What is Delayed Attribution?
Ever saw a paused ad above target ROAS - wondering if you made a mistake? Scratching your head wondering why?
Chances are you’ve encountered the media buying phenomenon – delayed attribution. It happens 24/7 and confuses every beginner Facebook advertiser.
Delayed attribution is when conversions are attributed AFTER the 1-day click/view window.
For example, let’s say my target ROAS on an ad was 2:
On day 1, I launch my ad. On day 3, I check my results. I see the ROAS is 1.5. So I pause. On day 7, I check my results. I see the same paused ad is at 2.5 ROAS. Why could this be? Well, that’s delayed attribution kicking in.
There are two reasons for this:
1) Delayed sales cycles. You have customers who take longer to make a buying decision than usual. Thus they may convert after you check your stats. Facebook attributes conversions on the day of the view/click, not the sale. So if I click an ad on Monday, and buy on Thursday, my sale shows up on Monday’s data – not Thursdays. This is especially prominent during the Black Friday/Cyber Monday weekend. Conversion rates skyrocket during this period as customers interact with ads early but wait to buy products on offer.
2) Delayed reporting. Facebook takes around 15 minutes to update conversions after the spend. So a customer could click on an ad and buy at 1 pm. You check the stats at 1.05 pm and see the ad is at 1.5 ROAS and pause. Then at 1.15 pm, the paused ads are now at 2.5 ROAS – and now above target.
You’ll want to calculate your attribution lift for your collection, so you can view the expected lift throughout the 28 days and not pause early.
You can see in the example below how the total revenue on this account (on the default 28-day click, 1-day view window) is €539K and we can deduce that:
a) €421.6K revenue within a 28-day click window, €177.1K within a 1-day view window. €421.6K + €117.1K = €538.8K in total revenue
b) €73K has come through between day 2 and day 7
c) €134K has come through between day 2 and day 28
d) Around 25% of the attributed revenue has come through AFTER day 1, so not all the buying decisions here are impulsive
To calculate your 1, 7 and 28-day attribution lift: Create a separate column in your reporting columns with these columns:
Purchase Conversion Value
Then underneath edit the attribution window and check these boxes:
View – 1-day
Click – 1 day
Click – 7-day
Click – 28-day
And include the default attribution window.
You should have your columns set up like this to view the ROAS lift (as shown in my screenshot below)
Default windows (28-day click + 1-day view)
Sum up the 1-day click and 1-day view ROAS numbers, and repeat for the 7-day click 1-day view AND 28 day click and 1-day view. (you can do this easily with google sheets and super metrics), and you’ll be able to see how your ROAS increases over time.
Let’s say you measure your Facebook ads performance from a 28-day click 1-day view attribution window. And your lifetime ROAS for a collection (excluding any sale/promo campaign) is 1 ROAS on a 1-day, 1.5 ROAS on a 7-day, and 2 ROAS on a 28-day.
We can assume it’s likely that ROAS will jump by 33% from day 7 to day 28. If your target is 1.7 on a 28 day click 1-day view attribution – and you’re at a 1.6 ROAS on day 7, you can confidently leave that ad running (even though it’s below target), as it’s likely it will be at 2.13 ROAS after a full 28 days. It’s not 100% guaranteed, especially if FB often takes credit for conversions from other channels in your account, but it’s very likely based on past data.
This is often one of the most misunderstood topics in Facebook ads. Use it to develop a better understanding of your results and how attribution works, and adjust your ROAS targets to get more scale out of your account.
Want to learn more about delayed attribution and how it applies to your brand?
Why not hop on a call with us? We can explain it to you in-depth, and if you want us to scale your account, we can walk through how we do that and what it’s like to enroll as a client.
Interested? Feel free to book a call with us here