Translation and notes
A readable adaptation of the article's core ideas, with the original source clearly credited.
The essay argues that startup investment returns concentrate in a few exceptional outcomes, and that some good opportunities initially look unpromising. Looking unpromising is not itself evidence of merit. What follows is an independent product investigation exercise, not the author's procedure or a complete retelling of the essay.
Imagine a team exploring appointment coordination between small shops and repair technicians. A colleague objects that shopkeepers already manage by telephone. Instead of building an app or debating the claim, ask about one recent repair: how many calls were needed to agree on a time, and where did coordination break down? Test only scheduling, not technician selection, repair quality, and payment together. With both parties' permission, follow one real request from available time slots to confirmation. Record missing information, late replies, and every intervention by the team. Collect only necessary data and agree on a deletion date.
Before starting, define the observation: can both parties confirm the appointment without another scheduling call? Afterwards, invite the customer to choose between the old route and the trial service for another request, without discounts or pressure. Record the choice and its reason, alongside the staff time needed to deliver the service. A useful service that requires constant manual coordination is a different finding from a tool people can use unaided. For this exercise, a possible limit is one week and three voluntary requests; that is a planning example, not a sample sufficient to establish market size or forecast growth.
At the review, separate observation, interpretation, and the next decision. No incoming requests may indicate poor access to suitable participants rather than no customer problem. A shop wanting help while a technician cannot respond is different from both parties rejecting the offer. Decide whether to change one step, repeat a bounded trial with a more suitable group, or stop this proposal. Assign responsibility for participant follow-up and data deletion. The useful output is an explicit account of what worked for whom, including unanswered questions, rather than a persuasive story about the company's future.
Selected advice from the article
- Suggested worksheet: objection, customer event, selected step, consent, observation, staff time, customer choice, next decision.
- For each review claim, ask whether it was observed, inferred, or remains to be tested.
Operating takeaways
- Convert a broad objection into a question about a recent customer event.
- Test one workflow step with consent and only the necessary data.
- Report manual coordination time alongside the customer's outcome and next choice.
- Distinguish missing requests, failed coordination, and a customer choosing not to return.
- Set an effort limit and review date; do not generalize beyond the trial's scope.
Quick feedback
What did this article make clearer for you?
Choose a reaction or share a question or experience. Useful responses can shape future notes.
This is a free adaptation, not a word-for-word republication of the original article.