. 

Medallia was founded in 2001 by a married pair of management consultants who could not get a hotel to notice they were unhappy. Over the next two decades, it helped build the category now known as experience management, raised six rounds from Sequoia, went public in 2019, and became an enterprise standard for finding out what customers think. In October 2021, at the peak of the software valuation cycle, Thoma Bravo took the company private for $6.4bn. The purchase was financed not against earnings, because Medallia had almost none, but against recurring revenue, with part of the coupon deferred until the company grew into the debt. The structure assumed growth would continue, but in reality, it had already begun to fade. 

By the end of 2025, deferred interest had compounded the loan to roughly $2.8bn, rates had more than tripled the cash coupon, and the Blackstone-led lender group declined to extend relief any further. In June 2026, Thoma Bravo handed over the keys and wrote off approximately $5.1bn of sponsor and co-investor equity in one of the largest PE losses ever recorded and one of the largest restructurings private credit has ever executed. 

In today's writeup, we begin with Medallia's product and the niche it occupies in enterprise software, then trace the company from its founding through Sequoia, the IPO, and the Thoma Bravo buyout. From there, we follow the path to distress, featuring aggressive ARR-based underwriting, a retention curve already deteriorating, and the rate environment that plagued the 2021 vintage. We’ll finish with the June 2026 recapitalization and what the new capital structure leaves Medallia to solve next. 

One note before diving in. When you read Pari Passu, you are not reading journalism; our research team includes former bankers, investors, and lawyers, and we understand very well that investing is the business of taking risk, and if you are never losing capital, you are probably not taking enough risks in the first place. This is the story of a private equity transaction gone, obviously, very wrong, and our language will reflect the economic reality of the loss. Thoma Bravo and the other institutional investors involved with this name, like virtually all the other sponsors we cover, have generated fantastic returns for their investors, and no negative comment included in our research should be interpreted as a comment on their track record or skills.

You check into a hotel and your room is not ready. You wait in the lobby for forty minutes, nobody explains why, and when you finally get upstairs the air conditioning does not work. Three days later, an email arrives asking you to rate your stay from one to ten. You fill it out or you do not, and in most cases nothing visible happens either way.

Zooming out, a hotel operator with several thousand properties generates hundreds of thousands of guest interactions a week through surveys, app sessions, call center recordings, front desk notes, and public review sites. Buried inside that volume is the fact that six guests on the fourth floor of one specific property have complained about the same air conditioning unit. Someone has to spot that signal within hours rather than quarters, figure out that it belongs with the maintenance supervisor at that property rather than a corporate analyst in another time zone, and get it in front of him before checkout. Doing that across every property and every customer channel at once is the problem Medallia was built to solve.

Medallia is an experience management company headquartered in Tysons, Virginia, founded in 2001 to help large organizations collect customer and employee feedback, make sense of it, and route problems to the people responsible for fixing them [1]. The company serves roughly a thousand enterprise customers, including at least seven of the Fortune 10, and is built primarily for large, operationally complex businesses rather than broad-based adoption across smaller customers. Medallia reports as a single consolidated segment, so there is no divisional revenue breakdown, but its last public filings split revenue roughly 80% subscription and 20% professional services in FY’21 (ended January 2021) [2]. Roughly eight million people use the platform in a given week [1]. The product makes more sense when broken down by what happens to the data: how Medallia collects it, interprets what was said, observes what was done, routes the issue, and supports the deployment around it. 

Collecting Feedback:

The foundation of the business is the Medallia Experience Cloud, which serves as the central platform for experience data inside a client organization. If you have ever received a post-purchase survey, a post-flight questionnaire, or a “how did we do” prompt after a support chat, you have interacted with the collection layer of a platform like this one. The Experience Cloud ingests those responses alongside unsolicited feedback from review sites, social posts, app store ratings, and messaging channels, and brings them together into a common customer or employee view [1]. 

Medallia’s difference lies in what happens after it collects the feedback. The system applies rules to determine which signals warrant a response, who owns that response, and how quickly it needs to happen. This is called closed-loop feedback, and it is the mechanism that turns a complaint into a work ticket assigned to a named person with a deadline. Medallia also brings customer and employee feedback into the same architecture, allowing a client to best determine the origin of a given problem [1]. 

Interpreting What Was Said:

Most of what customers tell a company isn't a number on a ten-point scale, but a sentence in a comment box, a complaint in a chat transcript, or a four-minute phone call with a call center agent. Making sense of that material requires a different set of tools than tabulating survey scores, and Medallia has invested heavily in building them. 

