Alfred Wahlforss was working out of choices. His startup, Listen Labs, wanted to rent over 100 engineers, however competing in opposition to Mark Zuckerberg’s $100 million offers appeared not possible. So he spent $5,000 — a fifth of his advertising finances — on a billboard in San Francisco displaying what appeared like gibberish: 5 strings of random numbers.
The numbers have been really AI tokens. Decoded, they led to a coding problem: construct an algorithm to behave as a digital bouncer at Berghain, the Berlin nightclub well-known for rejecting practically everybody on the door. Inside days, hundreds tried the puzzle. 430 cracked it. Some obtained employed. The winner flew to Berlin, all bills paid.
That unconventional method has now attracted $69 million in Collection B funding, led by Ribbit Capital with participation from Evantic and current buyers Sequoia Capital, Conviction, and Pear VC. The spherical values Pay attention Labs at $500 million and brings its whole capital to $100 million. In 9 months since launch, the corporate has grown annualized income by 15x to eight figures and performed over a million AI-powered interviews.
“If you obsess over clients, every thing else follows,” Wahlforss mentioned in an interview with VentureBeat. “Groups that use Pay attention carry the shopper into each determination, from advertising to product, and when the shopper is delighted, everyone seems to be.”
Why conventional market analysis is damaged, and what Pay attention Labs is constructing to repair it
Pay attention’s AI researcher finds contributors, conducts in-depth interviews, and delivers actionable insights in hours, not weeks. The platform replaces the normal alternative between quantitative surveys — which give statistical precision however miss nuance—and qualitative interviews, which ship depth however can’t scale.
Wahlforss defined the limitation of current approaches: “Basically surveys provide you with false precision as a result of folks find yourself answering the identical query… You’ll be able to’t get the outliers. Persons are really not sincere on surveys.” The choice, one-on-one human interviews, “provides you plenty of depth. You’ll be able to ask observe up questions. You’ll be able to form of double examine if they really know what they’re speaking about. And the issue is you possibly can’t scale that.”
The platform works in 4 steps: customers create a examine with AI help, Pay attention recruits contributors from its world community of 30 million folks, an AI moderator conducts in-depth interviews with follow-up questions, and outcomes are packaged into executive-ready studies together with key themes, spotlight reels, and slide decks.
What distinguishes Pay attention’s method is its use of open-ended video conversations moderately than multiple-choice varieties. “In a survey, you possibly can form of guess what it’s best to reply, and you’ve got 4 choices,” Wahlforss mentioned. “Oh, they in all probability need me to purchase excessive revenue. Let me click on on that button versus an open ended response. It simply generates way more honesty.”
The soiled secret of the $140 billion market analysis trade: rampant fraud
Listen finds and qualifies the proper contributors in its world community of 30 million folks. However constructing that panel required confronting what Wahlforss referred to as “probably the most surprising issues that we have discovered once we entered this trade”—rampant fraud.
“Basically, there is a monetary transaction concerned, which implies there shall be unhealthy gamers,” he defined. “We really had a number of the largest corporations, a few of them have billions in income, ship us individuals who declare to be form of enterprise consumers to our platform and our system instantly detected, like, fraud, fraud, fraud, fraud, fraud.”
The corporate constructed what it calls a “high quality guard” that cross-references LinkedIn profiles with video responses to confirm id, checks consistency throughout how contributors reply questions, and flags suspicious patterns. The end result, in keeping with Wahlforss: “Individuals discuss 3 times extra. They are much extra sincere once they speak about delicate matters like politics and psychological well being.”
Emeritus, a web-based training firm that makes use of Pay attention, reported that roughly 20% of survey responses beforehand fell into the fraudulent or low-quality class. With Pay attention, they decreased this to nearly zero. “We didn’t have to exchange any responses due to fraud or gibberish data,” mentioned Gabrielli Tiburi, Assistant Supervisor of Buyer Insights at Emeritus.
