Growth feels like proof. These five numbers tell the truth.

Maya has a number she loves. Ten thousand sign-ups for Tempo, her £10-a-month habit-tracking app, and growth of twenty percent month on month. It leads every investor update. It is the screenshot she posts, the line going up and to the right. It feels like proof the thing is working. It isn't proof of anything yet, and the first time Maya actually sits down and works out five other numbers, the picture changes completely. This issue walks that hour with her, one number at a time, so you can do the same for your own startup.

None of this needs a finance background. If you can run a calculator and be honest with yourself, you can do all five. I have kept Tempo deliberately simple, a single £10 plan, so the arithmetic stays clear. Swap in your own figures as we go.

The number that lies.

Sign-ups, downloads, page views, followers. These are the numbers that feel like progress, and founders reach for them because they only ever go up. The problem is that they say nothing about whether you have a business. A sign-up that never comes back cost you money and returned none. Ten thousand of them is not ten thousand pieces of evidence; it is one number, repeated, dressed up as traction. The five that follow are harder to face precisely because they can go the wrong way. That is what makes them worth trusting.

Number one: CAC, what a customer costs.

Customer Acquisition Cost is what you paid to win one paying customer. You work it out by adding up everything you spent getting customers over a period, then dividing by the number of paying customers that spending brought in. Last month Maya spent £16,000 on ads and another £4,000 on a marketing freelancer and her tools. That £20,000 brought in 400 new paying subscribers.

Two mistakes catch people here. The first is dividing by sign-ups instead of paying customers, which flatters the number and hides the truth. The second is counting only the ad spend and quietly ignoring the salaries, the agency, the tools and the discounts. If a person worked on winning customers, their cost belongs in the top of that sum. Be strict with yourself, because everyone else who looks at this later will be.

The two inputs hiding inside the next numbers.

Before we can value a customer, we need two things. The first is ARPU, average revenue per user, which for Tempo is simply the £10 a month they pay. The second is gross margin, the share of that £10 you actually keep after the direct costs of serving them. Payment fees, hosting and support cost Maya about £2 per user each month, so she keeps £8. That is a gross margin of 80 percent, and £8 of gross profit per user per month.

This is the step first-timers skip, and it matters more than any other. You value a customer on the profit they bring, not the revenue. Using the full £10 would overstate everything that follows by a quarter. Use the £8.

Number two: retention, the one that decides everything.

Now the number underneath all the others. Retention is the share of customers who are still with you as time passes; churn is its mirror, the share who leave. Of every 100 people who subscribe to Tempo, about 90 are still paying after month one, and Maya loses roughly ten percent of whoever remains each month after that. That is monthly churn of ten percent.

From churn you get the single figure the value calculation needs: how long the average customer stays. Divide one by the monthly churn rate.

Plot retention month by month and you get a curve, and that curve is the closest thing a startup has to a lie detector. At ten percent churn, 100 subscribers become about 53 after six months and 28 after a year. If the curve keeps sliding towards zero, you have a leaky product. If it falls and then flattens, holding a stable group of people who simply keep using the thing, that flattening is what product-market fit actually looks like on a chart. It is a far better signal than sign-ups, because it is people voting with their habit, month after month.

Number three: LTV, what a customer is worth.

Lifetime Value is the total gross profit one customer brings across their whole life with you. With the three ingredients in hand it is one line: gross profit per month, times the number of months they stay.

You will see LTV written as a single formula too: ARPU times margin, divided by churn. It is the same sum. Ten pounds times 0.8, divided by 0.10, is £80. However you write it, the number says one customer is worth eighty pounds of profit to Tempo over their life. Hold that against what one costs to acquire, and you have the number that matters most.

Number four: the LTV:CAC ratio, the scoreboard.

This is the one investors ask for and the one that tells you whether the model works at all. Put what a customer is worth over what they cost.

The rough rule the industry uses is that you want this above 3:1. Below one, you lose money on every customer, faster the more you grow. Between one and three, you make something, but not enough to cover the rest of the business, the salaries and the office and the years of runway. Much above five and the surprising read is that you are probably underspending on growth, leaving customers on the table you could afford to go and win. Maya is at 1.6. Every customer is worth less than double what she pays for them, and that has to carry the whole company. The line going up and to the right was hiding this.

Number five: payback, the cash-flow reality.

The ratio tells you if the model works eventually. CAC payback tells you how long your cash is underwater first, which for a startup with a finite bank balance is often the more urgent question. It is the CAC divided by the gross profit a customer brings each month.

So Maya spends £50 today and waits more than six months to get it back, all while churn is quietly removing customers from the base she is trying to build. Under twelve months is the usual comfort line; the shorter it is, the faster you can recycle cash into winning the next customer instead of raising more to bridge the gap. Now put all five numbers next to each other, because the fix only becomes obvious when you do.

Diagram · the five numbers, before and after one fix

Retention is the lever. Halving churn doubled the value of every customer Tempo already had, and turned a failing model into a working one, without touching the ad budget.

The trap: don't chase CAC, fix the bucket.

Faced with a 1.6:1 ratio, the instinct of almost every first-time founder is to make acquisition cheaper. Better ads, a sharper funnel, a referral scheme. And it helps: halve CAC from £50 to £25 and the ratio jumps to 3.2:1. But there is a bigger, slower lever hiding in the retention number, and most founders reach for it last when they should reach for it first.

Watch what happens if Maya leaves acquisition completely alone and instead fixes the reasons people leave. A better first week in the app, a nudge before the habit breaks, a reason to still be there in month three. Say that halves churn, from ten percent to five. Average lifetime doubles from 10 months to 20. LTV doubles from £80 to £160. And the ratio moves to 3.2:1, from failing to passing, with the same £50 CAC she had before. She spent nothing extra to acquire anyone. She just stopped losing the people she already had.

This is the mistake that burns real money. A founder sees an underwater ratio, assumes the fix is cheaper growth, and pours the next raise into acquisition. All that does is bring more customers into a product that cannot hold them, so you pay again and again to refill a bucket with a hole in it. Cheaper acquisition and better retention both move the ratio. Only one of them also builds a company that compounds, because retained customers stay, refer, and cost nothing to keep. Fix the bucket first.

What these numbers are really for.

For a later-stage company, these five are a dashboard for running the machine. For a startup still hunting product-market fit, they are something more useful: an honest instrument panel for whether you have found it yet. A great ratio built on a retention curve that flattens means you have something real, and pouring fuel on it is the right call. A great-looking growth chart sitting on a curve that slides to zero means you have a leak the growth is papering over, and more fuel just burns faster. The numbers do not care how the line looks on the investor update. That is exactly why they are worth more than the line.

Work them out once and they stop being finance jargon and start being a map. You will know whether to spend on growth or fix the product, whether you are ready to raise or need another quarter of proof, and which single lever moves your business the most. For most startups before product-market fit, that lever is retention, and these five numbers are how you prove it to yourself.

How we use this at Product Pieces.

Plenty of founders come to us with Maya's exact problem: a growth chart they are proud of and a quiet worry that it might not mean what they hope. The free Diagnostic is built to find that out, to read whether the piece your product function is missing is the senior judgment to tell real traction from a leak, and to point the next pound of spend at the lever that actually moves the business. Growth is easy to feel and hard to trust. Getting the five numbers straight, before the next raise goes into the wrong end of the funnel, is the piece we plug in.