What to expect from your web analytics

Many companies don’t know what to expect from their web analytics.   They implement an analytics solution, invest a lot of resources, and hope for the best. From there they review the most simple reports and traffic data (e.g. hits, popular pages, time on site). But web analytics can do much much more then that.  

Because web analytics is a relatively new area many companies do not have a set of expectations like they have for established departments like accounting, sales, and IT.  When any one of those departments aren’t living up to their expectations it becomes very clear very quickly.  With analytics the definition of success is murky.  There are no clear set of expectations.  This article will give some insight into what you should be getting and what you should expect from your web analytics.

A QUICK TEST
If you don’t trust your analytics enough to make decisions on your website your analytics solution is falling short.
It’s that simple. If the data from your analytics can’t be trusted there is something wrong. Usually there are good reasons not to trust your web analytics results. Some people just don’t feel good about some of the data their analytics solution gives them, and for good reason. These are a few of the common reasons:

• Sloppy Implementation – If the analytics solution was installed incorrectly from the beginning the data will be incomplete and dangerous. Most companies struggle with this step. The problem is compounded by complex website structures, lack of testing, and using the ‘vanilla’ or most basic level of implementation. This step is ongoing and is never “done”. As your site changes your code should be tested and updated regularly.
• Reports instead of insights – Data overload is a real problem. A bad web analyst will unload a pile of mind numbing reports filled with lots of data. When asked what it all means a bad analyst will spew more technical definitions and add more unneeded info. This is bad. A good analyst will give you simple ‘actionable insights’ about what the data is telling you. Good analysts interpret and refine the data. Without analysis the data is near worthless because no clear action is defined.
• Meaningful KPIs – A Key Performance Indicator (KPI) is number that tells you the health of you website. KPIs define customer satisfaction, shopping ease, quality of content, and customer intent. If your traffic fell by 20% one night a KPI should tell you why and how to fix it. KPIs are custom built for upper management, marketing, IT, and sales departments. There are few ‘one size fits all’ answers when it comes to KPIs. Each department should be focusing on a different set of KPIs. If your main KPI is ‘Hits’ on your website your analytics solution needs work.
• Positve ROI on web analytics – You should know for a fact that your web analytics saved you thousands of dollars in PPC, doubled your conversions, or saved you untold hours of wasted time by guiding your decisions. If all you have is a “hunch” that your web analytics is helping, this is a definite sign that your analytics is under performing! Your web analytics solution should be quantifying the value it brings to your company.
• Lack of Testing – A good analyst will always have A/B and Multivariate tests running on your site. They will be testing the most critical pages and optimizing there performance. The checkout pages, landing pages, and high traffic pages should always be improving. Instead of making sweeping changes or bull headed updates all changes should be done within statistical significance. If there is ever a conflict about the site the resounding solution to the disagreement should always be ‘Let’s test it!’.
• Know the reason traffic has decreased – When your website’s traffic is 20% below normal your web analytist should be able to tell you why. Blaming the weather, the economy, or saying it’s a fluke, are all signs that the web analyst doesn’t know. A good analytist should have your traffic sources segemented and KPIs set up on page and server performance. A good analyst will tell you that the dip in traffic is due to something definable. Exeptable answers might include loosing a major organic keyword, a high traffic landing page being broken, a server going down, a major PPC campaign being paused, or something more definate.

You should expect much more from your Web Analytics and your Web Analyst. The analyst position is so important that it should be given ample resources. With more resources should come higher expectations. So raise the bar and expect more from your web analytics.