How Microsoft, Sweetgreen, and Chubbies are utilizing AI interviews to construct higher merchandise
The pace benefit has confirmed central to Pay attention’s pitch. Conventional buyer analysis at Microsoft might take 4 to 6 weeks to generate insights. “By the point we get to them, both the choice has been made or we lose out on the chance to truly affect it,” mentioned Romani Patel, Senior Analysis Supervisor at Microsoft.
With Pay attention, Microsoft can now get insights in days, and in lots of instances, inside hours.
The platform has already powered a number of high-profile initiatives. Microsoft used Pay attention Labs to gather world buyer tales for its fiftieth anniversary celebration. “We wished customers to share how Copilot is empowering them to carry their finest self ahead,” Patel mentioned, “and we have been in a position to gather these consumer video tales inside a day.” Historically, that form of work would have taken six to eight weeks.
Simple Modern, an Oklahoma-based drinkware firm, used Pay attention to check a brand new product idea. The method took about an hour to write down questions, an hour to launch the examine, and a pair of.5 hours to obtain suggestions from 120 folks throughout the nation. “We went from ‘Ought to we even have this product?’ to ‘How ought to we launch it?'” mentioned Chris Hoyle, the corporate’s Chief Advertising and marketing Officer.
Chubbies, the shorts model, achieved a 24x enhance in youth analysis participation—rising from 5 to 120 contributors — through the use of Pay attention to beat the scheduling challenges of conventional focus teams with youngsters. “There’s faculty, sports activities, dinner, and homework,” defined Lauren Neville, Director of Insights and Innovation. “I needed to discover a option to hear from them that match into their schedules.”
The corporate additionally found product points by way of AI interviews which may have gone undetected in any other case. Wahlforss described how the AI “by way of conversations, realized there have been like points with the the youngsters quick line, and determined to, like, interview a whole bunch of youngsters. And I perceive that there have been points within the liner of the shorts and that they have been, like, scratchy, quote, unquote, in keeping with the folks interviewed.” The redesigned product turned “a blockbuster hit.”
The Jevons paradox explains why cheaper analysis creates extra demand, not much less
Listen Labs is getting into a large however fragmented market. Wahlforss cited analysis from Andreessen Horowitz estimating the market analysis trade at roughly $140 billion annually, populated by legacy gamers — some with greater than a billion {dollars} in income — that he believes are weak to disruption.
“There are very a lot current finances strains that we’re changing,” Wahlforss mentioned. “Why we’re changing them is that one, they’re tremendous expensive. Two, they’re form of caught on this outdated paradigm of selecting between a survey or interview, they usually additionally take months to work with.”
However the extra intriguing dynamic could also be that AI-powered analysis would not simply substitute current spending — it creates new demand. Wahlforss invoked the Jevons paradox, an financial precept that happens when technological developments make a useful resource extra environment friendly to make use of, however elevated effectivity results in elevated general consumption moderately than decreased consumption.
“What I’ve seen is that as one thing will get cheaper, you do not want much less of it. You need extra of it,” Wahlforss defined. “There’s infinite demand for buyer understanding. So the researchers on the workforce can do an order of magnitude extra analysis, and likewise different individuals who weren’t researchers earlier than can now try this as a part of their job.”
Contained in the elite engineering workforce that constructed Pay attention Labs earlier than they’d a working bathroom
Listen Labs traces its origins to a client app that Wahlforss and his co-founder constructed after assembly at Harvard. “We constructed this client app that obtained 20,000 downloads in in the future,” Wahlforss recalled. “We had all these customers, and we have been pondering like, okay, what can we do to get to know them higher? And we constructed this prototype of what Pay attention is at the moment.”
The founding workforce brings an uncommon pedigree. Wahlforss’s co-founder “was the nationwide champion in aggressive programming in Germany, and he labored at Tesla Autopilot.” The corporate claims that 30% of its engineering workforce are medalists from the International Olympiad in Informatics — the identical competitors that produced the founders of Cognition, the AI coding startup.
