A paid media strategy decides which channels you buy, what you are willing to pay for a customer, how the account is structured, and how budget moves between testing and scaling. It starts from your margin rather than from a budget, because the margin sets what a customer can cost before growth stops being worth having. This covers how to set the target, choose channels, structure spend, and scale without breaking what works.
Start from the margin, not the budget
Paid media plans usually start with a number someone is willing to spend. That is backwards. The first decision is what a customer is allowed to cost, and that comes out of the margin. Work out what survives from an order after product cost, shipping, fulfillment, fees and returns. Whatever is left is the room you have for acquisition and profit. Set the target cost per customer inside that room, and the whole plan has a spine.
This single number reframes everything downstream. Channels are chosen by whether they can deliver customers inside the target, campaigns are judged against it, and scaling continues while it holds. Without it, an account drifts towards whatever metric looks best that week. The details of the margin layers are in what is contribution margin (CM1, CM2, CM3), and the measurement side is in how to measure marketing ROI.
Capture demand before you create it
Paid channels do two different jobs. Some capture demand that already exists: search and shopping put you in front of someone actively looking for what you sell. Others create demand: social and video interrupt someone who was not looking and persuade them to want it. Both work. They do not work in the same order.
For most brands, capturing existing demand is where to start, because intent is already there and the feedback is fast and cheap to read. If people searching for your category will not buy from you at a workable cost, that is a signal about the offer, the price or the store, and it is worth knowing before you spend on persuading strangers. Once capture is working and saturating, creating demand is how you grow beyond the ceiling of what people already search for. Brands that invert this order often spend heavily teaching a cold audience to want a product that was not yet converting the warm one.
Structure: keep it simpler than feels natural
Account structure has drifted from craft towards restraint. Modern platforms optimize with machine learning that needs volume of data per campaign to learn. Splitting spend across dozens of tiny campaigns and ad sets starves each of the signal it needs, so nothing learns and everything underperforms. Consolidating spend into fewer, better-fed campaigns usually beats a beautifully segmented account.
The parts that still deserve separation are the ones you would decide differently about: prospecting apart from retargeting, since they perform differently and you would set different targets; markets or languages apart, since you would judge them separately; and product lines with materially different margins apart, because a single target CAC across products with 60% and 30% margins hides losses inside an average that looks fine.
Split budget between proving and scaling
An account needs a share of budget doing work that will not pay off this month. Testing produces the next winner; scaling harvests the last one. Run only the harvest and the account decays as creative fatigues and audiences saturate, which is why an account that was performing six months ago is quietly getting more expensive today. Run only tests and you never compound.
The practical rule is to keep the majority of spend on what has proven itself and a deliberate minority on what might replace it, and to protect that minority when results dip. The instinct in a bad month is to move everything onto the proven winner, which is exactly when the pipeline of future winners dies. That discipline of testing continuously, scaling what works and cutting what does not, is the loop the whole performance approach runs on.
How to scale without breaking it
Scaling is where good accounts get wrecked. Doubling the budget on a winning campaign overnight resets the platform’s learning and often turns a profitable campaign into an expensive one. Increases in steps, with time between them to let performance settle, hold up better.
Watch what happens to cost per customer as spend rises, because it will rise too. The platform buys the cheapest conversions first, so more budget means reaching further into less likely buyers. The question is whether cost per customer is still inside the target when you get there, and the honest answer usually arrives a week later, not the same day. Frequency is the other early warning: when the same people are seeing the ads repeatedly, the audience is saturating and the creative is aging, and no bid adjustment fixes that. New concepts do.
There is also a ceiling worth respecting. Every channel has a point where the next customer costs more than they are worth, and pushing past it because the growth target says so is how brands buy revenue at a loss. When you hit it, the next move is a new channel, a better converting store, or a better offer, rather than a higher bid.
When paid media is the wrong answer
Paid media amplifies whatever the business already is. If the store converts poorly, ads pay full price for visitors who leave, and the fix lives in conversion. If customers never come back, acquisition becomes a treadmill that costs more every month, and the fix lives in retention. If the price leaves no margin, no amount of media buying rescues it.
This is why the first question is whether acquisition is the constraint at all. Spending on ads for a problem that was never about traffic is the most common and most expensive mistake in the category, and it is what a diagnosis exists to prevent.
Frequently asked questions
How do I set a target CAC for paid media?
Work out what survives from an order after product cost, shipping, fulfillment, fees and returns. That leftover is the room for acquisition and profit. Set the target cost per customer inside it, and judge campaigns against that rather than against a borrowed ROAS benchmark.
Which paid channel should I start with?
Usually the ones that capture existing demand, like search and shopping, because intent is already there and the feedback is fast. Once capture saturates, demand-creation channels like social and video grow you past the ceiling of what people already search for.
How should I split budget between testing and scaling?
Keep the majority on what has proven itself and a deliberate minority on what might replace it. Protect the testing share when results dip, because cutting it is how the pipeline of future winners dies.
How do I scale a winning campaign?
Increase in steps with time between them, rather than doubling overnight, and watch cost per customer and frequency as spend rises. When cost per customer leaves the target, the answer is a new channel, a better converting store or a better offer rather than a higher bid.
Spending more and earning less?
We build paid media plans that start from your margin and scale only what holds. Request a strategy call and we will look at your targets and your account.
