Meta Ads for ecommerce brands that need a cost per acquisition, not a content calendar.

Facebook and Instagram are the cheapest attention you can buy and the easiest to waste. We run them as an acquisition channel with a cost target, not a content calendar.

What we do

What our ecommerce Meta Ads work covers.

Creative that is built to be tested

Not one hero video. A batch of angles, hooks and formats, shipped weekly, so the account is always learning rather than defending a guess.

Account structure that can be read

Campaigns arranged so you can tell which audience, which creative and which offer actually did the work.

Audience and exclusion work

Prospecting that reaches people who have never heard of you, and retargeting that stops paying twice for the same customer.

Offer and landing page alignment

The advert and the page it lands on say the same thing. Most Meta accounts fail here rather than in the targeting.

What we measure

How we score Meta Ads for an ecommerce store.

Every component in the Engine is scored out of 100 against its goal. These are the numbers Meta Ads is judged on.

Click through rate
Against the sector average for your category, not against last month.
Cost per acquisition
Blended and channel level, so scale is not hiding behind a good week.
ROAS
Reported next to contribution margin, because return on ad spend alone will happily bankrupt you.
Creative fatigue
Frequency and performance decay per asset, so we replace before the cost climbs.

Where it sits

Meta Ads is one component of a system.

Section 1 · Traffic

Meta Ads sits in Traffic, the first section of the Engine. It brings demand to the business. It cannot fix a page that does not convert, and it cannot fix a customer who never returns, so we score it against the two sections downstream before we recommend spending more on it.
The Ecommerce Engine™ Three sections, each limited by the one before it
Two members of The Obsidian Co. team reviewing an ecommerce account together.

We will not sell you Meta Ads because you asked for it. If the diagnosis says the constraint is somewhere else, we will tell you that instead, and we will say so in week one rather than in month six.

Questions

Meta Ads for ecommerce, asked and answered.

How much should an ecommerce store spend on Meta Ads?

Enough for the account to exit the learning phase and produce readable data, which in practice is $2,000 a month and up. Below that, results swing on noise and no honest agency can tell you what is working.

Do Meta Ads still work for ecommerce in 2026?

Yes, but the lever moved. Targeting is largely automated now, so the account is won on creative volume, offer and the page the click lands on. Stores that keep running one hero asset are the ones reporting that Meta stopped working.

What is a good ROAS for an ecommerce Meta Ads account?

There is no universal number, because ROAS without margin is meaningless. A 3x return is excellent at a 70% margin and loss making at 25%. We set the target from your contribution margin before we spend anything.

Can you run Meta Ads alongside our in-house team?

Yes. We often run acquisition while an internal team keeps organic social. What we will not do is run adverts into pages we are not allowed to look at.

Find out whether Meta Ads is what is actually holding you back.

Four weeks, no fee, and a diagnosis of the whole Engine in week one.

The rest of the Engine

The other components we run.