Recently, I had a chance to read through Avinash Kaushik‘s new book, Web Analytics, An Hour a Day. It’s a great book because it helps you learn how to view analytics the right way. In this article, I am going to outline what I took away from the basic philosophy Avinash defines in the beginning, as it sets the tone for the rest of the book.
Analytics From the Top Down
With Web analytics, it’s amazing just how much you can do to improve the performance of your site when you dive in deep and start digging through your data. One of the hardest parts about it, though, is figuring out how to focus on the right data – the data that gives you insight into what is happening with your Web site. It is really easy to get lost in the reports provided by the tools and not derive any critical new meaning.
Right from the start, Avinash gets you focused on the right way to think about analytics from the top down. One of the first questions he gets you focused on is, What is the purpose of your Web site? Much of your analytics strategy should emerge from the answer to this question. In addition, Avinash helps us understand a broader vision of analytics.
Analytics is much more than just the Web analytics tools that you can license today, for example: Clicktracks, Coremetrics, Google Analytics, IndexTools, Omniture, Unica Affinium NetInsight, Visual Sciences, and WebTrends. These are all leaders in the space of analyzing Web clickstream data.
Analyzing Web Site Performance
But there is so much more you can do than just tracking clickstream data. As Avinash tells us, clickstream data can help you “infer the intent” of your users, but ultimately, “inferring” is the best you can do with this type of data. That’s not to say these inferences are not valuable; they are.
However, you also want to look at the outcome. In other words, how is the Web site performing in terms of achieving its purpose? Are we getting the revenue, leads, or whatever we are expecting? What are the conversion rates? Early in the book, Avinash identifies this as your top priority in analytics. In fact, he says, “Is it a bit extreme to dump clickstream in favor of measuring outcomes first? Yes. Necessary? You bet.”
This ultimately leads you to the use of additional tools. The quality of information available from your traditional Web analytics tool is too poor for you to rely on this for your outcomes analysis. For this, you will need to do other things. At the end of the day, if you are running a PPC campaign, you need to know if you achieved your objectives. For example, if you are running an e-commerce Web site, you need to know how much money you made and get as much insight as possible as to why you did.
Avinash is also a big advocate of understanding the mindset of the customer, and why customers do what they do. This takes you into a wholly different direction in analytics. Finding out why depends on asking the customer directly. Remember, that clickstream data can only infer intent. To learn more, you need to go straight to the source, the customer himself or herself.
This means doing things such as usability testing and customer surveys. Clickstream analysis can contribute partially to this effort through the use of A/B tests and multivariate testing, but it’s a wise idea to invest a broader level of effort in other tools so you can really understand the customer experience.
Another area Avinash focuses on is competitive intelligence. Great services exist, such as comScore and Hitwise, that can provide you direct information on what your customers are doing. While this is obviously not for mom and pop Web sites, these tools are incredibly valuable for large businesses to utilize.
Avinash recommends that companies spend only 10 percent of their analytics budget on the tools, and then 90 percent on the people who will use those tools. At the source of this recommendation is the fact that Web analytics is hard. Unfortunately, it is much more than opening up your tool and looking at the visitor, unique visitor, and page view stats.
Data of this nature may be interesting from a general trending perspective, but if this is all you are doing in Web analytics, you aren’t really doing Web analytics. At the core of high value Web analytics is a real analysis of the data to increase the ROI you are getting with your site. This means things like customer segmentation, A/B and multivariate testing, using Site Overlays to see click-through behavior, and more.
While the discussion above is focused on the first few chapters of Web Analytics, An Hour A Day, the remaining 15 chapters are packed with no-nonsense, direct advice on where to focus your time to get the best results, specific advice on how to perform certain types of tasks, and how to build a data-driven culture.
If you are looking for a place to learn about analytics from the ground up, and also get into lots of specifics and details, this book is an excellent resource. In addition, 100 percent of Avinash’s proceeds from sales of the book are being donated to two great charities: The Smile Train and Doctors Without Borders.