Top 10 Web Analytics problems


Following discussions hosted by Jim Sterne during the Emetrics Summit in London, the attendees came up with the following list of major problems, encountered during a Web Analytics project. We've takken the liberty to select a few that we regularly also face.

  1. No objectives: amasing isn't it? yes well this is for most of my clients and probably the reason why they don't like me in the beginning as they get a beating in the form of : "what are you actually trying to acheive? you're spending thousands of euros on a website and for what? what's the bottom line and how can we measure it?". Once we got over that one, the fun part starts 😉
  2. Accuracy myth: it's not totally accurate, we all know it. As Avinash said, get over it. Data is there, more than ever before. Use it!
  3. KPI benchmarks (by industry): yes, conversion rates can be so different according to the type of website you have, what you consider to be conversion, how conversion is calculated, etc. My suggestion would be start with your own conversion rates and build up from there. The industry is still in it's inphancy, don't expect the published figures – if they exist – to make much sense, it's too soon!
  4. Standards & definitions: the Web Analytics Association is working on it (and doing a pretty good job by the way) so, for more information, cf.
  5. Senior Management Education. Personnally, I've given up on that one. Don't expect senior management to understand the difference between hits and page views but once you've resolved issue 1, you should be able to talk to them about their website in a way they can understand as well.
  6. Web Analytics for non e-commerce, retail website: it's getting better as the examples that I see or less and less about e-commerce. We use some e-commerce features to give fictive valuations to certain pages. I wasn't convinced in the beginning but was willing to give a try. Today it's become part of daily life as it allows for benchmarking of content value and actions.
  7. Data Integration (mobile, rss and BI): yes a huge issue this year. We are encoutering more and more systems feeding on one another from CRM or CMS to WA. Now, I'm not totally convinced that a WA tool should be the hub of it all but should at least allow for choice of where the data is going to sit, in the end.
  8. Privacy and legal issues, matters, disclaimers: I wanted to stress that one bause it's too often forgotten. My clients go "Yes, yes, no problem" and then I can wait for the legal department for over 6 months while the tagging is running and data is being collected. So, please, don't overlook the legal aspect of Web Analytics.
  9. Tagging as part of a process, before the launch: this is more about setting up procedures within a company. Once you start a campaign or any kind of communication around your website, make sure that adequate tracking is part of the project, before the project starts. After the campaign might be too late. Thus, adapt your checklists and processes so that you can take action for the next campaign based upon measurable findings!
  10. Measuring value: this is the last of the top 10 problems that I wanted to take into acocunt because I think that this is the way to go if we want this industry to move forward. We have to be able to attribute value to what we are measuring. It sounds stupid but believe me, it's not that easy. How do you attribute value to for example the top left box on your homepage? We use darwinist methods for the time being: what gets clicked upon stays, the rest is out but value is not only about clicks. How do you determine this? rules have to be set-up, models and tested out. These rules will also evolve depending upon the company's capacity to modify content easily and clearly define objectives of the website. Needless to say that we are not there yet.

I've taken from the Emetrics list my current top 10 Web Analytics issues. Some issues such as accuracy can be met by the vendors, others by more formal organisations such as the Web Analytics Association.
The rest is up to us, the Web Analytics professionals and our clients to work together in moving forward this young industry that could revolutionize the way marketing activities are undergone today.

You'll find here the top 10 Web Analytics problems found during the Emetrics 2004 @ Santa Barbara (Thanks to Stephan Spencer)


2 Responses to “Top 10 Web Analytics problems”

  1. Web Analytique » Les 10 principaux problèmes du Web Analytique Says:

    […] Top 10 Web Analytics problems (US): l’article original en anglais […]

  2. Web Analytics Wednesday April 2007 - Brussels - The Presentations « Blog Says:

    […] presented his view of what can go wrong on a Web Analytics project based on his 5 years experience as a WA Consultant. This was the first […]

Comments are closed.

%d bloggers like this: