Have you logged into Google Analytics recently and noticed a certain keyword slowly rise to the #1 spot in your Keyword Report?
Back in October 2011, Google announced it would encrypt (SSL) searches performed by those who are logged into their Google Accounts (Gmail, YouTube, Reader, Analytics etc) which means sites will no longer receive the referral search term from these searchers. This makes it difficult for SEOs like us to optimise websites for particular keywords or search terms but it also helps us identify how many people could be seeing personalised search results. In Google Analytics you will see all encrypted searches grouped under (not provided).
This change makes it difficult to know exactly which keywords are bring in quality traffic which then go on to convert into leads, and which keywords are attracting bad quality traffic. However, one side of the coin is that Google have said referral data (i.e the search term) IS passed onto Google Analytics if a searcher clicks on a PPC ad. In simple terms, ad clicks = £££ for Google. 95% of Google’s revenue comes from those ad clicks. That’s for another post maybe.
What is important for us as SEOs and Marketers, is that this change is hiding the keyword which actually generates leads and conversions.
The above image shows how ‘(not provided)’ is the number one keyword generating 434 visits and converting over 20% of traffic. Blimey! Yet I have no idea what those keywords are in order to produce more content and drive more traffic to the site from those search terms which have been hidden. It’s difficult to tell the client that their requirement to be on page 1 or top spot in Google for a particular keyword is actually working for them.
Since October 2011 this is the rate at which traffic is coming onto the Pauley Creative website from the keyword (not provided). All those hidden search terms now account for 3.5% of all visits to the site and it’s still growing.
Early indications show that encrypted search has now been rolled out in the UK (Google.co.uk) and other international sites (co.in, .es .fr) so if you have a website which spans across many countries and rely on SEO for traffic then you could be in for a bit of shock.
I’d love to know what % of your website traffic comes from the keyword (not provided). Some may be small depending on how tech-savvy your audience is and whether or not they have a Gmail email address in which case they may be logged into Google for most of the day. How does this change effect your site and by how much has it grown in the last 6 months or so?
Am I right in thinking the keyword (not provided) could infact be lots of different search terms including keywords that already feature on your list, its just that these searchers were logged in.
So on that basis it is likely to made up of the information analytics does give you. For example if construction company brings in 10% of your current traffic in organic search, you could guss that it brings in 10% of your (not provided) traffic?
Hi Emma
Correct. The search terms are all grouped together and batched under (not provided) or [secure search] if you are using another Analytics software, and come from those who are logged into their Google Accounts. Visits from (not provided) is increasing everyday as more and more people adopt Gmail and sign up to Google+ etc.
I don’t think it would be right to assume if 10% of your current traffic in organic search that it also brings in 10% of your (not provided) traffic.
I could have 200 visits coming from the keyword ‘construction company’ over a 3 months period.
I then do some SEO to bring in more visits from that keyword by increasing the rankings and creating more content to dominate page 1 of Google. I am also SEO’ing for the keywords construction co’s, construction company in Brighton and construction company south England.
After a few months, I also see that (not provided) visits have increased from 300 to 1000.
Now, which of the 3 keywords has seen an increase? What of construction company went from 200 to 900 and construction company made up the other 100? But then what if neither of them increased and actually it was an increase in the keyword ‘building company in brighton’?
You can see why it’s a problem.
Thanks for the response. Oh yes I can see that it is a problem in terms of analysis because you can only assume or guess rather than know what keywords are being used.
Hi priesh, this is really frustrating with google. How can we generate business now? Now even firefox has started with https, this will be like an end to business. Is there any solutions or any other options to it ?
Thanks for explaining it all. It’s such a frustrating change though, as I have set up a website in the last 2-3 months and 69% of my organic traffic is (not provided) so how on earth am I meant to start building the website up when I can’t see the most popular searches to get to my website? Very frustrating indeed.
Thanks for the comment Jack.
It is frustrating but what I would also do is look at the data within your webmaster tools account. Try and see if you can spot any search terms from there which don’t show up in GA which also include an increase in impressions.
This post might help: http://webanalysis.blogspot.co.uk/2012/03/finding-not-provided-keywords-in-google.html#axzz1qUYWR2jU
Have seen massive increases in (not provided) traffic in March. Getting well over 4% for all site visits on most of our sites now.
4%! Wow…..you will notice the increase starts around 5th of March as this this is when Google rolled out the secure search across Google.co.uk and .de and .es and .fr and others.
What is worrying is when the conversion rate (i.e percentage of downloads, enquiries, registrations etc) for (not provided) are the highest. Sigh.
Thanks for the post Pritesh
We have 2 Dutch sites; our product site shows 7% “not provided”, our contruction site 11,8%. Our French product site shows 11,4% and, funny enough, our German site 0%. I can’t explain that one.
As Anil Batra wrote in his blog, registering to Google Webmaster Tools seems to be the solution.
Strange.
I’ve just finished doing the month end report for a client. In February they had received 31 visits from (not provided). In March it jumped to 240….674% increase!!!
Hi Pritesh,
Thanks for the explanation. However, I’m almost certain there’s been a change of some sorts within the last month. Much like you, I’ve just finished doing a report for a client which has seen the not provided traffic go from 73 (1-Jan-2012 – 29-Feb-2012) to 576 (01-Mar-2012 – 19-Apr-2012)… 689% increase!
I work for multiple clients and comparing February to March has seen this increase across the board. It’s always somewhere between 400-700% of an increase in ‘Not Provided’ traffic.
Charles – Wow! (just picking myself up off the floor!)
What % of that is conversions? The rise could be the increase in people signing up for Google+ or other Google products which forces you to sign in to your account. Maybe.
For one of our clients they saw an increase of 400% in ‘not provided’ traffic – from 57 to 285. The conversion rate of those 285 was 3.51%. For this same client ‘not provided’ is the second highest organic term, and is accounting for 7.2% of their entire organic traffic for March. I’ve just sat and looked through 5-6 of our clients and my rough guess was slightly off, the lowest was a 250% increase and the highest was 650% – this is still a huge jump between the period of two months.
I can’t see that kind of increase simply coming from more people signing up for Google products – it seems too large!
Sorry to double post – I’ve just filtered my Analytics down by ‘not provided’ on a month by month. Regardless of which client I pick as soon as you hit March the graph jumps up. Something’s definitely changed around the March period. What that is exactly I’m still unsure.
Charles, If i remember correctly, i read a blog post from Econsultancy that (not provided) rolled out across all European Google domains around March time (i.e. Google.co.uk/.fr/.de/.es etc)
Okay triple posting here – sorry! (I wish there was an edit button)
Looks like I may have gotten to the bottom of it – http://insidesearch.blogspot.co.uk/2012/03/bringing-more-secure-search-around.html
As of 5th March Google introduced the SSL Encryption to Google.co.uk searches (being a .co.uk domain I’m assuming you guys are also based in the UK?). Looking at my own data it fits perfectly, from the 5th March onwards the amount of ‘not provided’ search traffic increases. What’s worrying about this is that on my search to find the reason behind this, I stumbled across news that over the coming months Firefox plans to make ALL of their default Google search SSL encrypted. Considering Firefox hold around roughly 25% of the browser market, this could be huge.