For 10 years I have been working with companies to optimize their websites to increase conversion rates. During this time, I’ve used what seems like dozens of web analytics tools.
The insight that is available is incredible if you don’t make the “7 Biggest Mistakes of Web Analytics”:
1. Improper Implementation
According to my friend, Stéphane Hamel the developer behind W.A.S.P., a FireFox plugin for detecting and reviewing web analytics implementations, nearly 100% of implementations are setup improperly. Common mistakes include: untagged pages, untagged or wrongly tagged transactions (i.e. transaction page is tagged like a regular page) and passing wrong values (especially for advanced tools like SiteCatalyst). In fact, several web analytics vendors websites also have these common errors.
To make sure your analytics don’t suffer you need to pick User Acceptance Tests (UAT) tests and re-run them specifically to look at tag quality (WASP can do a crawl that will work for content areas, but for transactions, there’s no better tests than actually running transactions to monitor the results.) Additionally, you should run continuous audits. Sites evolve and change, and the tagging quality suffers since different parts of the site may have been tagged at different times, etc.
2. No Goals Setup
The purpose of web analytics is to provide information about how well you are doing. Defining goals in the tool defines what game you are playing. How can you keep score without knowing where the goals are? Approximately 80%, of implementations have no goals setup. If you can’t define what is valuable to you, then how do expect to increase your results?
Even if you don’t do commerce you should have goals setup. Common non-commerce goals include tracking your inquiries, subscribers, white paper downloads, webinar attendees, etc. The key to success in setting up goals is aligning your goals with your customers’ goals.
3. No Segmentation
Repeat after me…”Not all traffic is equal.” Any analyst worth their weight in salt will tell you that the greatest insights don’t come from average and aggregated data but from slicing and dicing the data to produce intelligent segments. One of the most obvious is first time visitors versus repeat visitors (there are probably a couple of different segments in your repeat visitors too).
One of the features that thrilled me about the latest Google Analytics release is the ability to setup advanced segments. Bloggers surely know, that rss readers behave very differently than social media traffic. The reason FutureNow develops personas for our clients is to insure that we look at the website from these different perspectives/segments and measure them that way too.
4. Paying Too Much Attention to Irrelevant Data
“Web data is dirty data.” Never in the history of Analytics have we been able to collect so much information about visitor activity. This however leads to a lot of noise, due to things like cookie deletion, individuals browsing from multiple machines, different collection methods and definitions, etc. This is just one of the reason why you should never focus too much on data accuracy but on relative trends.
The second issue here is that all the people who rely on so many data points that they have no way of keeping an eye on them all. My core philosophy is that if you can’t relate all your reports back to how they fit into your financial statements then it should probably not be reporting it. Only focus on the metrics that you can actually control.
5. Not Setting up Milestone Events Documentation
With any luck, your business changes. You send out emails, run new campaigns (online and offline), make changes to your pages and other things that could impact your web results. However, most companies do a terrible job of documenting when these changes occurred and correlating them to their web analytics results. It would be great to have this information in your web analytics solution, but you could also run a private wiki that lets everyone on your team leave documentation about these changes.
6. Not Combining Quantitative Data with Qualitative Data
What I have learned in my 10 years plus of optimizing websites using customer-centric persona tools is that you need to be data driven but customer focused. If you forget to include Voice of Customer analytics in to your mix, you will tend to skew more to the cold data driven side (left side of the brain thinking) and neglect many of the opportunities on the softer voice of customer side (right side of the brain thinking). The data side is often focused on the what and how many, but the voice of customer side gives you additional insight into the why. Use tools like TeaLeaf, Bazaarvoice*, iPerceptions, Omniture Survey, ForeSee, and OpinionLab. Here is a great example of how to use web analytics and voice of customer analytics together. You can always start for free using 4Q from iPerceptions* to get some early wins.
7. Not Taking Action On the Data
Unless you are in the business of research, collecting data without acting on it may qualify for the definition of insanity (ever seen the underpants gnomes on South Park?). The purpose for investing in web analytics is to make data-driven, informed decisions and not just rely on gut instincts. The reason to invest in turning your data into insights is to become a smarter marketer and to produce better results. This is the very core of a Six Sigma approach to a continuous improvement process. Use the data to make changes to your website, feed your email campaigns and tools like Google Website Optimizer, Omniture’s Test & Target, CoreMetrics Intelligent Offer, or personalization tools like Monetate and SiteBrand and start increasing your results.
Once you have avoided these 7 biggest mistakes you are well on your way to web analytics success.
* Disclosure: I am on the Advisory Board of Bazaarvoice and iPerceptions.
Get Social: Join us on our Facebook Page & follow us on Twitter: http://Twitter.com/adCenterBlog