One of our sites is seeing ~100 sessions a day landing on the home page with a query string like the following appended:
Each one has the same value for "p" but a different "subid" and "uid."
I am trying to figure out where these variables are coming from.
I reviewed the Access Logs and Google Analytics and found that 98% of the traffic is new and the referrer is reported as our own website even though the access log indicates its the first page request for the ip.
Our site doesn't make use of these variables. Where they are coming from is a mystery.
I have explored the possibility that the variables are related to a cookie or tracking event from an ad or analytics service. I have yet to turn up evidence that corroborates this hypothesis.
The fact that it is masking or spoofing the referral makes me wonder if it is somehow botnet related. My suspicions about this grew stronger when I noticed that the traffic is evenly distributed across pages on the site; it has a uniform bounce rate and a uniform time on page of almost exactly 1 min; it is accessing minor navigation items (i.e. privacy page) at a greater frequency than ordinary; and last, there is not a common ip or user agent.
When I googled the uid portion of the query string I found other sites having the same query string with a different p value. Here are some examples:
So, it is not unique to us. I have looked under every stone I can think of so I am reaching out to the community for help. Has anyone run across this before? Do you have any thoughts on other things I could investigate that might turn up an answer to what is causing this?
---- 5/27/17 Update ----
We started redirecting traffic that matched the pattern to a page that contained a captcha. One month later over 6,000 sessions matching this pattern have gone to that page. None have passed the captcha. This does not appear to be human activity.
If you want to check your site's analytics for traffic matching this pattern you can use the following regex match against the landing page: