132. 7-Day Customer Discovery Playbook

Use this playbook when you have an idea, feature, segment, or pivot question and need evidence quickly.

The goal is not to prove that you are right. The goal is to learn whether a specific customer segment has a real, repeated, painful problem and whether the next step is worth taking.

Output by day 7

By the end of the week, you should have:

• 12 to 20 qualified conversations requested
• 5 to 10 actual conversations completed
• A pattern table
• A clear segment decision
• A continue, narrow, change, or stop decision
• The next experiment

Day 1: Segment and hypothesis

Write a narrow hypothesis.

Use this format:

We believe [specific customer] has [specific problem] when [specific trigger/event happens], and today they solve it by [current workaround].

Weak:

SMEs need better analytics.

Stronger:

D2C founders doing Rs 20 lakh to Rs 2 crore monthly revenue struggle to know which ad campaigns create profitable repeat customers, and today they combine Shopify, Meta Ads, spreadsheets, and agency reports manually.

Create a prospect list with:

Field	Notes
Name	Person, company, or account
Segment fit	Why this person matches the hypothesis
Role	User, buyer, influencer, or expert
Source	Warm intro, LinkedIn, customer list, community, event, founder network
Ask	Interview, workflow review, pilot, paid test, referral
Status	Not contacted, contacted, replied, scheduled, completed

Daily operating quotas

Discovery fails when it becomes a vague intention. Use a daily operating quota so the week has enough raw material.

Day	Minimum activity	Quality check
1	30 prospects listed, 10 exact-fit prospects marked	At least half the list should be people you can plausibly reach.
2	15 outreach messages sent	Messages should name a specific workflow or trigger.
3	10 more messages, 3 calls scheduled or completed	If replies are weak, change the segment or opening line.
4	3-5 calls completed	At least two calls should include a recent real example.
5	3-5 calls completed, 2 artifact asks made	Ask for a spreadsheet, screenshot, email thread, WhatsApp flow, report, or process doc.
6	Pattern review written	Count evidence, not enthusiasm.
7	Decision memo completed	Choose continue, narrow, change, or stop.

The quota is not there to make the founder busy. It protects against declaring a conclusion from two random calls and one friendly opinion.

Days 2-5: Conversations

Do not pitch first. Study reality first.

Use this script:

1. “Can you walk me through the last time this happened?”
2. “What triggered it?”
3. “How did you solve it?”
4. “Who else was involved?”
5. “What did it cost in time, money, risk, or stress?”
6. “What have you already tried?”
7. “What makes existing solutions not good enough?”
8. “If this got solved, what would improve?”
9. “Would you be open to reviewing a rough solution or pilot?”
10. “Who else should I speak with?”

Capture exact language. The way customers describe the problem is often better than the words founders invent.

Interview routing

After each call, route the person into one of four buckets.

Bucket	Use when	Next action
Strong problem	Recent example, clear cost, current workaround, useful next step	Ask for artifact, buyer/user intro, pilot problem statement, or paid diagnostic.
Problem but wrong segment	Pain exists but customer profile is not repeatable for your startup	Capture learning, ask who faces it more often, then deprioritize.
Curious but weak pain	Likes the idea but cannot describe recent behavior	Nurture lightly; do not build for them.
Expert/context	Consultant, investor, advisor, operator, or peer with secondhand market knowledge	Use for pattern context, not validation proof.

This keeps discovery honest. Experts can explain a market, but they usually cannot validate buyer urgency unless they are the buyer or close to actual purchase behavior.

What counts as signal

Strong signals:

• Customer describes a recent real example without prompting
• Current workaround is painful or expensive
• Customer has tried to solve it before
• Customer introduces another relevant person
• Customer shares data, documents, screenshots, or process details
• Customer asks about timing, price, or pilot
• Customer follows up without being chased

Weak signals:

• “Interesting idea”
• “This could be useful”
• “Keep me posted”
• “I know someone who might like this”
• Survey interest without behavior
• Compliments from friends

Day 6: Pattern review

Do not review conversations one by one. Review patterns.

Use this table:

Pattern	Evidence	Count	Strength
Same customer type	Who repeated?		Strong, medium, weak
Same trigger	What event creates urgency?		Strong, medium, weak
Same workaround	What do they do today?		Strong, medium, weak
Same cost	What pain is measurable?		Strong, medium, weak
Same buyer	Who controls budget or decision?		Strong, medium, weak
Same next step	Who wants to continue?		Strong, medium, weak

If every conversation teaches a different lesson, your segment is too broad.

Day 7: Decision

Choose one:

Decision	When to choose it
Continue	Same pain, same segment, credible next steps
Narrow	Pain is real, but only for a smaller segment
Change	Customer exists, but problem, buyer, or wedge is different
Stop	Pain is weak, access is poor, or no one will move

Write the decision in one paragraph:

We will [continue/narrow/change/stop] because [evidence]. The biggest remaining risk is [risk]. The next experiment is [experiment] by [date].

Evidence threshold

A discovery sprint should not require perfect certainty. It should create enough evidence for the next week.

Use this threshold:

Signal	Weak	Strong
Problem	People agree in theory	People describe a recent painful example
Segment	Mixed customer types	Same customer type repeats
Workflow	Vague frustration	Clear current process or workaround
Urgency	”Useful someday”	Trigger, deadline, cost, or repeated pain
Buyer	User interest only	Budget owner or decision path identified
Next step	”Keep me posted”	Pilot, intro, data share, payment, or another meeting

If you finish the week with five polite compliments and no workflow detail, the answer is not “build faster.” The answer is to narrow the segment, change the question, or find customers closer to the pain.

Red-team review

Before deciding to continue, force yourself to argue against the idea.

Red-team question	What to look for
Are we over-counting friendly people?	Friends, peers, investors, and community members may encourage you without buying.
Did anyone show current behavior?	Workaround, spend, workflow, urgency, or repeated follow-up.
Did we speak to the buyer?	If only users spoke, budget risk remains.
Is the segment still too broad?	Different triggers, tools, budgets, and buyers mean the segment needs narrowing.
Did anyone give a costly next step?	Time, data, internal intro, pilot work, or payment.
What would make this idea false?	Name the evidence that would make you stop or change.

If the idea survives the red-team review, the next week deserves a sharper experiment. If it does not, you have saved yourself from building into fog.

India field notes

For Indian founders, discovery often moves through trust networks before it becomes formal research. Warm introductions, alumni groups, founder WhatsApp groups, trade associations, local business communities, CA/CS/lawyer networks, and customer referrals can outperform cold survey links.

Be careful with two traps:

• People may agree to help because they like you, not because the problem is urgent.
• Small businesses may describe pain clearly but avoid payment or process change unless trust is high.

When possible, ask to see the current artifact: spreadsheet, WhatsApp thread, invoice, dashboard, manual checklist, MIS report, or screenshot. Artifacts reduce politeness bias.

Templates

Outreach message

Hi [Name], I am researching how [specific segment] handles [specific problem]. Not selling anything on this call. Could I ask you 20 minutes about the last time this came up for you? Your practical view would help me avoid building the wrong thing.

Conversation note

Question	Answer
Segment fit
Last real example
Trigger
Current workaround
Cost of problem
Existing alternatives
Buyer/user/influencer
Next step offered
Exact customer words

Pattern memo

At the end of the week, write one page:

Field	Answer
Strongest segment
Repeated pain
Current workaround
Trigger
Buyer or decision maker
Evidence strength	Strong / medium / weak
Biggest uncertainty
Next test

Evidence triage board

After every interview, put the evidence into a simple board. This prevents the founder from treating every conversation equally.

Evidence type	What to capture	How to use it
Direct quote	Exact words the customer used for the pain, workaround, or objection	Use in positioning, outreach, and problem statements
Workflow artifact	Screenshot, spreadsheet, invoice, checklist, WhatsApp thread, report, dashboard, or manual process	Decide what the product must fit into
Trigger	Event that makes the pain urgent now	Find better prospects and sharper outreach
Cost	Time, money, delay, risk, missed revenue, or emotional load	Judge whether the pain can support a business
Buyer path	Who owns budget, approval, procurement, or implementation	Avoid selling only to curious users
Contradiction	Customer behavior that conflicts with what they said	Protect against polite false positives

At the end of day 6, count evidence, not opinions. A useful discovery sprint should produce at least:

• Five recent examples of the problem.
• Three current workarounds.
• Three trigger events.
• Two people who can explain the buyer or approval path.
• One concrete next test: paid pilot, prototype, landing page, deeper segment interview, or stop.

If the board is mostly compliments, brainstormed ideas, or future hypotheticals, the sprint did not validate the problem yet.

Conversation quality control

Not every customer conversation is useful. Score each interview immediately after it ends.

Quality question	Good sign	Weak sign
Was the person in the target segment?	Yes, specific role and context	Friend, generic operator, or adjacent segment
Did they describe recent behaviour?	Specific recent example	Abstract opinion or future guess
Did they show a workaround?	Tool, spreadsheet, process, vendor, manual hack	”We should probably improve this”
Was there a cost?	Time, money, risk, revenue, stress, delay	Minor annoyance
Was the buyer path clearer?	Named approver, budget, procurement, owner	Unknown or “management will decide”
Did the call change your thinking?	New pattern or contradiction	Only confirmed founder belief

Give each call a quality score from 1 to 5. A week with ten low-quality conversations is weaker than a week with five strong ones.

Pattern synthesis worksheet

On day 6, do not summarize by memory. Use a worksheet.

Pattern	Evidence count	Strongest quote/artifact	Segment	Implication
Repeated pain
Current workaround
Trigger event
Buyer path
Willingness to pay
Product requirement
Contradiction

Separate three things:

Type	Meaning
Fact	What the customer said or showed.
Pattern	What repeated across customers.
Interpretation	What you think it means for the startup.

Founders often jump from one quote to product strategy. The worksheet slows that down. Evidence should earn its way into decisions.

Next experiment choice

End the sprint by choosing one next experiment. Do not end with “continue researching” unless the segment was clearly wrong and you need a new segment sprint.

Discovery outcome	Next experiment
Strong pain, clear buyer, no solution proof	Prototype or concierge MVP
Strong pain, unclear payment	Pricing conversation or paid diagnostic
Strong pain, unclear segment	Narrowed discovery sprint
Strong user pain, unclear buyer	Buyer discovery sprint
Interest but no urgency	Landing page around sharper trigger or stop
Weak/no repeated pain	Stop or choose a different problem

The output of discovery is not certainty. It is a better next bet.

Day 7 decision meeting

Do not end the discovery sprint alone in a notebook. Run a 45-minute decision meeting with co-founders, one advisor, or one trusted operator who is willing to challenge you.

Use this agenda:

Minute	Topic	Output
0-5	Restate the original hypothesis	One sentence: “We believed [segment] has [pain] because [reason].“
5-15	Review strongest evidence	Top three calls, artifacts, workarounds, or buyer signals.
15-25	Review contradictions	Evidence that weakens or changes the hypothesis.
25-35	Choose interpretation	Strong pain, weak pain, wrong segment, wrong buyer, or unclear evidence.
35-45	Choose next experiment	Prototype, paid diagnostic, pricing test, buyer discovery, narrowed sprint, or stop.

The meeting should produce one decision statement:

Based on [evidence], we believe [segment] has [pain] with [urgency]. Next week we will run [experiment] to test [next riskiest assumption]. We will stop or change direction if [kill signal] appears.

Use this decision rule:

Evidence pattern	Decision
Repeated painful workflow, named buyer, visible workaround	Move to prototype or concierge MVP.
Strong user pain, but no buyer clarity	Run buyer discovery before building.
Strong pain, but weak willingness to pay	Run pricing or paid diagnostic conversations.
Mixed pain across many segments	Narrow to the segment with strongest urgency.
Mostly compliments and future opinions	Do not build yet. Run better discovery or stop.
Pain exists but founder does not care about the market	Stop. Founder energy matters.

The discipline is to make the next experiment smaller and sharper. Discovery fails when it becomes endless. It also fails when founders use weak evidence as permission to build for months. The day 7 decision meeting keeps the company moving without pretending it knows more than it does.

Recruiting scripts by channel

Discovery quality depends on who you speak to. Use different asks for different channels instead of sending the same message everywhere.

Channel	Better ask	Why it works
Warm intro	”Could you introduce me to one person who directly owns [workflow]? I am trying to understand the last time it broke, not pitch them.”	Makes the intro specific and low-pressure.
LinkedIn	”I noticed [trigger]. Does your team own [workflow], or is this handled elsewhere?”	Lets the person redirect you if they are not the right buyer.
Founder community	”Looking for founders/operators who recently dealt with [specific event]. I want to study the workflow for 20 minutes.”	Uses context instead of broad survey language.
Customer/user list	”We are reviewing how people currently handle [workflow]. Could I ask about your last real example?”	Signals learning, not upsell.
Expert/advisor	”Which customer type feels this problem most painfully, and what would be weak evidence?”	Uses experts to improve targeting, not validate demand.

Avoid asking, “Can I pick your brain?” It is vague and makes the other person do work. A narrow ask gets better replies.

Artifact asks

Whenever a customer describes a workflow, ask for an artifact. Artifacts are useful because they show what people actually do, not only what they remember.

Useful asks:

• “Could you show me the spreadsheet or report you use today?”
• “Could you blur sensitive details and share a screenshot of the workflow?”
• “Can you forward a sample email or ticket thread with names removed?”
• “Can you walk me through the WhatsApp or Slack flow without sharing private data?”
• “Can I see the checklist, SOP, invoice, dashboard, or reconciliation file?”

Artifact rule:

If the customer…	Treat it as…
Immediately knows which artifact to show	Strong workflow evidence.
Needs approval but explains the artifact clearly	Medium evidence; follow up.
Cannot name any artifact or current process	Weak evidence; the pain may be abstract.
Shares data, sample files, or process screenshots	Strong next-step signal.

Do not collect sensitive data casually. Ask for redacted examples, screenshots, or walkthroughs where needed. Respect privacy, compliance, and customer trust.

Discovery scoring sheet

At the end of each call, score it from 0 to 2. This creates a more honest sprint review.

Signal	0	1	2
Segment fit	Wrong or unclear	Adjacent	Exact target segment
Recent example	None	Vague example	Specific recent story
Current workaround	None	Light workaround	Clear process, tool, spend, or manual effort
Cost of pain	Minor annoyance	Some time or stress	Money, risk, delay, revenue, churn, or founder attention
Buyer clarity	Unknown	Possible buyer named	Buyer, approver, or owner identified
Next step	None	Friendly interest	Artifact, intro, pilot, pricing conversation, or follow-up date

Interpretation:

Total score	Meaning
10-12	Strong discovery evidence. Move to the next experiment.
7-9	Real signal, but one major risk remains. Investigate that risk next.
4-6	Useful learning, weak validation. Narrow the segment or question.
0-3	Do not build from this evidence.

Scoring does not replace judgment. It stops the loudest anecdote from dominating the week.

Discovery saturation check

At the end of seven days, ask whether you have enough evidence to decide or whether you only collected a few interesting stories.

Use this saturation check:

Question	Good signal	Weak signal
Did the same pain repeat across similar people?	5+ exact-segment conversations mention the same situation, workaround, or consequence.	Each call produces a different problem.
Did customers describe recent examples?	They can name when it happened, who was involved, and what it cost.	They agree in theory but cannot recall specifics.
Did a current workaround appear?	Tool, spreadsheet, person, agency, WhatsApp group, checklist, or internal process exists.	The problem has no visible behavior today.
Did a buyer or owner appear?	Someone owns budget, risk, workflow, or behaviour change.	Users care but no owner is clear.
Did anyone offer a next step?	Artifact, intro, pilot conversation, pricing discussion, or follow-up date.	Friendly encouragement with no commitment.

Then decide:

Discovery result	Next move
Repeated pain, clear owner, current workaround, strong next steps.	Move to problem validation, pricing ask, or MVP scope.
Repeated pain but unclear buyer.	Run buyer discovery before building.
Pain is real but segment is mixed.	Narrow the segment and repeat discovery.
Many compliments, little behavior.	Stop pitching and restart interviews around current workflow.
No repeated pattern after serious outreach.	Park or kill the idea, or choose a new segment.

Discovery is not complete when the founder feels encouraged. It is complete when the next experiment is obvious.

Low-Response Rescue Plan

If day 3 arrives and almost nobody has replied, do not conclude the problem is dead yet. First diagnose whether the issue is segment, access, message, or ask.

Symptom	Likely issue	Fix within 24 hours
Warm intros reply, cold outreach does not	Access problem	Ask every warm contact for one exact-segment intro.
People reply but decline calls	Ask is too heavy or too vague	Offer a 10-minute question or async answer around a specific workflow.
People say “not my area”	Wrong role	Ask who owns the workflow, budget, approval, or pain.
People accept but no-show	Low urgency or weak trust	Confirm with a sharper agenda and a recent-event question.
Experts reply but buyers do not	You are easier to access than the market	Use experts only to identify buyer titles, triggers, and communities.
Everyone says “interesting” but avoids time	Problem may be weak or message abstract	Rewrite outreach around a concrete trigger and current workaround.

Use this rescue message:

Quick follow-up: I may have framed this too broadly. I am specifically trying to understand how [role/segment] handles [workflow] when [trigger] happens. Even a 5-minute reply to this question would help: what do you do today when this happens?

If response stays weak after a sharper segment, clearer trigger, and better access path, treat that as evidence. A startup needs a reachable market, not only a painful problem.

Good Note Versus Bad Note

Discovery quality often fails in the notes. Founders remember the emotional tone of the call and lose the evidence.

Bad note	Better note
”They liked the idea."	"They spend 3 hours every Friday reconciling Shopify payouts, Meta spend, and agency reports. They asked whether this could connect to Razorpay exports."
"Pricing may be okay."	"They currently pay Rs 25,000/month to an agency for reporting but said they would only pay for software if it reduced founder review time."
"Need dashboard."	"Founder checks CAC, repeat orders, refund rate, and campaign spend manually before every Monday growth meeting."
"Interesting SMB segment."	"D2C brand between Rs 30 lakh and Rs 1 crore monthly revenue has founder-led reporting pain; below that, owner does not track enough."
"Could be pilot."	"They offered to share last month’s anonymized report and agreed to review a prototype next Thursday.”

Good discovery notes contain:

• A recent example.
• The current workaround.
• The cost of the workaround.
• The buyer or owner.
• Exact customer words.
• The next action the customer agreed to.

Bad notes contain mostly founder interpretation. Better notes preserve reality so the founder can make a decision later without re-inventing the conversation.

Common mistakes

• Interviewing whoever is available instead of the target segment
• Asking whether people like the idea instead of studying past behavior
• Pitching before understanding the current workflow
• Counting compliments as demand
• Ending the week with “more research needed” instead of a decision

Done when

This sprint is done when you can explain:

• Which customer segment showed the strongest pain
• Which pain repeated
• What current workaround exists
• What evidence was strong versus weak
• What you will do next week

Related chapters

• 18. Customer Discovery Fundamentals
• 19. Running Customer Interviews
• 146. Customer Discovery Templates