Bring all your customer knowledge together, analyze it fast, and surface evidence-backed insights your team and AI agents can both act on.
Every company runs on a story it tells itself about its customers. The trouble is where that story lives: a half-remembered interview, a slide deck three quarters gone, a Miro board nobody reopens, a Slack thread that scrolled into the dark. Ask most product teams "did we already study this?" and the honest answer is a shrug. Condens, a software company out of Munich, built its entire business on turning that shrug into a two-second search.
The pitch is plain enough to fit on a napkin. Pull all your user research - interviews, usability tests, notes, recordings, screenshots - into one place. Transcribe it. Tag it. Let AI cluster and summarize it. Then keep it, forever, searchable, so the next researcher who asks a question the company already answered gets the answer instead of running the study again. Condens calls the goal moving "from research chaos to a single source of customer insights you can trust." Researchers call it, roughly, sanity.
User research has a strange economics. It is expensive to gather - recruiting participants, running sessions, sitting through hours of video - and almost free to lose. A finding that cost a week of work evaporates the moment the readout meeting ends. Six months later a new hire proposes the exact study that already ran, because the previous one lives on a laptop that has since been wiped.
Condens attacks that decay at both ends. On the front end, it gives researchers a working surface: a split-view where a raw transcript sits beside the tags, highlights, and clusters a researcher builds from it, so a claim in a report is always one click from the sentence a real person actually said. On the back end, it keeps every study as a searchable artifact. The analysis and the archive are the same object, which is the quiet trick - you don't file your insights away after the fact, they were filed the whole time.
That distinction matters more than it sounds. Plenty of tools promise a "repository," but most treat it as a filing cabinet bolted on after the work is done - a place you're supposed to remember to upload things to, which of course nobody does. Condens inverts the order. Because the analysis lives in the same system that stores it, the archive fills itself as a byproduct of doing the actual job. There is no separate step to skip, no export to forget. The insight and its evidence stay welded together, so a stat in a stakeholder deck can always be traced back to the exact clip where a user said it out loud.
Condens is, appropriately for a research tool, a product built by research. Before founding the company in 2018, Alexander Knoll interviewed more than 160 UX researchers around the world about what slowed them down. Only after that did he, Maximilian Hackenschmied, and Matej Svejda start building. Knoll is CEO; Hackenschmied runs product as CPO. The habit stuck: the team uses its own tool internally and ships changes based on what customers report, fast enough that one of them coined an internal hashtag for it - #BeLikeCondens.
There is a tidy irony in a UX research company doing that much upfront research on itself, and it doubles as the founding thesis. The 160 conversations were not a formality; they were the product spec. Researchers described the same recurring pain - hours lost re-watching recordings, insights trapped in one person's head, findings that couldn't survive the trip from analysis to stakeholder. Build to fix those specific complaints, keep talking to the people who have them, and the roadmap more or less writes itself. It is an unglamorous way to build software, and it is exactly why the product feels like it was made by someone who has actually sat through a bad interview at 5pm on a Friday.
There are really three products wearing one coat. The repository is the archive - a searchable home for sessions, artifacts, highlights, and tags. The analysis workspace is where the tagging and affinity-mapping happens. And Condens AI, added as the AI wave broke, sits across both: it suggests tags without creating duplicates, auto-clusters highlights by prompt or sentiment, writes summaries across sessions and tags, and answers questions in plain language. Transcription handles 85+ languages with speaker labels.
The AI angle comes with a line drawn hard in the sand. Condens connects to agents like Claude and ChatGPT so their answers stay grounded in a company's real interviews rather than guesses - and it states plainly that customer data is only ever used to run the service, never to train models for anyone else. In a market where "we added AI" usually means "we fed your stuff into a model," the flex is what Condens won't do.
The customer list reads like an airport departures board: British Airways, Volkswagen, Siemens, Accenture, KPMG, ZEISS, EON, H&M, Cisco, BMO, Just Eat Takeaway, Yelp, Freeletics, Pipedrive, TrueLayer, Moz. These are exactly the organizations with the most research and the most ways to lose it - many teams, many time zones, and compliance officers who care where the data sleeps. Condens answers that with SOC 2 Type 2, GDPR, HIPAA and CCPA compliance, PII anonymization, and a choice of EU or US hosting.
Who actually opens the tool day to day is a broader crowd than the word "researcher" suggests. Dedicated UX researchers are the core, but product managers, designers, and increasingly whole product teams live in the repository too - anyone who needs to know what a customer thinks and would rather not schedule a call to find out. That spread is the point of a shared repository: research stops being one team's private craft and becomes a resource the rest of the organization can pull from. The compliance surface is what makes the enterprise version of that trust possible. When a bank or a carmaker puts hours of recorded customer conversations into a third-party tool, "where does this data live and who can touch it" is not a nice-to-have, and Condens treats it as a first-class part of the product rather than a legal afterthought.
Here is the detail that makes Condens unusual in a category swimming in venture money: it has raised none. The company is bootstrapped and profitable, and treats that as a feature rather than a footnote. Being unfunded means no growth-at-all-costs pressure, so the team stays lean, keeps a flat structure where even leadership works customer support, and can obsess over the product instead of the next round. Students and educators get it free - a bet on the next generation of researchers, and a cheap-to-run one for a business that doesn't answer to a board.
The business model underneath is refreshingly legible. Condens sells subscriptions - a per-seat Lite plan for individual projects, a flat-rate Business plan aimed at small teams, and custom Enterprise deals for the big logos. No usage-based surprises, no free tier engineered to trap you, no confusing credits. You pay for seats, you get the tool. For a team of around a dozen people with roughly 5,000 followers on LinkedIn and a customer roster that punches far above that, the math clearly works, which is the whole argument for staying independent: a profitable niche tool with loyal users is a perfectly good business, even if it never becomes a headline.
Condens plays in the research-repository category alongside better-funded names - Dovetail is the giant, with HeyMarvin (Marvin), Great Question, UserBit, Aurelius and Looppanel filling out the field; the former EnjoyHQ once sat here too. The funded players compete on breadth and marketing budget. Condens competes on focus: a tool built by researchers, for researchers, by a team that doesn't have to chase a hockey-stick chart to keep the lights on.
| Trait | Condens | Typical funded rival |
|---|---|---|
| Funding | Bootstrapped, profitable | VC-backed |
| Focus | UX research only | Often broadening |
| Support | Everyone, incl. founders | Tiered teams |
| Your data → AI training | Never | Varies |
| Free for students | Yes | Varies |
Where does it fit in the market? Squarely in ResearchOps - the plumbing that decides whether a company's research compounds or resets to zero every quarter. As AI makes it trivial to generate answers, the scarce thing becomes trustworthy evidence to generate them from. Condens is a bet that the repository, not the chatbot, is the part that matters.
It is a modest bet, made without fanfare, from a city not famous for software. But the logos keep signing, the product keeps shipping, and somewhere a researcher just found in two seconds what used to take two weeks to redo.