Hi , i'm writing this as it seems i'm not the only one experiencing it
Considering a website with an average 1k-1,5k unique visitors a day ( not bots , no spam )
When i make an audience for retargeting with FB i get around 8K people only
which is for sure not possible, as our main traffic source is instagram
I think our pixel is set up correctly (?) as i had a analytics expert doing it, and Lookalike audiences works very very good
but there's definetely something wrong with the retargeting
as it's "new user" percentage on GA is 100% ( while on mailing list campaigns, seen that users might open the mail from different devices, it's 25% )
I've found other people having this problem on FB forums like Digital Marketer Engage forum, but none could give an explanation
Just a question, even if you can't help, what is your retargeting audience size compared to your traffic ? If i have 1k new users a day for 45 days ( most of them coming from instagram ) how can it be possible to have an audience of only 9000
and again, please consider that lookalike audience works VERY well when i set up campaign, so there must be something working fine
Your help is super appreciated !
I don't think the performance of your lookalike audience has anything to do with your problem, they are not practically linked.
In any case, can you outline the following:
1. Do you have the FB pixel core JS embedded in a persistent header/footer of your page?
2. If you go to the FB pixel data itself, what traffic count per day is it measuring? If your GA is saying 1000 people a day but the pixel is measuring far less events, then the problem becomes much clearer.
3. When you say retargeting audience, what specifically does this mean? What are your custom audience parameters? What is the audience time window?
Hi Zeno ! I thought that lookalike audience was created taking people visiting my store as a "base" for it
1- i have my pixel installed in
2- my pixel fires everytime there's an event like add to cart, view content, search, pageview. So in a day where i had 1000 people, if i check the pixel it just says "8000" which is i think a sum of all events that made it fire. I'm probably not on the right route here , please advise me if i'm doing wrong.
3- when i say "retargeting audience" i mean i want to show ads to people who has been on my website. I usually set it to anyone who visited collection page so i'm sure i avoid the bouncers or the not interested ones.
Thanks mate
Yes, the audience you select is used as a base for it - what I mean is, whether or not the ads perform well or not doesn't really reflect whether your pixel is working perfectly. You could be collecting 30% or all visits, 50%, 97%. You could be catching 100% of AddToCarts, 87% of checkout init and 47% of purchases (just imagine there is a tech issue). In any case, the audience produced by FB's modeling systems could still have plenty of users that convert and the ads might work well for you -- but this doesn't tell you anything about the coherency of your pixel setup. Hell, it could be placed on the wrong site entirely and by luck the lookalike audience may work! (remember they are quite big and broad).
Anyway, moving on...
2. Yes, correct. So things like page views and view content will add up and you'll see the aggregate result.
3. If you set it to people only visiting a deep page / funnel node, you would expect a small % of your overall visitors to fit into that, no? E.g. for every 1000 visitors, only 400 might get to that (your data may vary of course). Also, what is your attribution window - 180 days? Usually the pixel is pretty accurate, I would argue it will be more accurate than GA as it will only track logged in users as uniques and not dupe them. Your GA may very well do so, so your 45k unique users may actually only be 35k, and then from that you'd be looking at about 30% tracked visiting the deeper page, which is not outside the bounds of reason.
But in any case, its hard to gauge without seeing things under a microscope.
Thank you Zeno. I already contacted an IT for this before writing you. I thought it was ok ... i don't know, as i told you i found out other people have the same issue. Do you suggest Trackify ? ? It's a
I just set it up manually but then I'm pretty code savvy - stuff like Trackify should do the trick for most and they are ideal in that they will stay up to date (you'd hope) with any