How *ROAS goal* bidding protects margin

What minimum ROAS mechanics do#

ROAS goal bids selectively for predicted value: auctions likely to clear your multiple get competitive bids, while thin-value auctions get shaded down or skipped. This protects contribution per pound but slows spend wherever value predictions look uncertain. Meta documents the control in its minimum ROAS guidance, and the bid strategies guide contrasts it with cost cap.

The model can only protect value it can see, so every purchase must carry accurate value and currency. Missing values read as zero value, and zero value never clears any goal.

Setting the goal from margin maths#

Work backwards from contribution margin, not revenue vanity: a 60 per cent margin with a 20 per cent growth reinvestment target needs roughly a 2.5 multiple before ad spend, adjusted for returns, fees and fulfilment. Set the goal at the floor the business must clear, then add stretch only once delivery stabilises.

Keep the workings in the account record: future editors must know whether 3.0 guards margin or memorialises one cautious quarter. The ecommerce playbook shows the margin ladder, and the hub holds the planning tools.

Value signal the goal depends on#

Feed deduplicated purchase values through browser plus server paths with CAPI, confirm standard events carry value on every purchase, and sync offline or CRM revenue with offline conversions where sales close off site. Catalogue advertisers should verify variant IDs through catalogue setup so values attach to the right products.

Audit value match rates monthly: drifting values quietly retrain the goal onto fiction. Meta event quality scoring plus the CAPI guidance confirm the plumbing.

Ramp, budgets and patience#

ROAS goals ramp slower than cost caps because selective value bidding needs deeper evidence: fund roughly 50 purchase events a week minimum, freeze edits for a full cycle, and judge on weekly ROAS plus spend together, never one alone. Thin budgets under strict goals sit in Learning Limited until funding or pools grow.

Launch from volume evidence: run Highest Volume to discover achievable return, then set the goal slightly below it and tighten in steps. Learning status tells you when the goal has settled.

When the goal fails and what to do#

Goals fail where values lie, margins shift or creative cannot win value auctions: fix signal first, recheck margin maths second, refresh concepts third. Seasonal mixes distort ROAS fast, so segment prospecting from retargeting with prospecting splits before loosening anything.

Loosen deliberately for scale pushes and tighten for margin defence, each as a dated decision with a review point. Where unit CPA matters more than value mix, cost cap may fit better. Ask our team for a goal review when ROAS and profit disagree.

Seasonality and deliberate goal flexing#

Value mixes swing with seasons: gift buyers, sale hunters and full-price loyalists carry different baskets and different returns, so a goal tuned for October will strangle December volume or waste March margin. Plan goal flexes with the trading calendar: loosen 10 to 20 per cent for peak weeks to buy volume, then retighten as demand normalises. Seasonal planning sets the calendar.

Treat every flex as a dated decision with a review point, never a panic toggle, and keep prospecting and retargeting goals separate so one seasonal mix never drags the other. Sequence the moves with scaling plans and confirm value signal stays clean through the volume spike. Attribution explains why reported ROAS lags the flex.

Frequently asked questions#