Monday, December 19, 2011

One Prediction and Five Web Analytics Tips for 2012

For past few years I have made several predictions about Web Analytics. This year I am going to make only one prediction but will provide five tips for 2012.


This year the push will be towards “Multichannel Analytics”. Integration of various data sources, e.g. email, CRM, social media, call center etc. , with Web Analytics will take center stage.

Five Tips for Web Analytics

  1. Expand your web analytics to consider other data sources
    We all know by now that no one channel exists in isolation. Web, email, mobile, social media, catalog, stores, call centers etc. all impact each other. Web is a just one part of the customer’s experience and journey towards purchase. To fully understand customer behavior and optimize your marketing you have to go beyond web analytics and look at data from other channels.
  2. Move from “How Many” to “Who”
    Majority of the web analyst today analyze “How Many” e.g. how many people landed, how many bounced, how many converted etc. “How many” is a great start but it is time for you move to “Who”, e.g. who bounced, who did not convert etc. and then think about how to engage with those “Who” did or did not do something. (if you need help with this then ping me)
  3. Understand the data structure behind your web analytics data
    I am surprised that many web analysts today don’t understand how the web data is structured, how it is collected, where all the variables that are passed in your JavaScript end up at and how various data elements are related to each other. If you are one of those analysts, take some time to understand the data structure. Open a raw web server log file and start from there. If you company is porting the web analytics data into a database then open up that database and look under the hood.
  4. Learn SQL
    This is going to be critical. You can only do limited segmentation and optimization with aggregated data that is provided in the web analytics tools interface. To really understand customer behavior and capitalize on that you should be able to extract the data from the backend. Even if you are not going extract the data yourself, having an understanding of SQL will give you tons of ideas on segmentation, optimization and targeting.
  5. Make friends with “HiPP” (Highest Paid Person) and say goodbye to “HiPPO”
    HiPP is your friend, not foe. If you really want to create a culture of analytics in your organization then make friends with HiPP, get them on your side. You need their support. Stop using the term “HiPPO”.

Comments? Questions?

Web Analytics Jobs

No comments:

Post a Comment

I would like to hear your comments and questions.