Run Customer Discovery Like a Revenue Motion

The headcount data is the part nobody wants to print, so I will print it. Meta, Amazon, Microsoft, and Google all cut user research teams. Layoffs.fyi clocked 152,922 tech layoffs in 2024 and another 122,549 in 2025, and the dedicated UX research roles were over-represented in the cuts. By 2025, 49% of researchers said they felt pessimistic about the future of the field, up 26 points from the year before.

Meanwhile, the work did not disappear. It got routed.

A recent customer-research hiring analysis lays it out: 71% of organizations now have "people who do research" who are not dedicated researchers. 84% allow non-researchers to run studies. The product manager who would have filed a UXR ticket in 2024 runs the interview herself in 2026, with an AI moderator and an AI synthesizer doing the work that used to take a researcher three weeks.

Customer research is no longer a job title. It is a capability, and that capability is now spread across product, customer success, and marketing.

What Changed Between 2024 and 2026

Three things, in this order:

• AI moderation, transcription, and synthesis stopped being experiments and became defaults. 84% of academic researchers reported using AI tools in 2025, up from 57% the year before. 85% said it improved efficiency. The tooling crossed the line from "novel" to "the way we do it now."
• Pure-execution research roles shrank fastest. Senior strategic researchers remain in demand. The role that disappeared was the one whose job was running interviews, transcribing, and writing up findings. That is exactly the work the AI does cheaper and faster.
• Demand for research went up. The product, CS, and marketing teams did not stop needing answers about customers. They started getting them themselves, with the AI doing the moderation work.

Judd Antin, who ran research at Airbnb and Meta, has been writing about this for two years. His framing: the researchers who moved from execution to enablement are thriving. The ones who held on to execution work are the ones who got cut.

Who Actually Runs Customer Discovery Now

Three roles, in roughly equal share:

• Product managers running discovery on the features they own. The unit of work is the interview that proves or kills a hypothesis before engineering touches it.
• Customer success managers running renewal-risk and expansion conversations. The unit of work is the at-risk-account interview that the CSM can react to inside a 7-day window.
• Marketers running positioning, messaging, and ICP tests. The unit of work is the persona panel that validates a campaign before the budget is spent.

None of them are researchers. All of them need the moderator, the transcript, the signal scoring, and the synthesized answer in the same day, not three weeks later when the next quarter has started.

Run It Like a Revenue Motion

The thing every operator running discovery has in common: they think in cycles, not in studies. A study takes six weeks and produces a report. A revenue motion runs every week and produces a decision. The operator who got handed customer research wants the second model, not the first.

That changes what the research stack has to do:

• Self-serve setup. The PM books an interview slot the way she books a calendar event. No project plan, no kickoff call, no formal review process.
• AI-moderated interviews. Emma, our AI interviewer, runs the conversation. The PM does not get on the call.
• Same-day synthesis. The transcript is tagged with one of six expression signals: Sad, Angry, Confrontational, Neutral, Cheerful, or Enthusiastic. It is summarized and routed to the operator's tool of choice before she opens her laptop the next morning. These labels describe how a response is expressed in the conversation, not what a person privately feels.
• Patterns across conversations, not single-interview anecdotes. One interview is a data point. Twenty is a pattern. The system has to aggregate before the operator has to.

Read more about how we handle data, retention, and privacy in our Trust & Compliance Center.

The Wedge for Operators

The wedge is not "we built a better tool for researchers." The wedge is "we built the customer discovery layer your product, CS, and marketing teams already wish they had."

If you are running discovery without a researcher on the team, the question is not whether to hire one. The question is whether the system underneath you is treating you like an operator who needs an answer this week, or like a researcher who has six weeks to wait.

If the answer is the second one, start a free 3-minute Live Test Drive and hear what same-day discovery sounds like.