A store can show 4x ROAS in Ads Manager and still lose money on every new customer. The gap is product cost, shipping, payment fees, returns and discounts, none of which the ad platform sees. So we work as a digital marketing agency for ecommerce brands on one number: contribution margin, the profit left on an order after the product, the fulfillment and the ads that sold it. Meta ads, Google Shopping, SEO and the store itself are run together against it.

The short version
  • For Shopify and WooCommerce brands, mostly DTC, selling in the US and UK.
  • Meta ads to create demand; Google Shopping and Search to catch the demand that exists.
  • A feed Google can read, creative tested by concept, and a store that's fast on a phone.
  • The Meta Pixel plus the Conversions API, checked against real orders.
  • Monthly services run $500-$2,000 a month, month-to-month. Store builds run $1,200-$15,000, fixed price.

Why ecommerce marketing stops paying

What you see, and what's behind it
  • ROAS looks healthy, the bank balance doesn't. ROAS ignores product cost, shipping, fees and returns, and credits the ads with repeat buyers they didn't find.
  • More spend, flat sales. Every ad is a variation on one idea, and the audience tires of it.
  • Shopping clicks on the wrong products. Supplier titles, missing attributes, and one bid across very different margins.
  • Ads Manager and the store disagree. A browser-only Pixel losing events, or Purchase firing twice on the thank-you page.
  • Visitors arrive and leave. A slow theme, a stack of apps and a checkout with too many steps.

Contribution margin, not ROAS

ROAS doesn't know what the product cost. One order from an illustrative store, not a benchmark:

Per orderAmount
Order value$80.00
Product cost-$28.00
Shipping and packing-$9.00
Payment fees-$3.00
Returns allowance-$4.00
Margin before ads$36.00
Ad cost at 3x ROAS-$26.67
Contribution margin$9.33

Break-even ROAS for this store is 2.22x: $80 divided by the $36 left before ads. At 3x it keeps $9.33 an order. At 2x, a ROAS that looks respectable on a dashboard, it loses $4 on every sale. Whichever digital marketing agency ecommerce founders hire, the first question should be what a new customer is allowed to cost. The answer comes from this table, per product line.

What we do for ecommerce brands

Pick by the problem

How the pieces fit together

1

A margin model first

Cost, shipping, fees and returns for each product line. It sets break-even ROAS and the most a new customer can cost.

2

Tracking that matches the orders

The Pixel in the browser and the Conversions API from the server, both sending Purchase. Shopify notes that a browser-based ad blocker can prevent the Meta pixel from sharing data, while server-to-server data can't be blocked that way. Meta counts the two copies once when the event name and event ID match and they arrive within 48 hours of each other. More in Conversions API setup and Pixel not tracking Shopify purchases?.

3

A feed Google can read

In Shopping, the feed is the targeting. Google says it uses product data to match products to the right queries, and that missing or wrong data can stop ads and free listings from showing. Titles get rebuilt around how people search, and bids are split by margin.

4

Creative tested by concept

New reasons to buy, such as the problem, the proof or the offer, each tested on its own and judged on purchases. Ecommerce creative testing sets out the budget split.

5

Meta for new demand, Google for existing demand

Prospecting kept apart from retargeting, so you see what a new customer costs. Search and Shopping catch people already looking, with brand searches kept out of Performance Max's credit.

6

A store that closes the sale

Fast product pages on a phone, a short checkout, an order bump and one abandoned-checkout reminder. Clicks but no sales? separates store problems from ad problems.

7

A weekly reconciliation

Platform-reported sales against store orders, and contribution margin by channel.

Proof: a checkout and two ad accounts

We haven't published a physical-product store case study, so here is what we have, labeled honestly.

From a custom checkout we built for a coach and author's books
  • Order bump. 68% of single-book checkouts take the second book as well when it's offered.
  • Abandoned checkout. An order left unpaid for 45 minutes gets one reminder, which skips anyone who has paid since.
  • Server-side tracking. 100% of paid orders since August reach Meta from the server, deduplicated against the Pixel.

Those rates come from the store's own database. The site sells books and coaching, not physical products, and by the client's choice the case study publishes no revenue or order counts.

From two Meta ads accounts: climbing gyms that sell sessions through their own online checkout
  • One had run on a link click objective and never reported a sale. With purchase tracking and a sales objective, cost per booking fell from £28.82 to £5.13 on essentially the same budget. ROAS was 1.97x in the first phase and 2.95x across ten months.
  • The other had returned £0.82 for every £1 under its previous agency. Eight months after the handover it returns £2.99, and cost per booking is down from £38.78 to £9.44.

Both case studies, the rebuild and the handover, sell bookings rather than parcels, but the mechanics carry over to a store.

What it costs

ServicePriceTerms
Ads management (Google, Facebook or Instagram)$500-$2,000 a monthMonth-to-month, no setup fee, ad spend paid straight to the platform
SEO and AI search visibility$500-$2,000 a monthMonth-to-month, no long-term contract
Website, landing page or store build$1,200-$15,000, fixed priceAgreed in writing; 50% to begin, balance on delivery

Entry plans start at $500 a month. Full details are on our pricing page.

Who we're not for

Not the right fit
  • Stores whose margin can't carry paid acquisition at any ROAS the market will give. We'll show you that math first.
  • Brands that want ROAS reports but won't share costs and order data.
  • Anyone measuring success in followers, reach or add-to-carts.

Guides for ecommerce brands

Questions store owners ask a marketing agency

What does a digital marketing agency for ecommerce do?

It runs the channels that bring a store its customers, usually Meta ads, Google Shopping and SEO, and fixes tracking and the store when those are the real problem. It should report contribution margin and the cost of a new customer, not ROAS alone.

Should my store start with Meta ads or Google Ads?

Google first if people already search for what you sell, because Shopping catches that demand. Meta first if the product needs showing before anyone wants it.

What is a good ROAS for an ecommerce store?

One above your break-even, which is order value divided by the margin left before ads. A store keeping $36 of an $80 order breaks even at 2.22x. There's no universal good number.

Do I need the Conversions API if I already have the Meta Pixel?

In most stores, yes. Ad blockers stop some Pixel events, and the server copy fills the gap. Both should send the same event ID so Meta counts each purchase once.

Do you build Shopify and WooCommerce stores?

Yes, at a fixed price, with tracking wired in before launch, and we fix existing stores when speed or checkout is costing sales.

How much does ecommerce marketing cost with you?

Ads or SEO management runs $500-$2,000 a month, month-to-month, with ad spend paid straight to Meta or Google. Store builds are fixed-price, from $1,200, and most run $1,200-$15,000.

How should I choose an ecommerce digital marketing agency?

Ask what a new customer is allowed to cost in your store, and how they check the platforms' numbers against your orders. An agency that names a ROAS target before asking about margins hasn't done the math.

Want to see what a new customer really costs you, channel by channel? Book a strategy call and bring your margins and last quarter's orders.