Tag Archives: web analytics

Segmentation comes in twos.

segmentation comes in twos

I can’t even begin to calculate how many times I have been approached to add a metric or segment to our web analytics solution. These should be good requests for an organization, right? Unfortunately, this is not always a positive addition. The problem emerges because the proposed segment or metric has no business case; it may be that it’s just a cool new way to look at something.  This type of petition should never be implemented or allowed.

Analysis paralysis is not only caused by mindlessly digging into your data with no end goal in mind. Adding metrics and segments to your analytics with no real business case in mind also causes it. Ask the tough questions when an excited executive or colleague approaches you.

  • How will this make us more efficient?
  • What business action will this create?
  • How will this help us drive more revenue?

The third question is by far the most important — or at least it should be.

When I dictate that segmentation comes in twos, I mean that there should never be a group or segment created without a desired group being created with it. The simplest example of this is email. You send an email to a list of 100 subscribers, which is your first segment or group (sends group). You want these 100 subscribers to open the email, which in this case is your desired group (opens group).

Segmenting in twos with a group and a desired group will create possible action items or new segments for your business. In my example, using email leaves only one alternative for those not opening your emails; you need to test changing the subject line. This is the only way to drive the desired group to a new action and move these subscribers into the new group. This is an example of a business action created by segmenting in twos.

Question every request regardless who is making it (C level included). If the executive team is smart, it will thank you later. Don’t spend time creating things that will be fun to look at or may wow that one executive in a boardroom. Spend your time creating business actions to drive more business and compel your executive team to pay you more money for helping the company be more successful.

What they taught you in school was wrong. 1 + 1 = 3.

One of the most important lessons in online marketing is the realization that if 1+1=2, then you may have failed. Everything we track must have a purpose, and typically that purpose is to combine that metric with another metric to give us the business action we desire. Metrics should be created with the anticipation that they will spawn other meaningful metrics for the business. The days of “I think” this was a good campaign because sales seem to have gone up are over. We are in the age of data-driven marketing, where the data doesn’t lie but it does tell us if the marketer is. The new black for companies is in their ability to properly measure their success and failures. This pursuit to know the “why” is an endless one because as you continue to dive deeper, you realize that the original answer has now multiplied into 10 smaller possibilities. The continuous desire to acquire the answers will keep you from stagnation and complacency. Every marketing decision you make today should have a cost-benefit analysis at the ready to justify the business case. No metric or report should be collected or created without an action in mind that justifies its existence. We live in a data-rich world; it is about time we start acting like it.