Codnity Growth · Media buying
Media buying is deciding where your money goes, how much of it, and when to stop. The work that changes results is not creative volume. It is knowing what a new customer actually costs you after product, shipping and returns, then cutting the channels that never deliver one at that price.
Where the budget actually goes
These four are what we find in nearly every account we inherit, and none of them is a bidding problem. They are all measurement and decision problems.
A strong return on ad spend can still lose money once product cost, shipping and returns come out of it. Until the number you optimize toward is margin on a new customer, the algorithm is being rewarded for buying you the cheapest orders rather than the profitable ones.
Healthy paid accounts shrink as often as they grow. If every campaign that ever launched is still running, nobody is making decisions. They are adding. The savings from stopping usually outweigh the gains from optimizing.
You pay for the click and then ask the visitor to find the product themselves. This single mismatch wastes more budget than any bidding setting, and it is the first thing Codnity checks before touching a campaign.
If a monthly report has no section on what did not work, it is a sales document. You are being managed rather than informed, and the decisions that would have saved money are not being put in front of you.
Media buying decides where budget goes, how much of it, and when a channel is switched off. For a brand selling physical product that means the product feed, the campaign structure and the offer, tested against each other until the cost of a new customer is a number you can plan against.
Budget is set by how fast you need data, not by a percentage of revenue. A channel needs enough weekly orders to tell a real result from noise, so the honest floor is whatever buys that. Below it you are paying to guess.
ROAS measures revenue against ad spend and says nothing about whether the order made money. Contribution margin after product cost, shipping and returns is the number that decides whether to scale. A brand can hold a strong ROAS and still lose on every order.
| ROAS alone | Margin after costs | |
|---|---|---|
| Counts product cost | No | Yes |
| Counts returns | No | Yes |
| Separates new from repeat | No | Yes |
| Safe to scale on | No | Yes |
Shopping and search capture people already looking for the product, so they come first when the demand exists. Meta creates the demand and suits a product that needs showing rather than finding. Performance Max sits over both and needs a clean feed before it is worth switching on.
Google matches a shopping query against the feed, not against the campaign. Titles, attributes and images are what get compared, so a feed with thin titles and missing attributes loses auctions the bidding can never win back. Feed work is usually the cheapest performance gain available.
Platform-reported conversions overstate, and analytics under-reports. Codnity measures on the total that cannot be double counted: orders and enquiries against total spend for the period, reconciled to what the business actually banked. Channel numbers inform decisions; the blended number settles them.
Paid media multiplies whatever the page already does, which is why the order matters more than the budget. Sending traffic to a page that does not convert buys a faster version of the same result, so when the page, the feed or the tracking is the constraint, Codnity says so before quoting for management.
Where this stops
The second column is the one worth reading.
If your budget is too small to gather conversion data in a month, and you have no product or landing page worth the traffic, paid is the wrong first move and we will say so on the first call. Answer engine work and one good page buys you more per euro at that stage, and both are cheaper than us managing a budget too small to learn from.
We also expect the first six weeks to look worse before better, because fixing conversion definitions usually reduces the numbers the platforms report. That is the figure becoming honest rather than the campaign failing, and it is easier to hear now than in the first report.
Also in Growth
Next step
Give us your average order value, your margin on it, and what you are spending now. We come back with whether the maths works, which channel we would run first, and what we would switch off in week one.
Media buying FAQ