Noise or Music? - The Insights Blog

Understanding Web Analytics Accuracy – Whitepaper

Categories: Metrics understanding, Privacy and Accuracy, Pro Lounge / Comments: 8

Share Button

web analytics accuracy white paperI first wrote about web analytics accuracy in 2007 while working at Google. At that time numerous clients (big spending Google advertisers my team helped) were contacting their Adwords account managers asking why Google Analytics numbers did not match their AdWords click-through reports, or for that matter, match the other web measurement tools they were using.

These of course are legitimate questions. However there are a multitude of possible answers – not what you want to hear if you are the end-user trying to interpret your visitor reports! The original accuracy whitepaper (published in Feb 2008) explained all of the possible accuracy considerations I could think of at the time. It was a vendor agnostic accuracy check-list to help the end-user, and those that you report to, get comfortable with the data, its limitations and how to mitigate these.

Two years later and things have moved on. Accordingly I have updated the whitepaper to include all my latest thinking and add new data points from other people battling with accuracy – notably at study conducted by Paul Strupp and Garrett Clark at Sun Microsystems. The basic accuracy issues I originally describe haven’t changed – just augmented for April 2010 and now 19 pages…

If you are an agency with clients asking the same accuracy questions, or an in-house marketer/analyst struggling to reconcile data sources, this accuracy whitepaper will help you move forward. Feel free to distribute to clients/stakeholders. As before the whitepaper is vendor agnostic. That is, the principals and issues discussed are relevant for all on-site web analytics tools.


Download Whitepaper:
Understanding Web Analytics Accuracy

If you have further suggestions, or just want to let me know if it was use, please add your comments below. A retweet is always much appreciated :)

Share Button

Comments

  1. WSP says:

    Thank you for the whitepaper! I wish I had this two years ago. It would have saved me more than a few days of explanations, pointless reports and a couple of dollars. I definitely need to pass on the PDF to the new guys, here at the company.

    While working on several PPC accounts, I found accurate tracking and good loading times to be a challenge . A struggle, actually, on two fronts. Managing staff cares for leads and conversion rates, while web developers usually have no problem with web pages and forms loading 3-10 seconds, as long as the resources are secure and don’t misfire. Suggestions for improvements here need to be backed up with reliable data and an actual problem.

    Currently, Google Analytics will measure loading times for up to 10% of all traffic. Big sites can use the data. More traffic would mean better resolution and accuracy. Owners of smaller website should be careful though. Loading times between 5-10 seconds (according to GA) might not be as bad as one would think. If you are running a site of local importance and you measure good loading times there, than you shouldn’t really bother with data from other countries, which can bump up the averages.

  2. Metricks says:

    Hi Brian
    I had a query about accuracy in campaign reports, I’m having a problem where new visits are showing up in the last month for PPC campaigns that stopped well before the current time period (in Google Analytics) Presuming a selective cookie deletion issue possibly combined with attribution – any ideas?

  3. Hi Brian, would I be right in saying a lot of this is also in the latest version of Advanced Web Metrics?

    cheers
    Jon

  4. Lindsey says:

    My manager just sent me your whitepaper and I’m finding it really useful, particularly in understanding where people can fall prey to the data. I work at Compete.com doing much of their support and training so it can be difficult to understand where some of these misconceptions are coming from. Thanks for writing this, it’s a great resource!

  5. Thanks to Per Strid and Hailong (Scyan) for pointing out that Table 1 was mixed – the cells were labelled the wrong way round. Fixed now.

  6. Scyan says:

    Hello Brian, Glad to see the update version of this super white paper. I’ll read it carefully to check my own comprehension of web analytics data. Just a small notice: I’m reading the P5/19, the page tagging and logfile comparison table. Maybe something is not well placed on the table. Can you check this please? Thanks.

  7. Rob W says:

    Great Whitepaper! I have already printed out 20 copies and plan on giving them to clients that get “hung up” on why their website analytics data isn’t 100% accurate. Thanks for continuing to create these valuable white papers Brian!

Leave a Reply

Your email address will not be published. Required fields are marked *


7 − 1 =

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

© Brian Clifton 2015
Best practice privacy statement