Decide whether your business is ready for paid advertising by proving demand, calculating allowable acquisition cost and setting a loss limit before spending.
Short answer: Usually, no. First make at least 10 normal-priced sales through direct, referred or unpaid routes, calculate contribution per sale, and prove that the buying path works. Advertise earlier only when you already have equivalent offline evidence and can cap the test loss in advance, with one audience, one offer and one defined stopping decision.
Advertising can put an offer in front of people. It cannot tell you cleanly why an unproven offer failed. No sales might mean the audience, message, price, page, trust or fulfilment promise was wrong.
That makes early advertising an expensive bundle of unanswered questions. Manual sales feel slower, but they reveal what buyers misunderstood and what they valued before you pay to repeat the message.
Use the Paid Acquisition Proof Equation
The Paid Acquisition Proof Equation turns readiness into an economic claim: advertising result = orders × contribution per order minus media spend minus additional handling cost
You are ready to test scale only when you can support each input with evidence. The equation is simple. Obtaining honest numbers is the work.
- Demand: Question you must answer: Will the intended buyer pay the normal price?; Acceptable early evidence: Completed sales to buyers outside your support network; Dangerous assumption: Likes or stated interest equal demand
- Unit economics: Question you must answer: What remains before acquisition and overhead?; Acceptable early evidence: Actual price, refunds, materials, labour and fees; Dangerous assumption: Revenue is available to spend on ads
- Buying path: Question you must answer: Can a qualified visitor complete the purchase?; Acceptable early evidence: Observed enquiries, orders and drop-off points; Dangerous assumption: More traffic will repair friction
- Capacity: Question you must answer: Can you fulfil extra work at the promised standard?; Acceptable early evidence: Measured delivery time and available slots or stock; Dangerous assumption: Every additional order has the same cost
My view is that advertising should scale evidence, not create it. Some founders use ads as a market test because traffic can be purchased quickly. The problem is interpretation. Unless the rest of the system is known, a failed campaign does not tell you whether the market or the campaign failed.
Count normal sales, not merely organic clicks
“Organic” does not have to mean search traffic or social reach. A referred customer, direct sale, market transaction or personal outreach can prove that an appropriate buyer will pay. What matters is that the sale occurred on terms you intend to repeat.
Use 10 completed normal-price sales as a working readiness threshold for a straightforward small-business offer. It is not a statistical guarantee. Complex, high-value services may learn from fewer contracts, while a low-cost consumer item may need more. The sales should reveal actual delivery time, refund or rework risk and the questions buyers ask before committing.
Exclude support purchases, exceptional discounts and orders from people who could not reasonably be reached through the planned advertising. They may provide cash and operational learning, but they do not validate that acquisition route.
Calculate what one acquired customer can cost
Start with contribution from the first purchase: sale price minus product or delivery labour minus materials minus fulfilment minus transaction costs minus expected rework or refunds
That figure is your absolute break-even acquisition cost before general overhead. It is not a sensible target because spending all contribution on advertising leaves nothing for rent, administration, tax risk or profit. Decide how much contribution the business must retain, then allocate only the remainder to acquisition.
Do not include repeat purchases until you have observed them over a relevant period. A subscription button does not create retention. If three completed cohorts show reliable repeat behaviour, you can include a conservative portion and state the assumption.
For services, include the time spent qualifying poor leads. A campaign that produces 40 enquiries and one sale may look cheap in the advertising account while consuming two days of unpaid conversation.
Prove the buying path with people, not traffic
Watch suitable buyers attempt the same steps an advertisement would create. Can they understand the result, find the price basis, assess risk and take the next action without an explanation from you?
Record each stage: qualified visit or conversation, enquiry, accepted terms, payment, fulfilment and repeat purchase. Do not borrow a conversion rate from another business. Your price, trust, traffic source and buying friction make it non-transferable.
Fix obvious gaps before spending. If every manual buyer needs a call to understand sizing, an advertisement promising instant purchase sends people into a decision they cannot complete. Either keep the call or redesign the offer.
Measurement also needs a commercial end point. Clicks and impressions describe delivery of media. Judge the campaign by contribution after advertising and the quality of customers retained.
Set the loss and stopping decision in advance
A first campaign should test one audience, offer and buying path. Write the maximum cash loss, founder time and run length before it begins. The amount must be affordable if the result is zero.
Define what you will do at the end. If qualified visits occur but buyers abandon price, investigate price or value evidence. If nobody qualified arrives, inspect targeting and message. If sales occur below the allowable acquisition cost, repeat cautiously. Do not change five parts midway and call the combined result a test.
Avoid raising the budget merely because the platform reports that more data would help. More spending can improve measurement, but it cannot make an unaffordable loss responsible.
Worked example: Moor & Moss Candles
Moor & Moss Candles is a new Sheffield candle business. Its candle sells for £34. Wax, vessel and fragrance cost £11, packaging and fulfilment cost £4, and transaction costs average £1. First-purchase contribution before advertising and overhead is therefore £34 minus £11 minus £4 minus £1 = £18.
