Validate your business idea before the expensive part.

Mara asks landlords about their most recent turnover: how they found a cleaner, what was included, who held the keys and what delayed the next tenant. Those questions have details attached. 'Would you use my amazing new service?' mostly has an obvious polite answer attached.

For your customer interviews, ask about an actual occasion. What happened? What did they try? What did it cost in money or time? Who decided what to do? Ask follow-up questions when the answer surprises you, and leave enough silence for the person to think.

Keep a research invitation honest. If you promised a conversation without a sales pitch, honor that. You can ask whether they would like to hear about a test later.

Before the test, Mara writes an assumption: landlords in a small service area need weekend turnover cleans and will pay a price that covers travel, supplies and all her hours. She also writes what would weaken the idea: requests mostly outside the area, incompatible timing or a price nobody can accept.

In four randomized trials involving 759 firms in Italy and the UK, researchers found that training entrepreneurs to frame and test hypotheses increased the number of ideas they stopped. The 2024 study examined a structured training program, not a single worksheet. Leave room in your own test for evidence that changes the plan, including a reason to stop. (Camuffo and colleagues (2024): testing business ideas)

A first test might be one defined cleaning job with a written scope and a quoted price, after checking the requirements to do the work. A test for a physical product could start with a demonstration and customer feedback before accepting orders. Choose a test suited to what is uncertain.

If you do take a deposit or presell, explain what exists, what the customer will receive, the delivery timing and the cancellation or refund terms. Check the rules for that sale before collecting money. A cheerful mockup shouldn't leave someone thinking a finished product is sitting in a warehouse.

Price and interest are only part of the test. Record delivery time, supply costs, questions the customer asks and any rework. One sale can uncover a problem; it cannot establish repeat demand on its own.

If someone declines, ask whether they are willing to explain. They may like the service but need a weekday. They may already have someone reliable. They may need a different scope altogether. Avoid treating every no as a request for a lower price.

In Mara's example, several requests for Friday morning service would matter more than a stack of compliments about her logo. She could investigate that schedule, choose a different customer group or decide the business doesn't fit her available hours. All three would help her decide what to do before a bigger commitment.

Keep a short record of the assumption, who you spoke to, what you offered, the response and the next decision. Separate what people did from what you think it means. When the stories disagree, keep the disagreement visible.

You now have material for a working business plan: an offer with boundaries, costs that need checking and a next test. That's something you can work with when 'everyone loved it' wears off.