Alongside the new Shopify store, an older store kept receiving orders. Without a Meta pixel on that old environment, there was no good way to establish what part ads played in those sales. This is how I worked that out — and why my advice was not “more budget”.
Rivali is a family business in luxury household linen and nightwear, with a shop in Ghent. While the new Shopify store was being built, the old store stayed live. Orders still came in there, while that environment had no Meta pixel.
That situation made the ad account hard to assess. A visitor could discover the brand through an ad and order somewhere else later. That was a possible customer journey, but without additional data we could not reconstruct it for every order.
I worked on Rivali's e-commerce. To assess the campaigns, I first had to understand why store sales and ad reporting diverged.
That distinction is deliberate. Not everything I recommended was put in place.
A successful test event shows that this one test comes through. It is not the same as fully watertight tracking.
Platform reporting and total store revenue answer different questions. The ad account reports attributed conversions. The backend records orders. Bringing the sales data together gives a fuller picture of revenue, but in itself it does not prove which sales were caused by ads.
The product mix counted too. An order with bought-in brands could deliver a different contribution than an order with own products. One revenue figure was therefore not enough to set a responsible acquisition budget.
My advice was not to simply scale further. First it had to become clearer what costs an order could carry and how reliable the measurement baseline was. More ad spend was no answer to that uncertainty.
This case describes my approach to fragmented sales data, tracking problems and a range with different margins. It is not proof of profitable scale or of an exactly measured rise in revenue from my changes.
The original source exports are not fully available. That is why we do not publish an exact uplift, corrected ROAS or causal saving here that we cannot back up again.
Get the detailed breakdown: the approach behind the measurement baseline, order reconciliation, margin per product group and the scaling decision.
No exact client figures: the underlying exports are not fully available.
The full Rivali breakdown is on its way. If you do not see it within a few minutes, have a look in your spam.
Do you run a fashion brand that already advertises?
See how Northbound works →
Or book your free analysis now →
Thirty minutes, in your own numbers. You get your contribution margin, your break-even and where it leaks — whether we work together afterwards or not.