The owner spends £400 on an early campaign before establishing normal sales. The campaign produces 500 visits at an average media cost of £0.80 and 12 orders.
- Sales revenue: Calculation: 12 × £34; Amount: £408
- Contribution before ads: Calculation: 12 × £18; Amount: £216
- Media spend: Calculation: Fixed test spend; Amount: £400
- Result before general overhead: Calculation: £216 minus £400; Amount: £184 loss
To cover £400 of media from £18 first-purchase contribution, Moor & Moss would need £400 ÷ £18 = 22.23, so at least 23 orders. From 500 visits, that means 4.6 per cent would have to buy merely to reach advertising break-even. The observed result was 12 divided by 500, or 2.4 per cent.
Neither percentage is a market benchmark. The figures are illustrative. The campaign reveals a loss, but not whether targeting, trust, price or the buying page caused it. Ten earlier manual sales would have exposed several of those uncertainties for less cash.
Use a narrow exception responsibly
Advertising before organic sales can make sense when the business already has equivalent proof elsewhere. A retailer opening a second location may know the offer and economics. An experienced operator launching the same service in an adjacent area may have a credible contribution model.
It can also be a deliberate research expense when the loss is affordable and the hypothesis is narrow. Call it research, not customer acquisition, and do not use an inconclusive result to justify a larger spend.
If your product is regulated, makes health or financial claims, or is sold to consumers, advertising, evidence and cancellation requirements vary by jurisdiction and sector. Check current official rules and obtain qualified local advice before publishing claims or taking orders.
Related guides
Spend the next 14 days proving the inputs
List your last 10 completed sales and remove discounts or support purchases that do not represent the intended market. Calculate first-purchase contribution from actual delivery. Trace the buying path and fix the first point where suitable buyers need rescue.
Write an allowable acquisition cost that leaves money for overhead and profit. If any input remains guessed, use the next 14 days to make direct sales and observe it. Only then define one bounded campaign, its affordable loss and the decision you will make when it ends.
Frequently asked questions
What if all my competitors are advertising already?
Treat that as evidence that they value the channel, not proof that it is profitable for you. They may have lower costs, stronger repeat purchase, better conversion, established trust or a budget intended to defend market share. Some may also be losing money.
Inspect what audience and offer they appear to use, but calculate your own allowable acquisition cost from your own contribution. Do not bid simply to signal legitimacy. If competitors make attention too expensive for your economics, choose a narrower audience, a different route or an offer with more contribution rather than copying their spend.
Can advertising help with a new local opening?
Yes, when the underlying service has already sold and the opening has a real date, catchment and capacity. Use advertising to tell suitable nearby buyers what is available, not to discover whether anybody wants the category. Calculate contribution per booking, the number of slots and the maximum cost to fill them.
Separate opening curiosity from repeat demand. Planning, licensing and promotional rules vary by premises, council and sector, while consumer offers need accurate conditions. Verify local requirements and obtain qualified advice before announcing claims, discounts or availability you may not be able to honour.
Should a seasonal business advertise before its first peak?
Only after finding credible evidence that the offer, price and buying window fit the season. That evidence might come from advance orders, direct selling, prior industry experience or a smaller event with comparable buyers. Work backwards from the decision date so the campaign does not begin after customers have already booked alternatives.
Set stock or capacity limits and an affordable loss because you may not get another clean test for a year. Seasonal urgency does not excuse unsupported claims. Consumer, event and cancellation rules vary, so check current requirements for your market.
Can I advertise a product before the stock arrives?
Yes, only when you state availability and delivery timing accurately and can meet the obligations created by taking money. Distinguish an expression of interest from a paid pre-order. Explain cancellation, refund and delay terms before purchase, and keep funds and fulfilment risk under control.
Do not advertise an estimated arrival as guaranteed or continue taking orders after a known delay makes the promise false. Consumer and distance-selling rules vary by country and product. UK founders should check current official guidance and seek qualified advice for their particular pre-order arrangement.
Should I hire an advertising specialist for the first campaign?
Only after you can give them a proven offer, contribution limit, buyer definition, conversion event and affordable test budget. A capable specialist can improve campaign design and execution, but cannot decide your economics or create evidence you have withheld. Start with a narrow scope and evaluate contribution, not activity reports.
Clarify ownership of accounts, creative work, data and cancellation terms in writing. If you are still changing the product after every conversation, pay for customer learning before campaign management. Contract, privacy and advertising responsibilities remain yours even when delivery is outsourced.
What if my advertising makes sales but still loses money?
Stop scaling and calculate the loss per acquired customer. Separate media cost, discount, fulfilment, returns, sales time and repeat contribution supported by actual behaviour. A campaign can validate demand while proving that the acquisition model is uneconomic.
Raise contribution, reduce acquisition cost or improve the buying path one change at a time, then retest within a fixed loss. Do not justify current losses with an assumed lifetime value. If no credible change can bring acquisition below your allowable cost, the correct decision is to stop the channel even though it produces orders.
Comments (0)