Customer research that starts
with what people actually did, not what they say
What a customer says in an interview and what the same customer did on your site often disagree, and the disagreement is usually where the real insight lives. We run structured interviews alongside session recordings, support logs and CRM notes, so a finding is backed by behavior, not just a quote that sounded convincing in the room.
Where the money leaks today
Customer interviews are useful and routinely misleading in the same breath, because people are honest about their intentions and unreliable narrators of their own behavior. A customer will confidently explain why they almost did not buy, and the real reason, visible in a session recording, is often something smaller and more fixable than the story they tell afterward: a form that silently rejected their input, a step that took three tries, a sign-in flow that failed without a clear error.
We found exactly this pattern auditing an app with strong reviews and growing installs: thirty-nine people a month were failing a specific sign-in step, a concrete, fixable number that no interview alone would have surfaced, because the people who hit that wall mostly just left rather than calling to complain. Day-1 retention sat at 17.8%, day-7 at 6.5%, numbers that gave later interview findings about onboarding confusion real weight instead of leaving them as anecdotes.
Research that only interviews people, without checking what they actually did, risks confidently acting on the wrong story.
What we do
We build an interview guide around the specific open questions your business actually has, not a generic discovery template, and run 5 to 10 structured interviews with real customers or, often more revealing, people who considered buying and did not. Every interview gets cross-checked against whatever behavioral data exists: session recordings, support tickets, CRM notes, so a theme that emerges in conversation can be weighed against what actually happened at scale.
The synthesis we deliver is organized by theme and explicitly flags contradictions between what people said and what the data showed, because that gap is usually the most useful finding in the whole project. Recommendations are tied to specific funnel or product decisions you are actually facing, not a persona slide that looks good in a deck and changes nothing.
What we need from you
A way to reach real customers or recently lost prospects willing to talk, and access to session recordings, support logs or CRM notes if they exist. Tell us the specific decision this research needs to inform; a focused question produces a far more useful round of interviews than an open-ended “learn about our customers.”
How we measure
Whether the recommendations, once acted on, moved the specific metric they targeted, checked against the behavioral data the same way the original research was.
Price and timeline
| Option | Price | What it covers | Timeline |
|---|---|---|---|
| Launch or audit | from $1,200 | 5 to 10 interviews, behavioral cross-check, written synthesis | 2 to 3 weeks |
| Monthly management | from $900 / month | Ongoing rolling interviews tied to product or funnel changes | monthly, no lock-in |
| Full control, handover to your team | from $2,000 | Full interview guide, methodology and training for your team | 3 to 4 weeks |
Related
Customer research pairs naturally with a marketing funnel audit and informs positioning and messaging workshop work. The full build is on the analytics service page. Real examples: the Thailand store audit and the trading app funnel audit from session recordings.
Getting confident answers in interviews that do not match what your data shows? Get in touch and we will cross-check the two.
FAQ
How much does customer research cost?
From $1,200 for a round of 5 to 10 structured interviews cross-checked against behavioral data, with a written synthesis, delivered in 2 to 3 weeks.
How long does a research round take?
2 to 3 weeks, mostly gated by scheduling interviews with real customers; the analysis itself moves faster once interviews are done.
How many customers do we need to interview for this to be useful?
5 to 10 structured interviews usually surface the recurring themes for a focused question; more than that has diminishing returns unless you are deliberately covering very distinct customer segments.
What do you need from us?
Introductions to customers willing to talk, or a way to recruit them, plus access to session recordings, support tickets or CRM notes if those exist, since the behavioral cross-check is where this approach earns its value.
How do you report findings?
A written synthesis organized by theme, explicitly flagging where what customers said and what the data showed disagreed, with recommendations tied to specific funnel or product decisions, not a generic persona slide.