Medallia’s platform transcribes contact center calls and analyzes the resulting text for sentiment, recurring themes, and the underlying causes of negative feedback, which means a client no longer has to sample a small number of calls for quality review. Every call can be searched and scored, expanding what the client can measure while reducing the amount of manual review required. Medallia acquired much of this capability through its 2020 purchase of speech-to-text vendor Voci Technologies for roughly $60mm [1]. 

Observing What Was Done:

Asking people about their experience has an obvious limitation: people respond inconsistently, and most do not respond at all. The alternative is to watch what they actually do. Medallia's digital experience analytics module tracks how users move through a website or mobile app, identifying where they hesitate, repeat an action, or abandon a task entirely. A user who quickly taps the same button four times and then closes the app has told the company something specific without answering a single question. This capability also came through acquisition, with Medallia buying Decibel Insight in 2021 for roughly $160mm, its largest disclosed purchase [1]. 

Acting in the Moment:

Up to this point, the platform is mostly telling the client what happened and, in some cases, why. Medallia has increasingly pushed the platform toward intervening while the interaction is still underway. Journey orchestration tracks a customer's activity across online and offline touchpoints and adjusts what that customer is shown or offered in real time. Medallia added the capability through its acquisition of UK-based Thunderhead in early 2022. A related product, Mindful, handles contact center callbacks by holding a customer's place in a phone queue and calling back when an agent is available rather than requiring the customer to remain on hold [1]. 

Medallia has also added generative AI across the platform over the past two years. The tools can automate parts of text analytics configuration, summarize long customer histories, draft responses to feedback, and let business users query experience data conversationally rather than build dashboards manually. The tools are already seeing meaningful adoption, with more than 40% of Medallia's top three hundred clients enabling them within twelve months of release [1]. 

Running the Program:

The final piece is not a product. Roughly half of Medallia's software contracts are sold as managed services, meaning Medallia's own people help design the experience program, configure it, run it, and interpret the output for the client. This is a deliberate part of the model rather than a byproduct of enterprise software delivery. Many large organizations do not have internal people equipped to run a global feedback program, so Medallia sells them the team along with the software [1]. 

However, the same model creates one of the platform's most consistent criticisms. Because programs are heavily customized, changes often run through Medallia's own services engineers rather than the client, and a modification that looks simple can take days to implement. Industry practitioners have described the platform to Pari Passu as the Ferrari of experience systems: capable of a great deal in the hands of specialists, but demanding in that same way [1]. 

That tradeoff defines where Medallia competes best. Its principal competitor, Qualtrics, is built more heavily around self-service, allowing business users to configure and launch programs without the same level of technical support. That makes Qualtrics faster to deploy and cheaper to enter. It has used that advantage to gain share in the mid-market and push further into Medallia's core enterprise accounts, while Medallia has remained strongest in large, complex global deployments where customization and hands-on support are requirements rather than drawbacks [1]. 

The more important structural issue is that fewer customers want to fill out surveys in the first place. Response rates have been falling, frequently below 15%, as consumers are asked to rate an ever larger share of their daily transactions [3]. Medallia's own 2026 research found that 66% of experience practitioners believe customer journeys are improving while only 17% of consumers agree, and that more than half of consumers would prefer companies infer their satisfaction from behavior rather than ask them directly [4]. For a company built around surveys, that creates an obvious problem. Medallia has responded by pushing further into behavioral data, conversations, and other forms of passive feedback, which is where many of its acquisitions have taken the platform. 

Borge Hald and Amy Pressman, a married pair of management consultants who advised Fortune 500 executives on customer relationships, were frequent travelers who kept struggling to get a smoke-free room at a hotel they stayed in regularly. They were unhappy enough to steer friends elsewhere, and the hotel's management had no idea. The problem was not that the hotel lacked customer feedback but that the people running the property had no reliable way to see it and act on it. Hald and Pressman set out to build that system.

They incorporated the company as Berrypick in July 2000 and renamed it Medallia in May 2001. Hald and Pressman bootstrapped the company for roughly a decade, reaching profitability by 2003 before growth spending pushed it back into losses [5]. In 2012, Sequoia Capital came in with a first check of roughly $35mm and went on to lead every subsequent VC round, building a position of approximately 41% by the time it went public, one of the firm's largest ever at an IPO [6]. A $150mm round in 2015 valued the company at roughly $1.25bn, and a $70mm Series F in February 2019 took that to roughly $2.4bn. Hald ran the company as CEO from founding until 2018, when both founders stepped back from operating roles and the board hired Leslie Stretch, a software industry veteran, specifically to take the company public [7].

