-
Experiment on Pixel LLA vs. Uploaded List LLA
Posted by on April 6, 2021 at 11:17 pmHas anyone experimented with a direct test between Lookalike Audiences that come from the same store, just different “source” – in this case the fb pixel vs. an uploaded list?
Out of curiosity, I am currently testing this with one our accounts over the next few days and wanted to know if others have tried the same or have any additional data on this.
Scenario:
One campaign & ad set is using the Pixel data for a 1% LLA of the Past Purchasers in the last 30 days.Another campaign & ad set is using a 1% LLA of the uploaded customer list of Past Purchasers in the last 30 days from the same store.
The estimated lla audience sizes are both over 2M+ so I’m not worried about the overlap at this stage.
To my surprise, the LLA from the uploaded customer list has a very poor performance in comparison to the pixel LLA even though they are using the same website & data source of the same past purchasers.
I was expecting them to be similar – what are your thoughts?
-
8 Replies
-
I’ve been wondering this exact question. I hope to hear some other experience in here.
-
List size will influence that a bunch. If the match rate is low and the count is low, facebook will have a hard time generatjng a quality lla. Believe you need 50 conversions to get that lla to work. The pixel accelerates you past the learning phase. From my experience tho.
-
The only thing that could explain this is that pixel has more data points for FB to match it with, whereas you customer data set would be from email, name, contact no only?
-
Isn’t all fb LLA essentially just certain separations of friends…it’s a weak LLA at best.
-
Even direct pixel for page visitor LLA generally performs better, ATC & purchases are absolute bangers for me than an LLA on a custom list with a 1000 customers.
-
Pay attention not only for base audience sizes but how how many matches system managed to get front hose audiences. On-site pixel can transfer additional data (eg, email and phone) which help it to find matches. Uploaded list loses a lot in quality as you provide less data for each user
-
These are exactly the same users, so this comes down to how many users Facebook can match to your seed list.
If you have an equal number of users your pixel visitors are likely to generate more matches.
However, if you have buyers from more than 180-days ago your uploaded contacts may be a more robust list… i.e. users on your pixel list will fall off after that.
Just ensure that you keep your uploaded list up to date… best to do with a direct integration.
-
Depending on your setup, the audiences may be overlapping and Facebook is giving preference to the Pixel becuase its tracked by them so they have more data there
Log in to reply.