Nobody tells you the hard part
“Talk to your customers” is the most repeated advice in startups and the least specific. Who exactly? How many before it means anything? What do you ask so they stop being polite? The advice is free. The method is not.
Customer research, run at speed
Thirty customer conversations.
Started this week.
You've been meaning to do this for months. We start in days.
Every founder is told to talk to twenty or thirty people before building. Almost nobody does — there's a product to ship and the week is already gone. We run those conversations for you, starting now. One quick call to scope it, recruiting the next day, transcripts hitting your inbox in week one.
- Scoped in one quick call
- Recruiting starts next day
- Transcripts as they land
- Weeks, not quarters
Study specExample
- QUESTION
- Would finance leads at 10–200 person companies pay to replace their reconciliation spreadsheet?
- SAMPLE
- 18 × finance leads · SMB · currently on spreadsheets
- METHOD
- 60-min semi-structured interviews · recorded · transcribed
- OUTPUT
- Transcripts · Tagged findings · Decision memo · Live debrief
- TIMELINE
- Recruiting day 1 · transcripts week 1 · debrief week 3
Yours can start this week
// SECTION: THE GAP001
The evidence you need is in conversations you are not having.
Not because you don't believe in them. Because of what they cost you in time.
Done yourself, it eats a month
Finding twenty of the right people, chasing each one for a slot, running the sessions, then reading it all back — four to six weeks of your calendar, spent on scheduling. So it slips to after launch, which is exactly too late for it to change anything.
You cannot interview your own hypothesis
When the founder asks, people are kind. They say the idea is interesting and they'd probably use it, and you hear a market. A neutral third party gets the version where they tell you what they actually do today, and what they have already refused to pay for.
// SECTION: WHAT_YOU_GET002
What lands in your inbox.
Research you can check. Every conclusion traces back to the sentence that produced it.
SAMPLE::SCREENED01
The right people, screened
Real buyers and users in your category, sourced and screened against criteria we agree with you. Not a panel of professional survey-takers who will answer anything for a voucher.
SOURCED → SCREENED → YOU APPROVE THE LIST
INTERVIEW::STRUCTURED02
Interviews run by someone neutral
Thirty to sixty minute conversations, structured so the same ground gets covered every time, and open enough that people can tell you the thing you didn't know to ask about.
SAME QUESTIONS → COMPARABLE ANSWERS
RECORD::STREAMING03
Transcripts the day they happen
Recordings and full transcripts land as each session finishes — not bundled into a report six weeks later. Everything, not the highlights we liked. You're learning from day one.
SAME-DAY → NOTHING SUMMARISED AWAY
FINDINGS::TAGGED04
Findings with the evidence attached
Themes, not vibes. Each finding carries how many people raised it, which segment they were in, and the quotes underneath it. You can audit our reasoning instead of trusting it.
CLAIM → COUNT → SOURCE QUOTE
MEMO::DECISION05
A decision memo
What we would do next if this were our product, ranked, with the reasoning shown — including what we're unsure about and what it would take to be sure.
RANKED → REASONED → ACTIONABLE
DEBRIEF::LIVE06
A working debrief
A session with your team where we walk it through and stay in the room while you argue about it. Research that gets emailed and never discussed changes nothing.
WE STAY FOR THE ARGUMENT
// SECTION: HOW_IT_WORKS003
Four steps. You are involved in two of them.
The parts that eat your month are the parts we take — and we start the day after we scope.
- 01DAY 0Together
Scope
One quick call. We turn the open question you're carrying around into a research question that can actually be answered, and agree who needs to be interviewed for the answer to count.
- 02DAY 1We do this
Recruit
Sourcing starts the next morning. We screen participants against that definition and you approve the list before anyone is interviewed — so you never get a report built on the wrong people.
- 03WEEK 1We do this
Interview
Conversations start as soon as the list is signed off, and transcripts reach you the day each one happens. You're acting on what you hear in week one, not waiting on a final document.
- 04WEEK 2Together
Synthesise
We separate what people said from what it means, and give you both. Findings, evidence, a recommendation — then a live debrief where your team pushes back on all three.
// SECTION: USE_CASES004
The questions we get hired to answer.
If yours isn't here, it's probably a version of one of them. Ask on the quick call.
“Is this problem real, and expensive enough that someone would pay to remove it?”
The validation study. Run before the engineering spend, so the answer is still cheap to act on.
“What are they spending on this today, and what would make them move?”
What the current solution costs them in money, time, and headcount — and where the switching threshold actually sits.
“Does the way we describe this survive contact with the people it's aimed at?”
Positioning and messaging tested against real buyers, before it's printed on the homepage and the ad budget.
“Why did they actually churn?”
Exit interviews with the people who left. The cancellation-form reason and the real reason are rarely the same sentence.
“What's different about buyers in a segment we've never sold to?”
For teams moving up-market, into a new geography, or into a category where existing customer intuition doesn't transfer.
“Of the six things on the list, which would anyone notice if we shipped it?”
Prioritisation grounded in what users are working around today, rather than in whoever argued hardest in planning.
// SECTION: WHY_US005
Three reasons to hand this to someone else.
01
Speed
This is our full-time job, not the thing you'll get to after the release. Recruiting starts the morning after we scope, transcripts reach you the day they're recorded, and you have findings inside a month — while the decision is still live.
02
Independence
People soften the truth for the person who built the thing. They do not soften it for us. That gap is most of the value.
03
Method
Consistent questions, comparable answers, findings tied to sources. What you get back can be re-read in six months and still mean something — unlike a folder of call notes.
// SECTION: WHO_RUNS_IT006
Who runs your interviews.
Software engineer. I've built for Turing (a Google DeepMind project), Volvo, and Bosch, and led 15+ engineers building a payment gateway doing 2,000+ TPS across 50+ bank and NBFC integrations.
The range: real-time IoT and distributed systems, data pipelines for training LLMs, payments architecture and DevOps, ultra-low-latency voice AI. I learned the whole stack because to build a company you can't have a piece you don't understand.
Which is also how I know what this costs. Every one of those builds ran on assumptions somebody should have tested first, and the expensive ones were expensive because nobody made thirty phone calls before the sprint started. This is me making those calls for you — as an outsider your customers will be blunt with, and an engineer who knows which answers actually change what gets built.
I also run The Ideas Company — maps of the startup landscape for founders: every stage, who funds what, and what the words in a term sheet actually mean.
// SECTION: WHO_ITS_FOR007
Built for teams without a research function.
Early-stage through mid-size. Any segment.
Founders, pre-seed to Series A
Validating before the build, or assembling the evidence a raise is going to ask for. Often the first structured research the company has ever run.
Product teams at growing companies
Deciding what ships next quarter, and needing something better than the loudest stakeholder to decide it with.
Established companies launching something new
You have deep intuition about your current customers and none about this. A new product inside an old company starts from zero evidence, same as a startup.
If you already have a full-time research team, you don't need us. We're for the stage where the alternative to hiring us is nobody doing it at all.
// SECTION: FAQ008
Common questions.
How many people do you actually talk to?
It depends on the question. A validation study is usually somewhere between twelve and twenty conversations — enough that themes start repeating, which is the real signal you have enough. Pricing and positioning work across several segments needs more. We agree the number on the quick call, before you commit to anything.
How fast can you actually move?
Recruiting starts the day after the quick call, and you get transcripts the day each session is recorded — so you're learning in week one. Most engagements run two to five weeks end to end. Recruitment is the variable: consumers and SMB users are fast, senior enterprise buyers in a narrow niche take longer. You'll get an honest range on the call, not an optimistic one.
What does it cost?
It's scoped per engagement, because twelve conversations with SMB users and twenty with enterprise buyers are not the same job. You get a fixed number on the quick call, before any work begins — no hourly meter, no surprise at the end.
Can we sit in on the interviews?
On some, yes — and it's often useful. Not all of them. People answer differently when the founder is listening, and that difference is a large part of what you are paying us to avoid. We'll agree which sessions are open when we scope the study.
What if the research says the idea doesn't work?
Then we tell you that, and we show you the evidence for it. You're paying for the answer, not for agreement. Finding out over three weeks of interviews is the cheapest version of finding out — the expensive version is eighteen months of building.
Who runs the interviews?
We do. Not a subcontracted panel, not an outsourced call centre, not a form. The person who runs your conversations is the person who writes your findings — which is why the findings have judgement in them.
Do you sign an NDA?
Yes, and before the quick call if you'd prefer — we'll turn it around same day so it doesn't hold anything up. You'll be describing an unlaunched product to us; that's normal and we're set up for it.
We're not a startup. Does this still apply?
Yes. The problem isn't company size, it's evidence about a specific decision. An established company launching into a new segment has exactly as little data about that segment as a pre-seed founder has about theirs.
Do we own the outputs?
You do. Recordings, transcripts, findings, and the memo are yours, delivered in formats you can keep and re-read without us.
// SECTION: START
Start with the question you keep avoiding.
One quick call, no pitch deck. Bring the decision you're stuck on and we'll tell you whether research would settle it — including when the honest answer is that it wouldn't. If it's a fit, recruiting starts the next day.
Book a quick call30 minutes. No cost, no obligation. NDA beforehand if you want one.
// SECTION: BOOK009
Grab the next slot.
Thirty minutes with the person who'd actually run your study — not a sales rep reading a qualification script. Come with the decision; leave with a plan and a start date.
Prefer email? kvksatish98@gmail.com