Not provided in what?
In what particular tool/log/field?
I'm surprised there isn't much discussion on this one. This is above and beyond what folks have been discussing to date. There have been significant increases in "Not Provided" stats starting on Oct 30, 2011. I've had 10+ people confirm on Twitter and there are plenty of screenshots floating around depicting the increase. It is rather significant. This is not the measly increase we first saw when it was announced.
On one site I monitor, we were averaging about 500 per day. That jumped to 1,700 on Oct 31, 2011 and it then jumped to 4,200 on Nov 01, 2011 and appears it will be in the same range if not more once Nov 02, 2011 closes. Similar spikes are being seen by quite a few folks.
I am seeing the same spike in GA too since Oct 30. It now tops the keyword list whereas before it would be in 4th or 5th spot. Looking at the data it is mostly from US.
My traffic has also decreased since Oct 30 a bit so not sure if that is Seasonal related, weather (power outage in New England) or related to G tinkering.
I am seeing this as well it is now the number one result on my site. Did I miss some kind of announcement about this?
|On one site I monitor, we were averaging about 500 per day. That jumped to 1,700 on Oct 31, 2011 and it then |
You mean "(not provided)" traffic? or in general.
If the first then G just started not to provide referral data for SSL (logged in users) traffic.
|You mean "(not provided)" traffic? |
Yes, on a more significant scale than what was originally reported.
Keyword “Not Provided” By Google Spikes, Now 7-14% In Cases
^ I was expecting to see it reported here at WebmasterWorld first. Most appear to think I'm talking about the first SSL announcement. I'm not that far off the grid. Sheesh...
18% yesterday for us... Specifically the obfuscated, not provided, Google referrals.
yes, very annoying not set and not provided make up 20%, time to remove that code?
Not sure if it's connected, but 2 site's i monitor received dramatic spikes on 4 thru 5 Nov from [t.co...]
Everything's back to normal now - but it throws GA performance measurements out, so quite frustrating.
I'm not keen on flying blind.
I hit 19% yesterday. If the trend continues, I'm going to look at phasing in alternative analytics so I can compare an another solution with Google Analytics while data from the later is still largely available.
Given the lax response on this, I kind of feel like I'm over overreacting -- but in a case like this, I'd rather be safe than sorry.
Any other opinions?
Self correction: SSL queries of logged in users from G aren't visible to other analytics services either.
For some clients I've been seeing the number creep north of 20% (this for the last 7 days, i.e., 11/7 to 11/13). Not happy at all, but I think we're fairly helpless on this one...at least until Google starts getting more heat from Federal government inquiries into its monopolistic practices, at which point this issue should start getting more sunlight (i.e., "Protecting privacy? Really? Like you DON'T do for visitors clicking through paid ads?")
I'm seeing 14% "not provided" this week, and the "not provided" is converting nicely at above averages.
I'm a bit surprised over here that there's little discussion and apparently not much concern.
Also 9% "not provided" on a much larger site. 9% in this case is over 40K visitors. Ouch.
So this is possibly a useful hack to get some data back.
The usual caveats apply.
I will be testing it over the long holiday weekend.
Thanks chewy! Looking forward to hearing what you think.
if workarounds are the best we can come up with, that is a start.
my preference is, and has always been, that we need to pressure the powers that be to give webmasters reasonable access to information over obscuring and obfuscating a small but increasing subset.
professionals with the appropriate knowledge have the ability to 'read between the lines' and see the data that is missing.
but it is a great thing to be able to say to an executive "we know x is doing y".
it is not so great to have to say "it is a reasonable and educated guess that x may be causing y to happen".
We've noticed an increase in (not provided) on most of the websites we track. Even more strange though is that one of our client sites do not have even one instance of (not provided) though we do show (not set). I find it hard to believe that not one person coming to the site is logged into Google. Any ideas?