Before you spend on paid social, paid search (SEA/PPC) or any performance marketing, you need four numbers: your cost of goods sold, the margin left after it, the runway you can fund before the account must be profitable, and the lifetime value of a customer. From those we reverse-engineer the customer acquisition cost the business can afford, and the "scaling CPA" just above break-even where spend can grow. Without them there is no shared definition of success, which is how campaigns run for three months toward a target nobody agreed on.
At R17 Ventures AG we ask for these numbers before we agree to a single KPI. We are a digital performance agency paid on results, so the math has to work for you before it can work for us. If it does not, we will say so, including when the answer is that paid ads are not the right investment for your product.
After four years of meeting prospects for R17, the thing that still surprises us is how many businesses want to scale with paid ads without knowing what a customer costs them, what a customer is worth, or how long they can fund the gap between the two. None of that is a reason not to grow. It is the reason to start with the numbers.
There are four numbers a business should know before engaging an agency or launching paid media itself: cost of goods sold, margin, runway and customer lifetime value. The first is the one every business owner should be able to answer immediately; the others follow from it.
|
Number |
What it includes |
What it tells us |
|---|---|---|
|
Cost of goods sold (COGS) |
Production, shipping, handling, payment fees, anything that leaves with the product or service. |
The floor. Everything above it is what you have to work with. |
|
Margin |
What remains per sale after COGS, before advertising. |
How much of each sale can go toward acquiring the customer. |
|
Runway |
How long the business can fund the account before it has to be profitable: six months, eight, ten. |
Whether there is enough budget and time for the campaign to iterate to stable results. |
|
Lifetime value (LTV) |
Revenue from a customer across repeat purchases, subscriptions, upsells and additional services. |
Whether a first purchase needs to be profitable, or whether acquisition can be paid back over time. |
With these, you can work out your own profitability, and we can work out what we can promise you. Without them, both of us are guessing.
Enough to fund iterations, not just a test. Test budgets rarely produce the results anyone hoped for, because the results come from the rounds of learning after the test, and those take time. Meaningful, stable performance can take up to six months on a new account.
So the runway question has two parts. How long do you expect to wait for the first meaningful results, and can the business sustain itself for six, eight or ten months before hitting the thresholds that matter? If the honest answer is three months and the account needs six, the conversation should be about timing or budget, not about launching anyway.
A target cost per acquisition should be reverse-engineered from your costs and the return you need, not taken from a benchmark. Start with the cost of goods, add the margin you need to keep after all advertising expenses, and what remains is the maximum you can pay to acquire a customer. The KPIs for the campaign are then built to land there.
This is the approach we use most with clients who are new to digital, and it does something a benchmark cannot: it ties the campaign target to the economics of the business rather than to the industry average. There is no golden rule that says this ROAS or that CPA means success. It varies by industry, by product and by where the customer is in their lifecycle.
Once the break-even CPA is known, we set a scaling CPA: the acquisition cost just above break-even at which the client is making money and we can push spend aggressively. As long as the campaign holds that number, growth is funded by the growth itself.
Customer lifetime value decides whether paid media makes sense because, for most products, the first purchase does not cover the cost of acquiring the customer. The value sits in the long-term relationship: repeat orders, a subscription that runs for 12 to 48 months, maintenance, upsells or additional services on top of the original purchase.
SaaS and telco show this most clearly. Both carry high acquisition costs that pay for themselves because the customer stays. A common SaaS rule of thumb is to spend three to six months of subscription revenue to acquire a customer who will be around for a year or more. That only works if you have the runway to carry the upfront cost until it pays back.
At the other end, an e-commerce product with an average order value under $100 in Europe, the UK or the US is hard to make profitable on the first purchase. It can be done, and we have done it, but you should plan on it not being the case. And if the product is a one-off purchase at €30 with no repeat and no upsell, the honest question is whether paid ads are worth it at all. Sometimes the answer is no, and we would rather say that than take the budget.
Most of the time, though, lifetime value is where the upside hides. Once a customer has bought, you hold their contact details. Re-engaging them by email or SMS costs a fraction of acquiring a new customer. Most businesses pour everything into the first acquisition and forget the nurturing afterward, which is the cheapest growth they are not taking.
By agreeing the definition before the campaign starts. The how matters less than the fact that it is shared, between the agency and the client, and between the client's own stakeholders. There is nothing worse than a campaign running for three months before anyone discovers that nobody agreed what success looked like.
That definition is a North Star: what good looks like three and six months into the work. For one business it is a profitable account at a target CPA. For another it is a tested hypothesis, a minimum viable product that has proved the market exists. Both are legitimate. What is not legitimate is leaving it unsaid.
This is also why we are careful about what we optimize toward. We could get you more leads; that is not hard. If they do not convert, it does not help you commercially, and after a while it does not help us either, because our bonus rides on your growth. Quality over quantity is not a slogan here. It is how the model works.
Beyond budget and the processes to fulfill orders or call leads, a go-to-market needs three things: a product you can explain in a sentence, a clear picture of who buys it and why, and one person with the authority to decide. Roughly eight out of ten go-to-market requests that reach us are missing at least one.
“If you are not able to explain in simple words why someone should buy the product, probably your product is not ready to go to market.”
— Mariangela Coco, CEO, R17
Everybody is not a target. You need a clear profile of the person who could buy, and the specific situations in which the product is useful, because different people respond to different messages. We test hooks and topics at scale across a campaign, but we need the real use cases first to build the storyline around them.
Decision-making matters just as much. When one stakeholder briefs us and another, often the CEO or CRO, has a different vision, the work gets done twice and decision speed collapses. Three, four or five review rounds on a creative is not a process; it is a campaign standing still. Before launch, we ask who signs off, we ask for a single point of contact, and we ask that the mediation between internal stakeholders happens before the brief reaches us, so we are implementing a decision rather than sitting in the middle of one.
A scaling CPA is the cost per acquisition set just above the break-even point, at which the client is making money on each new customer and the campaign can increase spend aggressively. It is derived from the business's own costs and margin, not from an industry benchmark.
Up to six months on a new account. Results come from iterations after the initial test, and the runway needs to cover that period rather than a short test budget alone.
A common rule of thumb is three to six months of subscription revenue, justified by customers who stay 12 months or longer. The exact figure depends on churn, margin and the runway available to carry the upfront cost.
Yes, but it is the exception. Below a $100 average order value, first-purchase profitability is hard to reach in Western markets, so the model usually depends on repeat purchases, upsells and email re-engagement. A one-off product with no repeat potential may not justify paid ads at all.
Cost of goods sold, margin, runway and lifetime value, plus a shared definition of what success looks like at 90 days. Everything else, including the target CPA and ROAS, is derived from those.
Not sure your numbers add up for paid media? Send them to us and we will tell you what they support, including if the answer is "not yet" or "not this way."
R17 is a digital performance agency, paid on results: we work out what will actually move your numbers, then run the campaigns to get there.