Objectives: telling the machine what to want
The objective is the single most powerful choice you make. It's the steering wheel - everything beneath it bends to serve it. Pick the wrong one and a perfect campaign optimizes toward the wrong thing.
Remember the auction's Estimated Action Rate from Day 2 - "how likely is this person to take the action?" The objective defines what "the action" is. Choose "Sales," and Meta hunts for people likely to buy. Choose "Traffic," and it hunts for people likely to click - who may never buy. Same budget, same creative, wildly different people reached. The objective is the instruction set for Meta's AI.
1The six objectives (Meta's consolidated set)
Meta consolidated to six outcome-based objectives (older guides call this the ODAX set). Each maps to a stage of the customer journey:
2Map them to the funnel
The objectives line up with the classic marketing funnel. This is how you'll explain campaign architecture to clients:
A common structure: run Sales campaigns to capture demand from people ready to buy, while Awareness/Engagement campaigns build the future pool of buyers. But note the modern shift below - this layering is less rigid than it used to be.
Telling Meta your objective is like briefing a headhunter. Say "find me people who'll click my profile" and you'll get curious browsers. Say "find me people who'll sign a contract" and you get fewer names but real buyers. The headhunter optimizes for exactly the verb you give them. Most beginners brief the wrong verb because the cheaper verb (clicks) produces nicer-looking numbers - and then wonder why the clicks never convert.
3The counter-intuitive truth: optimize for the real money event
Instinct says: "I'm new, conversions are expensive, let me start with cheap Traffic and graduate to Sales later." Almost always wrong. If you want sales, choose Sales from day one. Here's why: Meta optimizes toward whatever you pick, and the buyers and the clickers are genuinely different people. Why different? Meta scores every user's likelihood of every event separately. Habitual clickers are cheap to reach precisely because few advertisers compete for them - they click everything and buy almost nothing. Buyers cost more per click because everyone is bidding for them. Cheap clicks aren't a discount; they're a different inventory. Optimizing for clicks trains the system on the wrong humans. Provided you can generate enough conversion events to learn from (Day 5), always optimize for the deepest event that matters to the business.
Why the cheaper objective often produces the more expensive customer. Numbers below are made up to show the mechanism - your results will vary:
Traffic buys the most clicks; Sales buys the fewest clicks but the right ones. The cost per click goes up, while the cost per customer comes down.
And the modern shift: with broad targeting and Advantage+ (Day 15), many accounts now run one consolidated Sales campaign and let delivery do the funnel work - the funnel remains how you think and report, less and less how you structure campaigns.
This is literally the first screen after clicking + Create. Two distinct choices, often confused (note: as of mid-2026, picking Sales or Leads may drop you into a streamlined Advantage+ setup where these are defaulted or collapsed - look for "Switch to manual setup" to see the full tree):
The optimization event (UI label: "Performance goal") is the precise action inside the objective. Under "Sales," you can optimize for Purchase, Add to Cart, Initiate Checkout, etc. This is the real lever - it tells the algorithm the exact moment of value to chase. Choosing "Purchase" vs "Add to Cart" sends the system after completely different people.
Two recurring errors: (1) Running Traffic campaigns and judging them on sales - "we got 10,000 visits and zero purchases!" Of course - you optimized for visits, not purchases. (2) Optimizing for a shallow event (Add to Cart) when they want a deep one (Purchase) because the shallow one is cheaper and looks better in reports. You'll constantly need to pull clients back to optimizing for the event that actually equals revenue. Vanity events are the enemy of efficiency.
Walk the objective → optimization-event tree in draft mode and watch the choice rewire everything beneath it. Nothing runs, nothing is billed.
- Click + Create → choose Sales → continue into the ad set (it stays a draft).
- If Meta drops you into a streamlined Advantage+ setup, look for "Switch to manual setup" or expand the ad set settings - the same objective → location → event tree is underneath. (Advantage+ gets its own deep-dive on Day 15.)
- Find Conversion location and the performance goal / optimization event dropdown. Write down every event you could optimize for (Purchase, Add to Cart, Initiate Checkout…).
- Go back and switch the objective to Leads. Open the same dropdowns - notice the entire event list changed. That's the objective rewiring the machine in real time.
- For one business you know well, finish this sentence in writing: "The deepest real-money event is ____; if it's too rare to learn from, my fallback event is ____."
- Discard the drafts.
Today's recap - 30 seconds
- The objective defines what "the action" is in the auction - it's the master instruction to Meta's AI.
- Six objectives: Awareness, Traffic, Engagement, Leads, Sales, App Promotion, mapping to funnel stages.
- Within an objective, the optimization event (e.g. Purchase vs Add to Cart) is the precise lever.
- Optimize for the deepest real-money event you can - clickers and buyers are different people.
- Never judge a campaign on an outcome it wasn't optimized for.