Stop Optimizing Platform ROAS Alone

The good ROAS, bad month pattern

Ads Manager can look healthy while finance is underwater. Platform ROAS does not know COGS, shipping, payment fees, or the discount that bought the click. It also does not know whether the order is a first purchase that never returns.

Teams then scale the winner and cash gets worse. The metric was precise and incomplete. Most brands calculate unit economics wrong by treating platform ROAS as profit.

Two formulas that should sit next to ROAS

Keep platform ROAS as a channel efficiency signal. Add these two numbers to every spend decision.

Contribution margin per order (simplified): Order revenue - discounts - COGS - shipping - payment fees. Divide by orders for a per-order view. Contribution margin ROAS (or return on ad spend on margin): total contribution margin attributable to the period / paid ad spend. If you cannot attribute margin perfectly, use store-wide contribution margin / paid spend as a floor.

Marketing efficiency ratio (MER): total revenue / total paid marketing spend over the same window (often 7 and 28 days). MER ignores last-click credit games between Meta and Google. A brand can have a Meta ROAS of 3.5x and a MER of 1.8x if other channels and brand demand are weak - that is a business problem, not a platform UI problem.

Worked example

Assume Meta reports $50,000 spend and $150,000 attributed revenue (3.0x platform ROAS). Gross sales in Shopify for the same window are $180,000. Average discount rate 10%, COGS 35% of pre-discount price, shipping net cost 8% of net sales, payment fees 3% of net sales.

Net sales after discount: $162,000. Rough contribution after COGS, shipping, and fees: $162,000 - (0.35 x $180,000) - (0.08 x $162,000) - (0.03 x $162,000) = $162,000 - $63,000 - $12,960 - $4,860 = $81,180.

Store-level margin return on that Meta spend alone is $81,180 / $50,000 = 1.62x if you naively assign all margin to Meta - already far less flattering than 3.0x. Reality is messier because not all revenue is Meta-driven. The point is not a perfect multi-touch model. The point is that scaling on 3.0x platform ROAS while contribution is thin is how brands grow into a cash crunch.

Decision rules that beat ROAS-only

Use thresholds the finance lead will recognize, not only the media buyer.

Do not scale a campaign on platform ROAS alone if contribution margin % after discounts and shipping is below the target floor for that SKU or collection
Prefer MER and new-customer contribution over last-click ROAS when brand search is stealing credit from prospecting
Separate brand search efficiency from cold prospecting in every report - mixed ROAS hides the real acquisition problem
Judge retention channels on cohort LTV and repeat rate, not first-purchase ROAS
When platform ROAS rises but MER falls, find the channel or discount that is buying empty revenue

What platform ROAS is still good for

Platform ROAS is useful for ranking creatives inside one channel, spotting delivery anomalies, and talking to the algorithm in its own units. It is a poor sole scoreboard for whether the business should spend the next $10,000.

Even imperfect margin beats precise vanity efficiency. Pull product cost and shipping into the decision, then judge channels by contribution and cohort repeat - not first purchase alone.

Questions

FAQ

Is ROAS useless?

No. It is a channel efficiency signal. It is incomplete for deciding whether growth is profitable once COGS, shipping, discounts, and retention enter the picture.

What should I optimize instead?

Contribution margin after product cost, shipping, and discounts, plus blended MER and cohort retention - not platform ROAS in isolation.

What is a healthy MER for DTC?

It varies by margin structure and growth stage. Many mid-market DTC brands target roughly 3-5x MER as a starting conversation with finance - but the only correct target is the one that covers contribution margin and cash goals for that brand. Copying a peer's MER without their COGS is cargo cult.

How do I start if cost data is messy?

Use approximate COGS by collection and a flat shipping assumption. Imperfect margin guidance still beats scaling on platform ROAS alone. Improve cost fidelity monthly; do not wait for perfect data to stop bad spend.

Want this on your account?

Thirty minutes. Bring the number that keeps you up.

Book a demo

Stop optimizing platform ROAS alone