Medallia priced its IPO at $21/share in July 2019, above a $16 to $18 range, raising roughly $326mm, and closed its first session at $37.05 for a gain of about 76% and a market value above $4.5bn [8].

As a public company, Medallia spent aggressively on capability. It bought Strikedeck and CoolaData in 2019, then Voci Technologies, LivingLens, Crowdicity, and Zingle through 2020, then Sense360 and Decibel Insight in 2021 [1]. The series of acquisitions totaled roughly $275mm, and they all pushed Medallia in the same direction, each adding a signal type that surveys could not capture. Voci brought speech, Decibel brought digital behavior, LivingLens brought video. Medallia was buying its way out of being a survey vendor. 

Then, in July 2021, Thoma Bravo took Medallia private at $34.00 per share in cash, a premium of roughly 20% to the unaffected price and roughly 29% to the unaffected 30-day average, valuing the company at approximately $6.4bn including debt and roughly $5.6bn in equity. The merger agreement carried a 40-day go-shop, which expired on September 4, 2021 without producing a competing bid. Shareholders approved on October 14, and the deal closed on October 29, 2021 [9]. 

Thoma Bravo funded the October 2021 buyout with a roughly $1.8bn 1L term loan and a $100mm revolver, against an equity check of approximately $5.1bn split between the firm's own funds and a co-invest vehicle. The term loan priced at S+650 and matured in October 2028, with 600bps of the spread payable in kind rather than in cash; a $100mm RCF at S+400 matured a year earlier. Debt represented just 28% of the $6.4bn purchase price, and equity sat nearly three dollars deep for every dollar of debt.

Medallia was not generating meaningful earnings at the time of the buyout. Its last full year as a public company produced roughly $477mm of revenue and ~$27mm of adj. EBITDA, a margin under 6%, and management was guiding to roughly break-even adj. EBITDA the following year [2]. With no meaningful EBITDA, lenders sized the loan to annual recurring revenue instead. Against $520mm of total run-rate revenue at close, the $6.4bn price implied approximately 12.3x. Against ARR, which excludes non-recurring professional services, it implied nearly 14x. 

Lending Against Revenue, Not Earnings:

It is worth explaining what an ARR loan actually is. Conventional leveraged finance sizes debt against EBITDA because EBITDA approximates the cash available to service it. High-growth software can break that logic, as customer acquisition costs are incurred upfront while subscription revenue arrives over time, so a healthy and growing SaaS company can still report little or no earnings. Lenders, including private credit, therefore began sizing some software loans against annual recurring revenue, focusing on whether the recurring base was large, sticky, and growing. Market convention put conservative recurring revenue loans at roughly 1.0 to 1.5x ARR, with private credit stretching toward 2.0x and the most aggressive 2021 deals reaching 2.5 to 3.0x. Medallia's $1.8bn sat against ARR of roughly $450mm to $470mm at close, or approximately 3.8 to 4.0x. Even including professional services, which was not recurring and carried almost no gross margin, the multiple was roughly 3.4x. Medallia was at or beyond the outer edge of the market on either measure.

The cleanest way to think about an ARR multiple is to translate it into the leverage it eventually becomes. ARR lending is ultimately a shortcut for normalized earnings: the lender is assuming that recurring revenue will grow and eventually convert into enough EBITDA to support a conventional debt load. At a 33% EBITDA margin, every turn of ARR debt becomes roughly three turns of EBITDA leverage, so 3x ARR is about 9x EBITDA. Growth is the other lever. To illustrate for Medallia, assume ARR doubles over five years, a roughly 15% CAGR, from about $460mm to about $920mm, while margins reach 33%. That produces roughly $300mm of EBITDA against $1.8bn of debt, or about 6x leverage, an ordinary figure for a sponsor-backed software company. That scenario best illustrates how to think about ARR to EBITDA conversion, and it also shows what Medallia’s structure required: continued strong growth and material margin expansion, delivered together.  

