Automated User Research Pipeline
Wire sales-call keywords to Slack to email to your calendar so customer interviews book themselves.
- Difficulty
- Moderate
- Time to result
- ~days to results
- Steps
- 6
- Confidence
- 93%
PMs stop talking to customers because sourcing them is friction. Kevin Yien's answer is to treat the sales team as a research team and build a plumbing chain — call-recording keyword alerts fire into Slack, an automation grabs the customer's email, sends them a templated note with a booking link, and interviews land on your calendar without you running a program. The principle underneath: a PM must never settle for looking through bent glass — reports, research summaries, and sales relays are all pre-processed by someone else.
Origin
Kevin Yien credits Beth Hills, a PM he hired at Mutiny, as 'the queen of automating customer research' — she built the Gong/Slack/Zapier system he describes. The complementary always-on intake tactic (a homepage popup offering to book feedback calls straight onto the PM's calendar) he attributes to Teresa Torres.
Core principles
- 01PMs need direct exposure to raw material — never accept research reports, sales relays, or market research as a substitute.
- 02Your sales team is a research team; call recordings are an unmined transcript corpus of your customers' actual language.
- 03Effort of setup is not a valid excuse for not talking to customers — automate the sourcing, not the conversation.
- 04Talking to customers does not tell you what to build; it updates your mental model of the customer, which then informs what to build.
- 05'I've already done 100 interviews, I know them' is a tempting lie — customers' worlds change continuously.
How to run it
- 1
Fix the structural blocker first
If a customer success manager or account owner forbids PMs from speaking to customers, that is a structural defect. Nothing downstream works until it is resolved.
Watch out Do not build automation on top of a policy that bans you from the calls it books.
- 2
Set keyword alerts on your call recordings
In Gong (or equivalent call-recording tool), set filters and alerts on the specific terms, phrases, competitors, or product areas you own — e.g. 'point of sale'.
Pro tip Every PM can name the handful of terms that map to their surface area — that list is the whole configuration.
- 3
Pipe the hits into Slack
Use the call-tool's Slack integration so each hit posts the transcript excerpt where the term was mentioned, along with the user and customer names.
Pro tip The excerpt alone is useful qualitative signal even on days you book nobody.
- 4
Automate outreach off the Slack post
Use Zapier (or similar) to trigger on each new Slack post in that channel, pull the customer's email field, and send a templated research-invite email via a tool like Customer.io.
- 5
Drop a booking link in the template
Embed a Calendly link reserved specifically for user research inside the email template so interested customers self-schedule directly onto your calendar.
Pro tip Pair this with an always-on funnel — userinterviews.com for B2B sourcing to your exact ICP, and a homepage feedback popup that books straight into your calendar.
- 6
Cover the channels sales doesn't touch
Find where people already talk about you — Reddit, X, forums, your support forum — and set up alerting there too, with a custom script written with an engineer if no off-the-shelf tool fits.
Watch out If there is nowhere at all that people discuss your product, that is itself a signal about product-market fit.
In the wild
Set 'point of sale' as a Gong term. A prospect on a sales call mentions it; Gong's Slack integration posts the transcript excerpt plus the customer name and contact. A Zap fires on that Slack post, pulls the email, and Customer.io sends a templated research invite containing a Calendly link.
→ Interviews with exactly the customers who just raised your product area appear on your calendar automatically, with zero recruiting work per interview.
Yien contrasts reading a report about a line cook with standing next to one. The report cannot carry the ancillary detail of the job — the pace, the flinches, the workarounds — that reshapes what you build.
→ Direct exposure updates the PM's mental model in ways no processed artifact can, which is why the pipeline exists to maximize direct contact rather than summaries.
Common mistakes
Consuming research instead of doing it
Treating UXR reports, sales anecdotes, or market research as your customer input means every insight has already been processed by someone else. Yien calls this looking through bent glass.
Deciding you already know your users
After enough interviews it feels safe to run on intuition and skip the next one. But customers' lives and contexts keep changing, and it is those micro-changes that determine what they'll need next.
Using setup effort as the excuse
'I don't have time to run a program' is precisely why the pipeline is worth an afternoon of wiring — the automation removes the recurring cost that kills the habit.
Is it for you?
Best for
B2B product managers with a sales team, call-recording tooling, and a chronic gap between how often they should and do talk to customers.
Not ideal for
B2C or self-serve products with no sales calls to mine, or organizations where PMs are structurally barred from customer contact (fix that first).
From the transcript
“it's been processed by someone and PM need direct exposure to raw material end of story”
“your sales team is a research team and if you don't view them that way you are missing out on half the value”
“gong has an integration with slack you set up this alert it posts to slack the excerpt of the transcript where it was mentioned along…”
From the episode
Unorthodox PM wisdom: Automating user insights, unselling job candidates, logging every decision, more
Kevin Yien (Stripe, Square, Mutiny)