Traffic is easy. Margin is the hard part.
An online shop can grow revenue and lose money at the same time, and most reporting is arranged so nobody notices for two quarters. Ecommerce marketing done properly is a conversation about contribution margin that happens to involve advertising.
- Contribution margin
- Search, shopping, paid social, email
- A product-level profitability read
Revenue is the metric that hides the problem.
Platform dashboards report revenue and return on ad spend. Neither accounts for cost of goods, shipping, payment fees, returns or the discount that closed the sale. A campaign at a 4x return can be losing money on a product with a thin margin and a high return rate.
The second problem is attribution. Every platform counts the same sale, so the sum of reported revenue across channels regularly exceeds what the shop actually took. Budget then gets allocated on the basis of whichever platform claims hardest.
The third is that a catalogue is not uniform. A handful of products usually carry the profit, a long tail breaks even, and some actively lose money once returns are counted. Spending evenly across them is a choice nobody consciously made.
The first useful number in most ecommerce accounts is not a marketing number at all. It is contribution margin by product.
Once that exists, half the budget questions answer themselves.
Spend against margin, not against revenue.
We start by building a product-level view of what actually contributes: price, cost of goods, shipping, fees, typical discount, return rate. It is unglamorous and it is the foundation for everything after it.
Then the feed. Shopping and Performance Max are fed by product data, and most catalogues ship with titles written for a human browsing a category page rather than for a machine matching a query. Fixing titles, attributes and product types is usually the single highest return piece of work in a shopping account.
Campaign structure follows margin. Products that contribute get budget and bid support. Products that do not get excluded or run at a bid that reflects what they are worth. Email and retention carry the repeat purchase, because acquisition economics only work if someone comes back.
We will recommend excluding products from advertising. That feels wrong to most owners and it is almost always right.
Advertising a product that loses money faster is not growth.
Deliverables.
Product-level margin model
What each product contributes after cost of goods, shipping, fees, discount and returns. Built once, maintained, and used to make every budget decision after it.
Feed audit and optimisation
Titles, attributes, product types, images and availability corrected so shopping campaigns can match the queries the products deserve.
Shopping and Performance Max
Campaigns structured around margin tiers rather than one campaign holding the whole catalogue, with exclusions where exclusion is the right answer.
Paid search for intent
Brand defence, high-intent non-brand terms, and the category queries worth paying for. Separated so you can see which is which.
Paid social for demand creation
Creative built for the feed, tested in sets, aimed at the products that can carry an acquisition cost.
Email and retention
Abandoned basket, post-purchase, replenishment and win-back flows. The cheapest revenue in ecommerce and the most commonly neglected.
Tracking that survives
Server-side where the platform supports it, with order values passed as net contribution rather than gross revenue wherever the data allows.
Reporting in one place
Total spend, total revenue, blended acquisition cost and contribution margin together, so no platform gets to grade its own homework.
How the work runs.
Economics first
Build the margin model. Until it exists, every channel conversation is speculation.
Feed and tracking
Fix the data layer. Campaigns built on a broken feed or broken tracking cannot be optimised, only guessed at.
Structure and launch
Rebuild campaigns around margin tiers and intent, with a clear plan for what each one is for.
Test and reallocate
One variable at a time, with enough volume to mean something. Budget moves towards contribution.
Retention
Once acquisition is stable, the flows that make the second purchase cheaper than the first.
Straight answers.
Do I need a certain revenue level for this to be worth it?
Roughly, there needs to be enough order volume for a test to resolve inside a month. Below that, the honest recommendation is usually to fix the product page, the feed and the email flows and leave paid media alone until volume justifies it.
Which platform do you work with?
Shopify, WooCommerce and most of the common hosted platforms. The marketing work is largely platform independent. What changes is how easily we can get clean data out, and Shopify and WooCommerce are both straightforward.
Will you handle the product photography and copy?
Product copy and feed titles, yes. Photography we would normally bring in, and we will say so rather than quietly subcontracting it and marking it up.
What about marketplaces like Amazon?
We will give you an honest view on whether a marketplace fits your margin, but running marketplace accounts is a specialism of its own and we are not going to pretend otherwise.
How do you handle returns in the numbers?
As a cost against the product that caused them. A category with a thirty percent return rate is a different commercial proposition to one with three percent, and averaging them hides exactly the thing you need to see.
Where this connects.
Tell us what’s next.
Send over the problem and the numbers you have. We will tell you honestly whether we are the right people for it.