Record assumptions
Before the test begins, note the intended customer, exact offer, price, expected costs and the result you hope to observe. Be specific: “contact ten salon owners and secure two paid bookings for simple product photographs” can be checked; “see whether people like photography” cannot. Also record what would make you adjust or stop the test. This prevents you from changing the goal afterwards to make weak results appear successful.
Track contacts
For each person approached, record the date, who they were, the message or channel used, their response, questions and objections. Do not rely on memory or save important evidence only inside scattered WhatsApp chats. Patterns matter: several prospects asking whether delivery is included may show that the offer is unclear, while silence from an unsuitable group may say little about demand. Protect private information and use simple labels or initials when full customer details are unnecessary.
Measure action
Count different actions separately: messages opened, replies, requests for details, quotes, referrals, preorders, deposits and completed sales. Likes and encouraging comments can support visibility, but they are weaker evidence than someone agreeing to pay. Record actual spending, time used, cancellations, revisions and delivery problems as well. A test with two sales may still be unsuitable if each order takes far longer or costs much more than expected.
Review
At the end of seven days, compare the results with your original assumptions. Which customer responded, which words attracted questions, what price resistance appeared and what did delivery really require? Decide whether to continue, adjust or pause. Continue when there is credible action and the work is manageable; adjust when interest exists but the offer, customer or process is unclear; pause when costs, safety, demand or personal commitments make the idea unsuitable for now. One short test guides the next decision—it does not guarantee the future.