The Berghain billboard stunt generated roughly 5 million views throughout social media, in keeping with Wahlforss. It mirrored the depth of the expertise conflict within the Bay Space.
“We needed to do this stuff as a result of a few of our, like early staff, joined the corporate earlier than we had a working bathroom,” he mentioned. “However now we fastened that scenario.”
The corporate grew from 5 to 40 staff in 2024 and plans to succeed in 150 this yr. It hires engineers for non-engineering roles throughout advertising, development, and operations — a guess that within the AI period, technical fluency issues in every single place.
Artificial clients and automatic choices: what Pay attention Labs is constructing subsequent
Wahlforss outlined an formidable product roadmap that pushes into extra speculative territory. The corporate is constructing “the flexibility to simulate your clients, so you possibly can take all of these interviews we have completed, after which extrapolate primarily based on that and create artificial customers or simulated consumer voices.”
Past simulation, Pay attention goals to allow automated motion primarily based on analysis findings. “Are you able to not simply make suggestions, but additionally create spawn brokers to both change issues in code or some buyer churns? Are you able to give them a reduction and attempt to carry them again?”
Wahlforss acknowledged the moral implications. “Clearly, as you mentioned, there’s form of moral issues there. Of like, automated determination making general could be unhealthy, however we can have appreciable guardrails to be sure that the businesses are at all times within the loop.”
The corporate already handles delicate information with care. “We do not prepare on any of the info,” Wahlforss mentioned. “We will even scrub any delicate PII robotically so the mannequin can detect that. And there are occasions when, for instance, you’re employed with buyers, the place when you unintentionally point out one thing that could possibly be materials, non public data, the AI can really detect that and take away any data like that.”
How AI might reshape the way forward for product improvement
Maybe essentially the most provocative implication of Pay attention’s mannequin is the way it might reshape product improvement itself. Wahlforss described a buyer — an Australian startup — that has adopted what quantities to a steady suggestions loop.
“They’re primarily based in Australia, in order that they’re coding throughout the day, after which of their evening, they’re releasing a Pay attention examine with an American viewers. Pay attention validates no matter they constructed throughout the day, they usually get suggestions on that. They’ll then plug that suggestions immediately into coding instruments like Claude Code and iterate.”
The imaginative and prescient extends Y Combinator’s well-known dictum — “write code, talk to users” — into an automatic cycle. “Write code is now getting automated. And I feel like discuss to customers shall be as nicely, and you will have this type of infinite loop the place you can begin to ship this actually wonderful product, nearly form of autonomously.”
Whether or not that imaginative and prescient materializes will depend on elements past Pay attention’s management — the continued enchancment of AI fashions, enterprise willingness to belief automated analysis, and whether or not pace actually correlates with higher merchandise. A 2024 MIT study discovered that 95% of AI pilots fail to maneuver into manufacturing, a statistic Wahlforss cited as the rationale he emphasizes high quality over demos.
“I am consistently have to emphasise like, let’s be sure the standard is there and the small print are proper,” he mentioned.
However the firm’s development suggests urge for food for the experiment. Microsoft’s Patel mentioned Pay attention has “eliminated the drudgery of analysis and introduced the enjoyable and pleasure again into my work.” Chubbies is now pushing its founder to provide everybody within the firm a login. Sling Cash, a stablecoin funds startup, can create a survey in ten minutes and obtain outcomes the identical day.
“It is a whole recreation changer,” mentioned Ali Romero, Sling Cash’s advertising supervisor.
Wahlforss has a unique phrase for what he is constructing. When requested in regards to the rigidity between pace and rigor — the long-held perception that transferring quick means reducing corners — he cited Nat Friedman, the previous GitHub CEO and Pay attention investor, who retains an inventory of one-liners on his web site.
Certainly one of them: “Gradual is pretend.”
It is an aggressive declare for an trade constructed on methodological warning. However Listen Labs is betting that within the AI period, the businesses that hear quickest would be the ones that win. The one query is whether or not clients will discuss again.
