8 Google Analytics Sins: Are You Guilty?

analytics failIf you use a “set it and forget it” implementation of Google’s free Analytics tool, your data may be “accurate” but not as useful to you as it could be.

I wrote a tutorial called “8 Stupid Things Webmasters Do To Mess Up Their Analytics” and submitted it to Marketing Pilgrim’s 3rd Annual SEM Scholarship Contest. It explains in detail how you can optimize your Analytics Settings to collect relevant data that will instantly improve your conversion rates and help your web analysts make more meaningful conclusions and decisions.

So you’ll have to check out the entire article over at Marketing Pilgrim to find out the 8 reasons, as per the contest rules, I cannot post the article here.

Being a contest, I really appreciate your help! You can’t “vote” for me in the traditional sense, the contest works like this:

Between May 28th and 30th we’ll publish the qualifying entries and give each article 4 weeks (from publication) to attract as many readers as possible. Entrants are encouraged to provide articles that are optimized for the search engines and also utilize the various search and social media channels to promote their work.

On June 30th, we’ll select five entries that have achieved the highest overall traffic. In an effort to encourage quality traffic–not just quantity–each entry can earn two additional bonuses.

Bonus 1: If an entry achieves an average “Bounce Rate” below 75% it will be awarded a bonus equal to 10% of its total “unique views.”
Bonus 2: If an entry achieves an average “Time on Page” greater than 2:00 minutes, it will be awarded a bonus equal to 10% of its total “unique views.”

On July 7th, we’ll announce the grand prize winner. The winner will be the entry that receives the most votes from our panel of expert judges.

I need help to get to the top 5 which means I need a lot of traffic and a lot of engagement (few bounces, long reading time). Given that there are so many fantastic entries, competition is stiff. But I know I can count on our amazing Get Elastic readership to help, here’s how:

  • Follow the link to read the full entry

  • Please read the entire post, even if you think it’s boring
  • If it’s really too boring for you, just take a long, hard look at the pretty pictures – for at least 2.5 minutes. If you stare long enough, they start morphing into Transformers robots – no joke, try it!!
  • Share the love – link to it from your blog, give it a Sphinn, Stumble it (thumb it up or review it), bookmark it in Del.icio.us, Twitter it, post it to Facebook, email it to a friend, share through Google Reader or however you share content!

Muchas gracias!

Related Articles

11 Responses to “8 Google Analytics Sins: Are You Guilty?”

  1. You’ve got my vote…reading it right now!

  2. Thanks John, and thanks for the comment on the MP article too :)

  3. It was a great article Linda. I read it, Sphunn it, and Stumbled it. Good luck!

  4. Franz says:

    Excellent writeup, hope you win! I lingered and didn’t bounce :)

  5. alex says:

    vertical forward integration – increase value for customer i mean – and later sell nice packages- could be good for google – theyve got such huge base of customers. i ddnt know about anaytics.. they could go for all user experience management platforms, and after maybe on demand CRMs (here maybe they would have to improve their reputation for privacy of data)..

    dont you think that thier user interface assuming popularity of google is one of the wose in the world – lack of interface to access all sevices added for the individual user

  6. Phil says:

    Linda, I would add…

    * Forgetting to tag links in email campaigns
    * Forgetting to separate navigational searches from search engine keyword, by using a custom filter for Direct Search or on page addIgnoredOrganic
    * Not changing keywords and url`s to lowercase (GA groups “KEYWORD” and “keyword” differently)
    * Allowing log spam by not adding an including only hostname: (www\.|)mydomain\.com
    * When using ppc – not recording actual search keyword by using a custom filter e.g $A3 in (\?|&)(q|p)=([^&]*) and
    * Forgetting to manually tag paid search landing pages in yahoo and msn.
    * Not segmenting visitors by type (e.g return visitors behave differently to new visitors, or Hunters behave differently to Browsers http://tinyurl.com/4x2k73 )

    When installing:
    * mixing urchin & ga.js
    * Forgetting to tracking mailto: clicks
    * Forgetting to tag links between 2 domains or links to/from an externally hosted shopping cart.

    Note: here is a better example of an exclude multiple IP`s in GA: ^(xxx\.xxx\.xxx\.xx1|xxx\.xxx\.xxx\.xx2)$

    IP exclusion only works if the user has a static IP (most home ISP`s are dynamic) and the cookie exclusion you mention uses up the user-defined field in GA. This empty field is far better used for customer segmentation.

    Personally, I recommend using snipurl.com or tinyurl.com to create a link to http://www.mydomain.com/index.htm?utm_source=Exclude_My_Visit&utm_nooverride=1 then just excluding Campaign Source: “Exclude_My_Visit” this will last for 6 months or until cookies are cleared.


  7. @ Phil

    Fantastic comment. If there was a contest for best follow up comment, this gets my vote.

    * When using ppc – not recording actual search keyword by using a custom filter e.g $A3 in (\?|&)(q|p)=([^&]*)

    I have seen a few solutions for this (and this is a MAJOR issue) but have yet to implement any. The ones I have seen required using a modified version of the tracking script. Can you elaborate on this filter method for us or point us to a descriptive post (or get a hold of us and do a guest post!)?

  8. @Phil,

    Yup, that was golden. Thanks for sharing your tips and you really should identify yourself with a website for recognition!

    I agree with Jason, we could make an entire post out of your comment.

  9. Phil says:


    You can easily create a new GA profile to test the custom filter without effecting your main GA profile.

    One of the drawbacks of using the JavaScript “Sleuth” on-page method; is that it effects all profiles, not just individual ones.

    —- Filter method (needs to be add in order)

    1. Include PPC only visits
    Include only filter
    Campaign Medium (cpc|ppc)

    2. Clear User Defined
    User Defined: (.*)
    User Defined: (not set)

    3. Campaign Term > User Defined
    Campaign Term (.*)
    User Defined $A1

    4. User Defined + Referral Keyword > User Defined
    User Defined: (.*)
    Referral: (\?|&)(q|p|query|qs|qt|encquery|k|rdata|searchExpr|szukaj|terms|text|wd)=([^&]*)
    User Defined: $A1 ($B3)

    5. User Defined to lowercase
    Lowercase: User-Defined

    The above is an adaptation of this post: http://www.ga-experts.com/blog/2006/11/how-to-get-detailed-ppc-keyword-data-from-google-analytics/

    —- JavaScript “Sleuth” method

    For details read:

    —- Which to Choose?

    I checked some of the GAAC sites and only ROI revolutions uses this method. Even Brian Clifton`s website (advanced-web-metrics.com) does not use the JavaScript method, so voting with GAAC feet = custom filter wins by about 20 to 1.

    —- Other points

    I am still experimenting with these so I can not approve them yet…

    * Using 2 trackers and 2 user defined fields

    * Keyword funnel script this stores the first and last keyword/referrer. This helps overcome first click wins problems in GA and is useful for monitoring effect of branding campaigns, aswell as
    viewing keywords higher up the users buying process.
    http://www.ga-experts.com/keywordTracker.js (urchin.js version)

    Thanks for the offer of guest post, I am flattered :-) However, I would need to end ties with my current employer in order to be able post openly. So for the time being it is necessary for me to post as myself, not representing any organisation.


  10. RE: filter 2…

    2. Clear User Defined
    User Defined: (.*)
    User Defined: (not set)

    Should the second user defined entry be in the Output To -> Constructor field and should the value contain the brackets or not?

    re: filter 4…

    The comment display area is too small and cut it off: end of the regex is =([^&]*)


    shouldn’t there be a comma between $A1 ($B3) ?

    If I get this working nicely, I will create a full tutorial on it for everyone with a screencast (with Phil getting full credit).

  11. [...] up “8 Stupid Things Webmasters Do To Mess Up Their Analytics”. As Linda shares at Get Elastic, she’s trying to win the Marketing Pilgrim SEM Scholarship contest with this one. So read her [...]

Leave a Reply

© 2014 Get Elastic Ecommerce Blog. All rights reserved. Site Admin · Entries RSS · Comments RSS