Saturday, November 21, 2009

Guide to making Web Analytics more accurate!

A recent eMarketer study identified top issues with the accuracy of analytics. Most prominently, users said they can't drill into the data (42%), and called out the issues with marketing attribution (32%), campaign tracking code (25%) and cross site analysis (20%). These factors become prominent as the business evolves over time and market driven changes are incorporated across the site. However, with some planning and framework, we can define and implement a robust analytics strategy that is flexible and adapts to the evolving business needs and site-wide changes.

The first step is to take a step back and outline the core strengths of the business, around which metrics will be defined. This means that if we are a content focused site, we need to ensure the page view depth, SEO traffic growth, etc., are top measurement priorities. Further, monetization strategy of this traffic should be defined, in close coordination with sales and business development. (The exercise needs to be repeated at the page and/or section level). For an ecommerce site, it will be highly desirable that conversion metrics are aligned with the strategy, traffic engagement and monetization.

Web sites have three key ingredients that define its performance.
1 - traffic into the site
2 - navigation behavior within the site
3 - conversion from the site (usually lead/revenue generating click)

The key to a robust site analytics data is to be able to connect the above 3 ingredients in a flow that helps explain the value of the user coming into the site. In my previous experience, we ran into several issues in trying to develop reporting that provided this level of visibility.

Basically, a campaign code assigned to a particular promotion needs to be tracked through the navigation cycle, up to the point where user leaves the site. This helps us understand which campaign drives most traffic, how do users from various campaigns interact with our experience, which campaign drives maximum value towards our internal goals (engagement, lead, revenue, etc.).

Ideally, it is best to work with ONE vendor in implementing the tracking and reporting system, so as to ensure consistency, data integrity and institutional knowledge - identifying a vendor that provides the best possible solution becomes much easier, once the step one above has been defined at the onset. Onsite staff that understands the business and the analytics implementation can then provide insights and recommendations that are sustainable.

No comments: