Do six requests justify building PDF export?
The team has ten interview notes. Six ask for PDF export; four ask for CSV. A colleague wants to build PDF first.
The question
RelayDesk and these interviews are fictional. Calls are not independent companies: multiple people can speak for one company. The sample was recruited for discovery and is not representative of the whole market.
| PDF requests | CSV requests | |
|---|---|---|
| Company A | 4 | 0 |
| Company B | 2 | 0 |
| Companies C–F, one person each | 0 | 4 |
Ten calls represent six companies. No purchase or adoption commitment was recorded.
You may spend up to two hours testing an export manually with two consenting companies. A product commitment needs the product lead’s approval.
“Choose your next action and explain the trade-off.”
Not saved by this site. Your browser may restore text when you navigate back.
A strong answer, step by step
Open each step in turn. This is one strong answer, not the only one.
Step 1: What does the tally show?
What would you do here? Open the move you’d make.
Separate calls from companies
The strong move. Repeated voices can dominate a call count.
Call PDF the preference of 60% of customers.
Tempting, but it misses something. The ten calls are neither ten independent companies nor a representative customer sample.
See how a strong candidate says it
“PDF appears in 6 of 10 calls, but only 2 of 6 companies. CSV appears in 4 calls from 4 companies.”
Why it works: Repeated voices can dominate a call count.
Step 2: What do you ask next?
What would you do here? Open the move you’d make.
Ask about the export’s use
The strong move. A requested format is a proposed solution, not yet the underlying job.
Start coding PDF to reward the loudest customers.
Tempting, but it misses something. Volume alone does not explain the need or establish value.
See how a strong candidate says it
“Ask what each company needs to do with the export, who uses it and what happens today.”
Why it works: A requested format is a proposed solution, not yet the underlying job.
Step 3: What is the smaller test?
Companies A and B explain that their accountants need a table they can import into an existing system. Both systems accept CSV. Neither company needs a printable report.
What would you do here? Open the move you’d make.
Test the accountant workflow
The strong move. The new context suggests a common job that can be tested cheaply.
Assume every company can now use CSV.
Tempting, but it misses something. The two companies’ clarification does not prove the solution works for them or everyone else.
See how a strong candidate says it
“Test whether a manually prepared CSV lets those accountants complete the job before committing to a PDF feature.”
Why it works: The new context suggests a common job that can be tested cheaply.
Step 4: What would count as learning?
Preparing, observing and recording one export test takes 30 minutes. Each of the two companies agrees to try two exports, a week apart.
What would you do here? Open the move you’d make.
Observe completion and repeat use
The strong move. The test fits the two-hour limit and checks repeat use, not only approval in a call.
Declare product-market fit after both companies say thanks.
Tempting, but it misses something. Polite feedback is not sustained use or broad market evidence.
See how a strong candidate says it
“Two exports at 30 minutes each for each of two companies use 120 minutes. Watch import completion and ask whether they use the second export without help.”
Why it works: The test fits the two-hour limit and checks repeat use, not only approval in a call.
The answer, in short
- Decision
- Run the bounded CSV workflow test and bring its results to the product lead.
- Because
- The original tally mixes people with companies and hides the job behind the format.
- What would change it
- Observed failure to import, repeated use, or a different underlying job could change the solution.
Common traps
- Do not turn a discovery sample into a market share estimate.
- A requested feature is not proof of the best solution.
Check your answer
- I checked the unit of counting.
- I asked what the user needs to accomplish.
- I kept the test within budget.
- I defined evidence beyond positive comments.
Last reviewed Nov 10, 2026