What are the main attribution models?
Meta’s default attribution model is last-click within its attribution window. Meta uses a 7-day click, 1-day view window by default. If someone clicks your ad and converts within 7 days, that ad gets full credit. If they only viewed (didn’t click) and converted within 1 day, the view gets credit. This is a single-touch, last-touch model.
How Meta’s default attribution works
Here’s a practical example of how Meta assigns credit:
Under Meta’s last-click model, the retargeting ad gets full credit for the $120 purchase. The video ad and carousel ad that introduced the customer and built interest? They show zero conversions in your dashboard.
This is why people often think retargeting is their best-performing campaign and prospecting is “wasting money.” The attribution model is hiding the full picture.
Attribution models in plain English
Think of it like scoring credit for a team goal in basketball. The player who makes the final shot (last click) gets the stat. But the player who stole the ball, the one who made the pass, and the one who set the screen all contributed. If you only reward the final shooter, you’ll bench the players who create opportunities. Attribution models are different ways of splitting credit among everyone who touched the ball. Last-click attribution is like only tracking who scored. Data-driven attribution is like watching the film and figuring out which plays actually led to baskets.Common attribution model mistakes
Only using last-click and killing prospecting campaigns
Only using last-click and killing prospecting campaigns
Last-click attribution systematically under-credits top-of-funnel campaigns. If you see your prospecting campaign showing a 1.2x ROAS and your retargeting campaign showing a 8x ROAS, the retargeting number is inflated. Retargeting only works because prospecting brought those people in. Cut prospecting and watch your retargeting results collapse within 2-3 weeks.
Comparing ROAS across different attribution windows
Comparing ROAS across different attribution windows
A campaign measured on 7-day click will always show higher ROAS than the same campaign measured on 1-day click. If you’re comparing performance across campaigns or ad accounts with different attribution windows, you’re comparing apples to oranges. Standardize your windows before making any decisions.
Not understanding view-through conversions
Not understanding view-through conversions
View-through conversions count when someone sees your ad (doesn’t click) and converts within 1 day. These are controversial. Some are real: someone sees your ad, remembers your brand, and Googles you later. Others are coincidental: someone was going to buy anyway and happened to scroll past your ad. Don’t ignore view-throughs entirely, but don’t treat them the same as click-through conversions either.
Over-crediting retargeting campaigns
Over-crediting retargeting campaigns
Retargeting ads often show 5-10x ROAS while prospecting shows 1-3x. This doesn’t mean retargeting is 5x better. Retargeting audiences are people who already visited your site and showed intent. They were likely to convert regardless. Many brands find that pausing retargeting only reduces total conversions by 10-20%, not the 50%+ the attribution numbers would suggest.
How attribution models relate to other metrics
How to think about attribution correctly
1
Know what Meta is doing by default
Check your attribution settings in Ads Manager. Meta defaults to 7-day click, 1-day view, last-click. If you’ve never changed this, that’s what you’re using. Go to your campaign settings and confirm.
2
Compare attribution windows side by side
In Ads Manager, use the “Compare Attribution Settings” feature to see your results under 1-day click vs. 7-day click. If a campaign only looks good on 7-day click, those conversions are happening days after the click. Not necessarily bad, but worth understanding.
3
Don't judge prospecting by last-click alone
If your prospecting campaigns show a 1.5x ROAS on last-click, check whether your retargeting audience is growing. If prospecting is feeding new people into retargeting, it’s more valuable than the 1.5x ROAS suggests. Try pausing prospecting for a week and see what happens to retargeting volume.
4
Use incrementality tests for the real answer
The only way to truly know a campaign’s value is to run a lift test. Meta offers Conversion Lift Studies that compare a test group (sees ads) to a holdout group (doesn’t). This measures actual incremental conversions, not just attributed ones.
5
Let AdAdvisor surface the patterns
AdAdvisor analyzes your conversion data across your full account, helping you see which campaigns are actually driving results vs. which are just claiming credit from other campaigns’ work.
See the full picture behind your conversions
Attribution models are just one lens. AdAdvisor combines your conversion data, spend, creative performance, and audience insights to show you what’s actually working, not just what’s claiming credit.Try AdAdvisor Free
Connect your Meta account and see which campaigns are truly driving conversions.
ROAS Calculator
Calculate your break-even ROAS so you know when a campaign is actually profitable.
Related terms
Attribution Window
Time period for counting conversions
View-Through Conversions
Conversions from ad views, not clicks
ROAS
Return on ad spend