Unit economics first
What a customer is worth, what you can afford to pay for one, and what margin survives at each spend level. The number everything else is judged against.
PPC & Paid Media
PPC & Paid Media
Service / 04
Paid ads are arithmetic, not magic
If a customer is worth more than they cost to acquire, you scale. If they aren't, no amount of targeting fixes it. Most accounts we inherit have never had that number calculated properly — so we start there.
The problem
The dashboard says it's working
Platform-reported ROAS looks healthy. The agency's monthly report is green. And yet the bank account doesn't reflect any of it, because the numbers being celebrated were never the ones that mattered.
Ad platforms report on their own performance, with their own attribution window, counting conversions they take credit for. Measure against revenue instead of platform-reported conversions and the picture usually changes — sometimes a lot.
An ad account optimised against the wrong number will get very efficient at producing the wrong result.
The second problem is quieter: nobody worked out what a customer is actually worth. Without that, there's no ceiling on acceptable acquisition cost — so "good ROAS" becomes whatever the platform says it is, and scaling decisions get made on vibes.
What's included
Six parts of running an account
Campaign management is one of them. The rest decide whether the management is pointed at anything useful.
Tracking & attribution
Pixels verified, events deduplicated, server-side tracking where browser signals fall short. Without this, everything downstream is guesswork.
Campaign build
Structure that lets the platform learn without wasting budget — sensible consolidation, clean audiences, exclusions that actually exclude.
Creative testing
New angles on a continuous cycle, not a set-and-forget launch. On Meta especially, creative is the targeting — it decides who the algorithm finds.
Landing page & funnel
The cheapest improvement in most accounts isn't the ads — it's the page they land on. We'll tell you when that's where the problem is.
Reporting against revenue
Weekly, measured against your numbers rather than the platform's. Including the campaigns we stopped and why.
The number that decides everything
Can you afford to advertise?
Before any campaign runs, this is the arithmetic. Put your own numbers in — it takes ten seconds and it's the same calculation we do in the first week of any engagement.
What a typical order is worth
After product and shipping costs
Ad spend ÷ orders
Profitable, with room to scale
You keep 425 per order after ad costs. Break-even cost per sale is 1,125.
Break-even ROAS 2.22×
Simplified deliberately — it ignores returns, repeat purchase and overheads, all of which move the answer. If the result is close to break-even, those details decide it, and that's the conversation worth having on a call.
How it runs
The first ninety days
Week 1 Audit
The numbers, before the ads
Existing account reviewed, tracking checked, unit economics worked out. If the maths doesn't support paid acquisition, this is where we say so — before you've spent anything with us.
Week 1–2 Foundations
Fix the measurement
Pixels verified, conversion events deduplicated, server-side tracking where needed. Running campaigns on broken tracking just buys expensive misinformation.
Week 2–4 Launch
Small, then structured
Campaigns launched at a budget that can afford to learn, with targets agreed beforehand. Early spend is buying information, and we'll say which is which.
Week 4–8 Testing
Find the angle that works
Continuous creative testing against the same offer. Most accounts are won or lost here — one message usually outperforms the rest by a wide margin, and it's rarely the one anyone predicted.
Month 2–3 Scale
Increase spend where the maths holds
Budget goes up only while cost per sale stays under the ceiling we calculated. Scaling past that point buys revenue at a loss, which is easy to do and hard to notice.
What we won't tell you
That we can guarantee a ROAS. Nobody can — it depends on your margins, your offer, your market and your competitors' budgets, most of which neither of us controls. An agency promising a specific return is describing a best case as though it were a forecast.
We also won't tell you paid is always the answer. If your margins are thin and your average order is small, the arithmetic above may simply not work — and we'd rather show you that in week one than bill you for six months of finding out.
And sometimes the ads are fine. If the traffic converts badly, the problem is the landing page or the offer, and spending more on ads makes it worse. We'll say so even though fixing ads is what you hired us for.
What you get
Every week, without asking
- Unit economics modelled at the start — your real ceiling on cost per sale.
- Weekly reporting against revenue, not platform-reported conversions.
- Fee and ad spend separate on every invoice, with no markup on spend.
- Full account ownership — your business manager, your pixels, your data.
- Creative test log showing what ran, what won and what we killed.
- Direct access to the strategist managing the account.
Questions
About paid media specifically
We won't quote one before seeing your numbers, and we'd be cautious about anyone who does. What we will do in week one is calculate your break-even ROAS — the figure below which you're losing money — and agree a target above it. That's a number grounded in your margins rather than our marketing.
A fixed monthly fee, separate from ad spend, both itemised on every invoice. We don't take a percentage of spend — that model pays an agency more for spending more, which is a conflict of interest exactly when you need honest advice about scaling back.
Two constraints. Below roughly 50 conversions a month, the platforms can't optimise properly, so the spend has to support that. And below a certain level our fee eats the returns, in which case you're better off running it yourself. We'll tell you honestly which side of both lines you're on.
Signal in two to four weeks — enough to see which direction things are moving. A real verdict takes a full purchase cycle, longer for considered purchases. Early results are also unreliable in both directions: a strong first fortnight isn't proof any more than a weak one is.
Yes — static, video and copy, tested continuously. If you have a team or existing assets we'll work with those instead and focus on angles and testing. On Meta in particular, creative does more work than targeting, so this isn't a side task.
You do. Ad accounts, pixels, audiences and domains sit in your business manager with you as owner, and we're added as a partner. Some agencies run clients through their own accounts, which means the pixel history and audiences stay behind when you leave — worth asking anyone you talk to.
Also available
Pairs well with
Start with the numbers
Send us access and we'll audit what's running, check whether the tracking is telling the truth, and model your real ceiling on cost per sale. You keep the findings whether or not you work with us.
No pitch deck, no pressure. Just a conversation about your numbers.