In the seven years since Google Analytics launched, the realm of analytics has changed hands from the purview of the IT department to the marketing department and beyond. But just because we can readily access the tools does not mean that we are necessarily putting them to good use. To gain some insight into this we interviewed Brian Clifton, author, consultant and trainer, who specializes in performance optimization using Google Analytics.

Brian acted as head of Web Analytics for Google Europe from 2005-2008 and recently released the third edition of his "Advanced Web Metrics with Google Analytics," used by students and professionals worldwide. 

CMSWire: Do people collect data that is useless or unnecessary?

Brian Clifton: No, and let me say that I don't think collecting *any* kind of data is useless. However, it can be distracting and generate a higher level of "noise." The mistake I see most often is people not focusing i.e., they simply look at all of the data all of the time. Beyond the high level reporting of metrics such as total visitor counts, total revenue etc., looking at all of the data all of the time has very little analytical value. Therefore, what you need to do, is chop up the data into meaningful groups of similar data points i.e., segment.

Example segments to enhance the signals from the noise:

  • Customers
    • A segment that includes only visits from existing customers -- even if they do not purchase from your site when they visit. This is why this segment is so valuable -- these visitor have bought from you before!
  • Engagers
    • After customers, Engagers are your next most valuable group of visitors. They have shown engagement, i.e., made an effort, perhaps even given you their contact details. Examples include registrations, contact requests (form submissions), file downloads, ranking products and content, comments and feedback, video views, clicks on social love buttons etc.
  • Mobilers
    • A segment that includes only visits from mobile devices. Mobile is a very different experience to browsing on a bigger screen. Lots of implication for your web site structure/design if this segment is significant to you.
  • Non-Bouncers
    • A segment that excludes all bounced visits i.e., removes single page visits that have no engagement. The theory is that these visitors are clearly not interested in your website, so remove them from your analysis.

Of course these segments are not mutually exclusive and you can be very creative by combining them.

My point is that data "noise" is not something to be discarded. It's only noise relative to a specific question you are trying to answer at a particular time. When you look at answering a different question, that noise can be a rich data stream indeed (people have won nobel prizes for analyzing noise!). Therefore, always collect the full spectrum of web data and user segmentation as your tool for focus and analysis.

Editor's Note: To get more insights from Brian, follow him on Twitter @BrianClifton

CMSWire: How can a company decide which data to collect? Are there common metrics/strategies that work across the board or does it change on a case by case basis?

BC: My approach when I first engage with a client is not to show or look at any analytic reports. I usually get an odd response to that when I have been hired as the web analytics expert! However, the first meeting is a workshop where I sit down with the marketing team to understand their needs and abilities. By needs, I mean understanding what tasks they perform as a group and therefore how they can measure the success (or not) of that. A metric that measures success is known as a Key Performance Indicator (KPI).

For example, as a group we investigate:

  • What is producing the lead generation (call to action)?
    • this can be a contact-us form, interactive quote generator, registration sign-up
  • What is driving the call to action?
    • Video demo, fact sheet (PDF download), on-site testimonials/press/PR, your detailed support section, special offers
  • Which are the key landing pages?
    • Often this is neglected with a default answer of "the home page." However, where a visitor arrives on your site is critical to getting them to convert.
  • What are the key marketing channels used to drive visitors?
    • Usually a mix of SEO, AdWords, banner ads, email marketing and social activities. Often these are not tracked correctly. By that, I mean the arriving visitor is tracked, but not assigned to the correct channel/campaign they came from. In fact, 9 out of 10 times I find email visits incorrectly labelled as a "direct" visit. That is, campaign tracking is not being used.

Once you understand what constitutes success, the actual KPI measurement is relatively straight forward. For example, what proportion of visits (the conversion rate) watch your product demonstration to completion simply becomes the number of video view completions divided by the total number of visits. Tools like Google Analytics automatically track your KPI conversion rates for you. You just need to tell it what the success metric is.

So the process is:

  1. define your KPIs
  2. measured your KPIs i.e., implement a best practice setup of your tracking tool
  3. segment KPIs to understand what is working for you

Most people start at #2 -- focusing on implementation without the focus on the important metrics. That results in a lack of buy-in from senior management from the get-go and leads to frustration all round.

CMSWire: What key changes have you seen in analytics and analytics tools over the past few years?

BC: The BIG seismic shift was when Google acquired Urchin Software in 2005 and later that year launched Google Analytics (14-Nov-2005 -- a date I remember well as we celebrated the launch in Madrid with the Google Spanish office team!). The two major changes were making the product free to use and specifically creating a product that was targeted at marketers when the rest of the industry was targeting IT teams. It was, and still is, a revolutionary approach that even now the industry has not adjusted to.

Following this, the industry scrambled to consolidate and survive i.e., mergers and acquisitions -- it is after all hard to beat free as a price tag. That said, I do feel the industry spent way to much time, energy and money trying to pick holes in the GA product rather than innovating in their own unique ways. In those early days it was part of my job at Google to counter a lot of the myths and mistruths that were spread around.

Today companies are much more comfortable with Google collecting anonymous web data on their behalf. At the beginning of 2012, it was estimated that half the web and 45% of Fortune 500 companies use Google Analytics.

CMSWire: What do you see as the future direction of analytics?

BC: For me, the big challenge moving forward is the ability for tools to be able to stitch together the multi-device world. The explosion over the last two years of smartphones, smartTVs and tablets means that the user journey is one that the current tracking technologies (i.e., cookies) cannot cope with.

For example, one person using all of these devices online is tracked separately as 3 different unique visitors. Combine this with the fact that smartTVs and tablets are often shared devices at home, and you can see the problem rapidly becomes untenable. Solving this issue, is the where the industry needs to head if web analytics is to remain a viable tool for marketers and website designers.

Incidentally, companies such as Microsoft, Google and Apple are capable of solving this issue as they control the "Internet view." However, apart from the technical challenge there are also privacy issues.

Privacy is something that has been in the background since the beginning of the web and has only recently started to become more publicly discussed by the mass media. The industry need to balance the tracking needs of website owners with the right to not be tracked as an individual. My view is that this is a much greater challenge than the technical problem.

A number of industry luminaries advocate that web analytics will soon merge with business intelligence -- I am not so sure. My reasoning is that the vast majority of visitors to your site are anonymous, with only a small fraction (typically 5 percent) going on to identifying themselves either by registering, subscribing or purchasing from you.

Business Intelligence is all about data mining your customers. That is clearly important. However optimizing your site and its marketing to the other 95 percent of unknown visitors clearly has a much bigger potential to your business. That is where web analytics fits i.e., outside of BI, though communicating with your BI tools when they become a customer is key and has always been possible.

CMSWire: Is there a gap in experts who can provide analytical skills?

BC: There is certainly a very large skills gap (there always has been in this industry) with most "analysts" doing the job part-time i.e., usually less than one-day per week fitted around their other duties. Even when the skills gap is filled, often the person is junior from a business point of view. That is, they can be technically and analytically very knowledgeable, but are not part of the strategic business core of the organization.

A “junior” person driving website change -- be it how to best drive traffic to your site, or how to optimize your conversion of an anonymous visitor into a customer -- is always going to end in frustration. Because the powers that be don’t trust the business accumen of the person advocating the change. When that gap is closed, web analytics will really take off and there really is so much to be done to improve the web ...