Lenders underwriting on ARR follow a handful of screens to judge whether growth justified near-term losses. Medallia performed poorly on several of them. The best-known screen is the Rule of 40, which adds revenue growth and profit margin and treats 40 as the benchmark for a healthy balance between the two. Medallia grew 19% in its last public year at a non-GAAP operating margin of roughly 2%, for a score of approximately 21, down from roughly 28 the year before. Sales efficiency was weak as well. Medallia added roughly $70mm of subscription revenue in fiscal 2021 after spending roughly $181mm on sales and marketing the year before, implying a magic number of approximately 0.4 against a convention that treats 0.75 as efficient and 1.0 as excellent [2]. Medallia was paying more than two dollars of sales cost for every dollar of new recurring revenue. The third screen is net revenue retention, which tracks a cohort of existing customers over a year and nets upsells against downgrades and churn. Above 100%, the installed base grows on its own; below it, the company has to win new business simply to stand still. Medallia's was 126% at the end of fiscal 2018 and 116% a year later prior to the IPO. Both are strong figures, and 116% still places a business comfortably among healthy enterprise software companies and well above the average stressed PE-backed software company. The direction, however, was more concerning than the absolute level. Net revenue retention is one of the clearest indicators of whether an ARR loan will grow into itself, and Medallia’s had already fallen ten points in a single year, two years before the structure was underwritten. 

The term loan’s PIK feature followed the same logic. A borrower without meaningful earnings cannot support a fully cash-pay coupon, so lenders initially allowed 600bps of the S+650 spread to accrue to principal instead, before stepping down to a 400bps PIK rate in 2023. The deferral was meant to bridge the period before the company reached a defined EBITDA threshold and converted to conventional cash pay. That structure works if the borrower grows into the debt. If it does not, the balance compounds while the underlying business fails to catch up, and the feature that preserves liquidity early on becomes part of the problem later.

Given those metrics, the obvious question is how lenders got comfortable with the structure. The likely answer is that the deal was struck at the top of the software valuation cycle, when public comparables supported double-digit revenue multiples, and the working assumption across the asset class was that enterprise software growth would continue more or less indefinitely. Another part of it is the sponsor. Thoma Bravo has built a reputation as the most operationally capable buyer in software, and a TB deal carried a premium in the credit market as well as the equity market, which is to say lenders were potentially getting conviction from the firm's stellar track record in addition to their underwriting of Medallia. The last part of it was the cushion. Roughly $5.1bn of equity sat beneath the loan, and with a lender detaching below 30% of TEV, enterprise value could fall by more than 70% before lenders faced any impairment. In our view, that combination is what made an otherwise aggressive underwrite look defensible in 2021. 

To be successful, the structure required four things to be true. It required ARR to keep growing. It required that growth to convert into margin, since the loan carried performance targets that would have flipped it from ARR-based to conventional EBITDA-based treatment [11]. It required the 400 basis points of PIK to be a bridge across the margin-building years rather than a permanent feature. And finally, it required rates to stay roughly where they were, because the entire structure floated.

The Rollup Continues:

All four assumptions ended up breaking, but before any of that, Thoma Bravo went shopping.

Within ten months of closing, Medallia had completed multiple acquisitions, including Thunderhead, a UK journey orchestration vendor, and Mindful, a contact center callback business acquired from Alpine Investors. The acquisitions pushed Medallia further in the direction it was already heading: away from simply measuring customer experience and toward acting on it in real time. Thunderhead and Mindful added orchestration capabilities on top of the existing survey, speech, and digital analytics platform, expanding the product while also increasing the integration burden. That work unfolded amid repeated leadership changes, with Leslie Stretch, the software veteran the board had hired to take Medallia public, succeeded by Joe Tyrrell, an interim period under Mike Lipps, and Mark Bishof taking over in 2025 [1]. Turnover at the top is rarely a coincidence, and in PE-backed companies, it often is a leading indicator of operational trouble. While the rollup’s individual purchase prices were not public, the financing footprint was. In the second quarter of 2023, Medallia drew incremental term debt alongside the original facility. Blackstone's combined position across its two lending vehicles jumped $376mm in that single quarter, and since Blackstone held roughly 54% of the original loan, the draw was likely closer to $700mm. By then, roughly $150mm of PIK had already accrued to the original loan, putting total funded debt at approximately $2.65bn less than two years after a $1.8bn buyout.

With the platform assembled, the margin for error in the original underwriting had narrowed considerably. 

The Retention Problem:

The operational problem came first. Growth slowed as Qualtrics pressed upmarket from the mid-market with a self-service product that deployed faster and cost less, while contact center vendors began packaging conversational analytics into the telephony purchase itself, quietly removing Medallia from selection processes that never happened. The damage was more visible in retention than in reported ARR. ARR did grow under sponsor ownership, from roughly $460mm at close to approximately $600mm by 2026, but that is growth of about 5% a year against a structure underwritten to something closer to 20%. Pari Passu sources describe the retention picture as where the business actually came apart, driven by a run of large logo losses.