Since Jan 20 I've observed a steep decent on the Passerby count In 3 different organizations
Is there any change in how it is being calculated?
Is there any publicly available documentation on the algorithms used for the analytics?
Found the Answer in this post Meraki and iOS 14 MAC Address Randomization - Cisco Meraki
on this Thread Documentation Digest: January 21st to January 25th - The Meraki Community
It has to do with the drop of randomized MAC addresses from the passerby count.
I'm not convinced by that article (it from IOs latest update in September and the changes in passerby count is on 20 Jan), maybe there is another explanation.
Yeah the MAC Address randomization has been in place for a while.. probably something else but we haven't experienced it on our end, might just be a configuration issue/change on your network?
No, two different Networks on two different organizations, no changes on the configuration log.
Org A
Org B
It seems that visitors data is no affected, but there's a Clear drop in Passerby sice 20 Jan.
The fact that its two different organizations definitely has to mean that there's been some systematic change. I'm surprised there aren't more recent posts about this subject?
@Juanchomann our all seemed to go right down on the 21st January, is that the same for you?
The article you referred to was posted on the 22nd so I think the ignoring of random MAC addresses was enabled on the 21st:
@cmr that article is from September, and the update on 22st is "Updated Screenshots" . Looking back until July there is noticeable growing trend in passerby until the change in Jan 20. Maybe you're right. Hope that someone can verify that.
Haven't done any change or update on my side since in January, and still I see a massive drop..
@Nacouzi that is the time that Meraki stopped counting randomised MAC addresses in the passers by statistics, up until then they were counted so that is why there is a sudden drop. Randomised MAC addresses all follow a prescribed format so it is easy to discount them.
ok, so it explains it.
Although they are devices with randomized addresses, they are still real for a time window and can be counted. Is there the possibility to change this default as they provide over in Cisco DNA?
It doesn't explain it alone, especially that there is a sudden drop, as if on January 21st, all Apple users updated to iOS 14 and enabled MAC randomization, which is practically improbable.
same for me. This being said, I think it is much more accurate right now. I have a retail store, I think it used to count the cars' WIFI, now it counts only the pedestrians WIFI, which is much better in my use case.