How to Test Demand for a Business Idea in 14 Days

How do you test demand for a business idea in 14 days?

Test a business idea by asking a narrow group of buyers to take a costly action for a specific offer. In 14 days, define the claim, observe the current workaround, interview likely buyers, present an honest offer, and measure what people do. Search interest and compliments are context, not demand. A deposit, signed pilot, approved purchase step, or qualified introduction is stronger evidence because it costs the buyer something.

This is a pre-investment test, not proof that a company will work. The goal is to replace one large assumption with a smaller set of observed facts before committing more time or money.

Day 0: write the claim and the kill criteria

Write one sentence before opening a survey tool or buying traffic:

We believe [specific buyer] will take [specific action] to obtain [specific outcome] under [price, timing, and delivery conditions].

Each bracket matters. “Small businesses want better reporting” cannot be tested because the buyer, problem, offer, and action are undefined. “Finance managers at 20 to 100 person agencies will book a 30-minute workflow review for a weekly cash forecast that replaces a manual spreadsheet” can produce evidence.

Set the decision rules now. For example, continue only if the test produces three qualified conversations, two requests for a concrete next step, and one buyer willing to enter an approved purchasing process. Those are hypothetical thresholds, not benchmarks. Choose numbers that fit the transaction, sales cycle, risk, and audience size. Write the stop rule before results arrive so a weak response cannot be relabelled as success.

Also set a cost cap, an end date, and prohibited tactics. Do not create fake scarcity, false testimonials, invented customers, or a product capability that does not exist. The Federal Trade Commission’s advertising guidance says advertising claims must be truthful, not misleading, and supported by evidence when appropriate. That rule applies to a test page too.

Days 1 and 2: find evidence of current behaviour

Start with behaviour that already exists. A market is easier to believe when the target buyer is already spending money, staff time, political capital, or attention on the problem.

Use Google Trends as directional context only. Google says Trends data is sampled, aggregated, and normalised, with values scaled from 0 to 100. It is not absolute search volume. Compare several descriptions of the same job, record the geography and period, and save the exact terms or topics used. Google also distinguishes a search term from a topic, so document which one you selected.

Write a one-page evidence note. Name the buyer, current alternative, visible cost, likely trigger, and the fact that would disprove your view. If you cannot identify a current behaviour, the next two days should investigate the problem rather than promote a solution.

Days 3 and 4: interview the buyer without selling the answer

Speak with people who have recently faced the problem or own the relevant budget. Ask for a specific past event:

• What happened the last time this problem appeared?
• What did you do next?
• Who was involved in the decision?
• What did the workaround cost in time, money, delay, or risk?
• What caused the issue to receive attention?
• Which approved option did you reject, and why?
• What would have to be true for the organisation to change its process?

Do not ask whether someone “likes” the idea until the behavioural account is complete. People can be polite about a concept they would never buy. Record the difference between the person with the problem, the person who approves spend, the person who uses the result, and anyone who can block implementation.

Five thoughtful interviews may teach more than fifty shallow survey responses, but five is not a universal sample or a claim of representativeness. Keep a source log and note how each participant was recruited. If everyone came through one friend or one industry group, label that concentration.

Days 5 and 6: build the smallest honest offer

Turn what you learned into an offer with a buyer, outcome, scope, price or price range, delivery date, exclusions, and next step. It can be a manual service, a paid diagnostic, a design partner pilot, a preorder with clear terms, or a proposal for an existing procurement process. It does not need finished software.

The offer must say what exists today. If the service will be delivered manually, say so. If payment is refundable under stated conditions, put those conditions in writing. If the test collects personal or business information, collect only what is necessary and state how it will be used. For contracts, deposits, regulated claims, privacy obligations, tax, and consumer rights, obtain qualified legal, accounting, or compliance advice in the relevant jurisdiction before accepting commitments.

The test should isolate the hard assumption. If the main uncertainty is willingness to pay, a free waitlist does not test it. If the main uncertainty is implementation effort, a payment button does not test whether the buyer can approve access, data, or workflow change.

Days 7 to 10: put the offer in front of a bounded audience

Choose one channel that reaches the defined buyer. That might be direct outreach to a documented list, a partner introduction, a small event, an existing audience, or a tightly capped advertising test. Keep the audience, message, and offer stable long enough to interpret a response.

If using search advertising, Google Ads Keyword Planner forecasts can provide forecast clicks and cost. Google says forecasted impressions account for bid, budget, seasonality, and historical ad quality. Treat forecasts as modelled context, not observed demand. Record the settings and compare the forecast with what actually happened.

Track the full path:

1. People who received or saw the offer.
2. People who met the buyer definition.
3. People who responded.
4. People who completed a discovery step.
5. People who accepted a costly next action.
6. People who declined, with the reason when they volunteered one.

Do not optimise merely for click-through rate. A clever headline can increase clicks while attracting the wrong buyer. The useful question is whether qualified people move toward the transaction under the stated conditions.

Days 11 and 12: ask for the next costly action

A costly action is not limited to money. In a large organisation, introducing a budget owner, sharing a redacted workflow, scheduling an implementation review, completing a vendor form, or obtaining internal approval may be more meaningful than a casual deposit.

Choose an action that resembles the real path. A business with a six-month procurement cycle should not interpret a two-day purchase failure the same way as a consumer download. Instead, test whether a real sponsor will begin the next required step and whether the organisation can clear an obvious blocker.

Record objections without arguing them away. “No budget this quarter,” “security review will not allow it,” “the current workaround is adequate,” and “the result is valuable but this team does not own the decision” point to different problems. One concerns timing, one feasibility, one urgency, and one buyer selection.

Days 13 and 14: calculate, decide, and write the next test

Use simple counts before constructing a story around them:

• Qualified response rate = qualified respondents divided by qualified people reached.
• Discovery completion rate = completed discovery steps divided by qualified respondents.
• Costly-action rate = buyers taking the selected action divided by completed discovery steps.
• Acquisition cost for the test = direct test spend divided by qualified costly actions.

Recompute each figure from the raw counts. If the denominator is small, show the count beside the percentage. Three actions from ten completed conversations is more informative than “30 percent conversion” on its own. Do not annualise a two-week result or present the test as a forecast of company revenue.

End with one of four decisions:

• Stop because the problem, buyer, or urgency was not supported.
• Change the buyer because interest appeared elsewhere.
• Change the offer because the problem was real but the transaction was wrong.
• Continue to a larger test because qualified buyers took the expected costly action.

The memo should preserve the original claim, recruitment method, audience, offer, counts, objections, cost, limitations, and decision. It should also name the one assumption the next test will address. That is the useful output of 14 days: not certainty, but a smaller and more expensive-to-ignore set of unknowns.

If the result supports another step, our working method starts with the decision and works backward through evidence, constraints, and sequencing. The reverse-engineering framework shows how the same discipline applies to a whole business. What we do and how to work with us explain the research and execution context.

Sources

• Federal Trade Commission: Advertising and Marketing Basics
• Google Trends: FAQ about Google Trends data
• Google Trends: Search terms and topics
• Google Ads: About Keyword Planner forecasts

This article is general information about testing demand for a business idea. It is not financial advice. Decisions that commit your money, sign contracts, or make regulated claims belong with a licensed financial adviser or another qualified professional.