Diese Seite gibt es auch auf Deutsch.Zur deutschen Version
Skip to content

Fundamentals · Guide

Choosing the Right Optimisation Objectives for Meta Ads

Campaign objective, optimisation event and bid strategy: small settings with a big impact. How to optimise for profitable revenue instead of cheap clicks.

By Denys Lichtenstein Prefer us on Google

Cover image: Choosing the Right Optimisation Objectives for Meta Ads

Two accounts with identical budget and creative can perform completely differently – purely because of the objective settings. The optimisation objective tells the algorithm what to learn on. Choose it wrong and you get exactly the wrong thing: lots of cheap clicks instead of profitable purchases.

Campaign objective: almost always “Sales”

Meta offers objectives like Awareness, Traffic, Engagement, Leads and Sales. For e-commerce, Sales is generally the right choice – even for cold audiences. Traffic campaigns deliver cheap clicks from people who like to click but rarely buy. That feels good in the reporting and still burns budget.

Optimisation event: learn on the purchase

Within the Sales campaign, you choose the conversion event to optimise for:

  • Purchase – the standard. The algorithm looks for people who buy.
  • AddToCart / InitiateCheckout – only as a bridge when you lack the volume for clean purchase optimisation.
  • ViewContent – almost never useful for performance, because it’s too far up the funnel.

For this to work, Meta needs enough signals: as a guideline ~50 purchases per ad set per week (the “learning phase”). Anyone below that should consolidate campaigns and pool the budget instead of spreading it across many small ad sets.

Bid strategy: control vs. volume

  • Highest volume – Meta spends the budget and gets as many conversions as possible. Good for learning and scaling, without a fixed efficiency limit.
  • Cost cap – you set a target CPA. More control over efficiency, but delivery can collapse if the cap is too tight.
  • ROAS goal – for accounts with stable, high conversion volume. Needs reliable signals, otherwise it optimises on distorted numbers.

When in doubt, start with highest volume, collect data, and only switch to cost cap or a ROAS goal once the signal base is stable.

Don’t forget attribution

The attribution window (e.g. 7-day click / 1-day view) determines which conversions Meta credits itself with – and what it optimises for. Important: this is the platform’s view, not the whole truth. That’s why you additionally measure via MER and your real numbers. More on this in our article MER instead of ROAS.

The most common mistakes

  • Optimising for traffic instead of sales, because the clicks look cheap.
  • Too many small ad sets that never get out of the learning phase.
  • A cost cap set too early and too tight – delivery dies.
  • Only looking at platform ROAS and ignoring the contribution margin.

Unsure whether your campaigns are optimising for the right objective? In the free account check we take a close look at exactly that.

FAQ

Frequently asked questions

Which event should I optimise for – purchase or add to cart?

Whenever possible, the purchase. The algorithm optimises best when it learns on the event that actually brings you revenue. Only if you consistently have too few purchases per week (roughly 50 per ad set as a guideline) is an event further up the funnel (e.g. AddToCart) a sensible bridge.

What does the 50-conversions rule mean?

Meta needs around 50 optimisation events per ad set per week to leave the learning phase and deliver stably. Anyone who doesn't reach this volume should consolidate campaigns, pool the budget or choose a more frequent event.

Free account check

We apply exactly this know-how to your setup – in 30 minutes we analyse your account and show you 3 concrete quick wins.