Nobody books the call
You email ten users a calendar link. One replies. Zero show. The personalized feedback email? Crickets — even from the power users.
Code got cheap. Knowing what to build didn’t.
Ask your users why, while they can still tell you.
UserTold.ai runs a voice interview about the moment a real user gets stuck, in the same session — the quote, the screen, and the page path stay together. You and your coding agent review the evidence before it becomes work.
Voice + screen + navigationMCP · CLI · REST$0.25 / recorded minute
REC · interview in progress — trial user, day 3study: onboarding-friction
navigation/pricing/settings/account← back ×2/settings/accountscreenvoice
01:12 observed02:22 said02:41 asked why
Observed01:12
Opens account settings. Scans. Goes back. Returns and scans again.
/settings/account → back ×2
Said02:22
“I expected billing under account settings. I went back twice and still couldn’t find it.”
Asked why — 19 seconds later02:41
The debrief asks about the exact moment it just watched — while the user can still explain what they expected and where they looked.
Evidenceconfidence 0.91
Quote, replay, and page path stay attached.
Quote, replay, and page path stay attached.
Draft workClarify where billing settings live — 3 linked moments from 3 users.
✓ Pushed to LinearVerified by a human or project-aware agent first. Sources travel with the issue.
● WatchingWhen the Linear issue completes, new interviews are checked for the same evidence coming back.
The gap
You have the users. You keep missing the moment.
Analytics say what, never why
The funnel shows where they left. Session replays pile up that nobody has time to watch. The reason walks away untold.
Feedback arrives without context
A churn survey says “too complicated.” Which screen? Which task? By the time you can ask, they’re gone.
UserTold asks at the only moment users can actually answer: while it’s happening, inside your product.
How it works
From a real moment to a merged fix
Each step produces an artifact you can open. Nothing in the chain asks to be taken on faith.
→ a widget in your product
Embed once
One script tag. Choose the posture per study: silent observation, a guided task, or an open conversation.
→ a recorded session
Meet users in the moment
An AI interviewer talks with users in their language, or quietly captures real usage — voice, screen, and navigation.
→ source-linked evidence
Get evidence, not opinions
Struggling moments, desired outcomes, workarounds — each with the quote, the timestamp, the page path, and the replay.
→ a verified issue
Ship it, then watch
Related evidence groups into draft work. You verify and push to Linear or GitHub with sources attached. When a linked Linear issue completes, UserTold watches whether the evidence comes back.
Evidence-first
Open the evidence. Don’t trust the summary.
AI research tools hand you conclusions. UserTold hands you conclusions with the receipts attached — every finding stays linked to what a person actually said and did, so you can check the interpretation before acting on it.
Evidence · struggling_momentses_xyz789 · 02:22
“I tried this flow three times and still cannot find where to change billing.”
Replay 02:22/checkout/step-3Confidence 0.912 similar moments
The same evidence, as your agent sees it
{
"signal_type": "struggling_moment",
"quote": "I tried this flow three times and still cannot find where to change billing.",
"confidence": 0.91,
"intensity": 0.8,
"interviewRef": "ses_xyz789",
"timestamp_ms": 142300,
"page_url": "/checkout/step-3"
}For coding agents
Your agent can ship. Now it can find out what’s worth shipping.
UserTold is MCP-, CLI-, and API-first. A coding agent can design a study, wait for real users to take part, and read the evidence — in structured JSON, with schemas published at discovery — before it proposes a fix.
MCPmcp.usertold.ai/mcp — OAuth, tools for studies.* evidence.* work.* interviews.*
CLIusertold study create --title "My Product Study" --activate --format json
HandoffVerified work lands in Linear or GitHub with quotes and source moments attached.
See how agents use UserTold →# an agent-run study, end to end
read usertold://projects
call studies.create
… real users interview in your product …
call evidence.list
call work.create_from_evidence
… human or agent verifies the grouping …
call work.push → Linear UT-214On purpose
Three things UserTold refuses to do
Recruit strangers
No panels, no rented respondents. It interviews people already using your product. If nobody shows up, it says so — it doesn’t fake a sample.
Interrupt a struggle
During observation, stuckness is evidence. The interviewer doesn’t jump in to rescue, hint, or steer — it asks why afterward, when the moment is on record.
Fake certainty
What a user said, what was observed, and what the model inferred are never blended. Every finding carries its source and a confidence score.
Pricing
Priced like an API, not a research department
Pay for recorded interview minutes. No subscription, no seats, no per-interview minimum — billed on exact recorded seconds, itemized in Billing.
Managed AI
Inference included$0.25 / recorded minute
One predictable UserTold charge. We operate and pay for the interview AI.
5 min $1.2520 min $5.0060 min $15.00
BYOK
Your OpenAI key$0.15 / recorded minute
UserTold charges the platform fee. OpenAI bills inference directly to your provider account.
5 min $0.7520 min $3.0060 min $9.00
Questions
Before you ask
From your product. UserTold interviews users you can already reach — it does not recruit, and it does not guarantee responses. In-product timing is why response rates beat cold calendar links: the user is already there.
The next thing you ship should be something a user told you.
Embed the widget, run your first study on yourself in five minutes, and read the evidence it produces before you point it at real users.