Note: all these numbers are from a real account we manage.
When agencies tell you how good they are because they get amazing ROAS (Return On Ad Spend) numbers, there are things you need to watch out for. Of course everyone likes to show off great ROAS numbers, but they can definitely be misleading.
Granted, this situation you’re about to see is fairly extreme, but it’s a good scenario for this case.
Recently I looked at a campaign I was working on and saw this ROAS:
WOW! Cool, right?
While this number is completely real and accurate, it’s definitely not the whole story. And while no one would probably ever pass this off as a real, consistent result that they can get you, it’s worth asking questions. More than likely these reported ROAS numbers come with a story.
Here’s the rest of this story, and things to think about.
When we open this up a bit, we see these 8 purchases were based on $0.01 spent.
How is that possible? If I could get that return on every penny spent, that would be awesome!!
Let’s look deeper.
Now I see the purchase conversion value is $203.98.
$200 is nothing to sneeze at, of course. But there’s still more to it.
Then we see under ‘Delivery’ that this campaign (or AdGroup) is actually off.
Now it’s making more sense. This definitely isn’t a sustainable ROAS, especially when the campaign is off.
What happened here is pretty simple, but it has to do with attribution and conversion windows.
The typical conversion window for Facebook is up to 3 days, minimum. Meaning conversions continue to be reported up to 3 days after they actually may have happened.
So, a purchase from an ad that happened on Monday, might not get reported to Ads Manager as a conversion until Thursday. While some of the reported data is close to real time, some of it isn’t.
(FYI, the Google Ads reporting window can be even longer, I’ve seen up to a 15 day delay.)
This particular campaign was on for a while, then we turned it off and caught it at the exact time it spent 1 cent. And because it had a spend on it and conversions reported from the previous few days, it was able to calculate a ROAS, and what a ROAS it is!
It’s important to know that if someone is saying they got some amazing ROAS, it might be a unique situation on a single campaign. All of the other campaigns in this account during this same time period had more regular ROAS numbers and this super high ROAS didn’t actually bring the total blended ROAS of all the accounts up that much.
In order to know what’s realistic, make sure to ask them what the ROAS was overall for all campaigns for a longer time period, say 30 days minimum. 30 days or longer will take care of these crazy anomalies so they don’t skew results too badly.
