Google Analytics Question

Discussion in 'Internet Marketing' started by Fergal, May 19, 2011.

  1. Fergal

    Fergal
    uix_expand uix_collapse
    Premium Member
    Premium Member

    Joined:
    Nov 18, 2007
    Messages:
    10,575
    Likes Received:
    1,165
    I did one search for the following in Google.com on 14th May "windsor site:http://www.businessadviceforum.com". That search is now showing up in Google Analytics for this site, which is great.

    I did not repeat that search since the single search I made on the 14th, also that search term was not showing up in Analytics until I made it on the 14th. However, the Google Analytics report for http://www.businessadviceforum.com for the 17th May - Traffic Sources > Keywords it says that we had 11 visits from that search term on that date - screenshot.

    I really don't believe that there were 11 other searches for that obscure term on the 17th. So my question is where are those 11 visits coming from?

    Thanks!
     
  2. mysandyda

    mysandyda
    uix_expand uix_collapse
    Member

    Joined:
    Jun 2, 2010
    Messages:
    225
    Likes Received:
    7
    Google doesnt lies, I am sure in it so I think it can be the truth.
     
  3. John Peterr

    John Peterr
    uix_expand uix_collapse
    New Member

    Joined:
    Jan 22, 2010
    Messages:
    20
    Likes Received:
    0
    Hi,
    I Don't think it is a surprising thing, may be other people have also types the same term as you had typed.
     
  4. Charles-John

    Charles-John
    uix_expand uix_collapse
    New Member

    Joined:
    Feb 23, 2011
    Messages:
    24
    Likes Received:
    4
    It is unlikely that people use such a query. Judging by the time on the site and the number of pages visited it must have been you :D. By the way does BAF has only 447 visits from SEs a day?? Looks like the report is totally wrong.
     
    • Like Like x 1
    #4 Charles-John, May 24, 2011
    Last edited: May 24, 2011
  5. JPM

    JPM
    uix_expand uix_collapse
    Active Member

    Joined:
    Jan 10, 2011
    Messages:
    928
    Likes Received:
    179
    The point Fergal is trying to make is that it hadn't been posted at all before he searched it, so it is exceptionally coincidental that it happened a further 10 times on the same day, especially as it is so obscure.

    Fergal, I notice that, according to your screenshot - that none of the visitors from the search term were new, does this mean that analytics mistakenly recorded you as clicking it 11 times when in fact you just visited?

    Does that make sense, not sure if I articulated myself properly.
     
  6. Charles-John

    Charles-John
    uix_expand uix_collapse
    New Member

    Joined:
    Feb 23, 2011
    Messages:
    24
    Likes Received:
    4
    Actually he made the search on 14th and the additional searches were recorded on 17th, if they were on the same day may be it could be better explained.

    Any idea which were the landing pages for that keyword??
     
  7. timpane

    timpane
    uix_expand uix_collapse
    Member

    Joined:
    Apr 26, 2011
    Messages:
    51
    Likes Received:
    0
    Judging by the time on the location and the number of sheets travelled to it should have been you
     
  8. Fergal

    Fergal
    uix_expand uix_collapse
    Premium Member
    Premium Member

    Joined:
    Nov 18, 2007
    Messages:
    10,575
    Likes Received:
    1,165
    That's correct Charles-John, the searches were showing up a couple of days after the day I made the search. The number one result and the result I was looking for when I completed that search was Tying a Tie - Windsor Knot.

    You could be right there, the stats for this site as shown by AwStats are much higher than those shown by Google Analytics. I don't believe Analytics is reporting correctly but I find it useful for uncovering info such as sites that are referring visitors to us and what keywords people are using to find us.
     
  9. PaulPinnacle

    PaulPinnacle
    uix_expand uix_collapse
    Member

    Joined:
    Aug 24, 2010
    Messages:
    275
    Likes Received:
    59
    It doesn't quite explain it, given the discrepancy with dates etc., but it's a good starting point to start digging from in general. When you drill down into that specific keyword, the fact that the bounce rate, pages/visit, etc aren't uniform (e.g. either 0% or 100%) also seems to suggest otherwise at an initial glance. The abnormally high P/V and avg time on site (along with the relatively low bounce rate) also make the results stand out a little.

    It certainly is an odd one Fergal. I've nothing I can think of off the top of my head to explain it, but let me do a little digging tomorrow (working on an updated analytics lecture anyway so directly related) and see if anything jumps out. The figure on the P/V would make me throw my hat at some sort of bot/scraping activity, but I've no explanation for why that specific term would suddenly emerge as the entrance path. Scratching my head at the moment.
     
  10. PaulPinnacle

    PaulPinnacle
    uix_expand uix_collapse
    Member

    Joined:
    Aug 24, 2010
    Messages:
    275
    Likes Received:
    59
    This is quite common, but in most cases relatively straight forward to resolve. How large is the discrepency between the log files and the GA stats? Some difference is always going to occur and to be fully expected, once the gap between the two becomes large (e.g. double digit percentage) it suggests it might be down to more than just the difference in systems.

    The most common issue here is for the code to be missing from pages on the site. The easiest way to pick up this problem is spotting a lot of self referrals in the stats (e.g. BAF showing as a referral site to BAF), which can then lead you to the relevant pages too (or to look for patterns within the pages so you can identify the root cause, such as a missing include). Probably the next most common issue is for buggy JS code to cause the 'correct' GA code not to fire, but it's getting a little trickier to spot at this level. Beyond the obvious, it can get into really bespoke situations for individual sites but you can see a list of the most likely causes here directly from google (also worth looking through the tools available for diagnosing problems. Personally I like using Charles as it gives a huge amount of information, but for usability you might be better starting off with something like the Google Analytics Tracking Code Debugger).
     
    • Like Like x 1
  11. gomix

    gomix
    uix_expand uix_collapse
    New Member

    Joined:
    May 26, 2011
    Messages:
    2
    Likes Received:
    0
    Google statistical always inaccurate, I believe of this point
     
  12. Fergal

    Fergal
    uix_expand uix_collapse
    Premium Member
    Premium Member

    Joined:
    Nov 18, 2007
    Messages:
    10,575
    Likes Received:
    1,165
    Thanks for all of that Paul, very helpful as usual. AwStats shows us getting well over twice the amount of visits shown by Google Analytics, I will go through those links on Google and try to figure out what the issue is.
     

Share This Page