April 22, 2010

Measurement

My entire career is based on the measurement of ecosystems.

When you hire me for a Multichannel Forensics or Online Marketing Simulation or Digital Profile project, you ask me to "tell the story" of "how everything fits together over time". Given how everything fits together, you ask me to find ways to reduce or reallocate your marketing budget.

97% of the measurement audience focuses on the measurement of events. We pretend to measure multi-dimensional events --- the online marketer tries to tie telephone orders to e-commerce orders, in order to properly measure paid search activities --- the catalog marketer tries to tie e-commerce orders to telephone orders, in order to properly measure catalog activities. If we can prove that a series of activities generate a positive outcome, then we can defend additional activities, can't we?

The biggest difference between my projects and campaign-based analytics is "purpose". When you use web analytics to measure a campaign, you know whether the campaign worked or not. Your analysis fails to communicate what impact the campaign will have on your business two or three years from now.

Your typical Multichannel Forensics or Online Marketing Simulation or Digital Profile project does the opposite. These projects do not consider the impact of specific campaign. Instead, these projects focus on the impact of a family of activities (i.e. search or e-mail marketing) on the future ecosystem of a business.
  • If you double your e-mail marketing frequency, what happens to store sales two years from now?
  • If you stop your paid search program, what happens to e-mail marketing results three years from now?
  • How might social media shoppers evolve differently than traditional catalog shoppers?
I have yet to find anybody in the web analytics world who focuses on ecosystems. Web Analytics evolved to measure campaign effectiveness, or event effectiveness, or visit habits. The element of time is not of primary concern in this discipline.

Measurement has never been easier, or cheaper. And yet, we're not measuring everything we should be measuring. Too few of us are measuring ecosystems. To many of us are measuring lighting bolts within thunderstorms, events that cannot be replicated, events that do not influence what will happen two days or two years from now.

5 comments:

  1. Anonymous5:27 AM

    Kevin, you are completely correct, in fact this was the very topic I covered at eMetrics albeit under a different presentation title. I'd like to connect with you offline and share a few of our approaches with you, perhaps collect some feedback and receive any constructive criticisms.

    Ross - MadisonMetrics

    ReplyDelete
  2. Great post, and so right on! I've focused on this kind of measurement with my businesses for years, but have never been able to express it this clearly.

    I've always thought of this "ecosystem measurement" as getting your hands deep in the data and there being some intuitive feel and understanding of how it all fits together.

    Formalizing that quantitatively isn't addressed through current analytics tools, but if done well, would be incredibly powerful indeed.

    Again, great post that got my wheels turning!

    ReplyDelete
  3. Very smart, but very challenging as well. As a digital media company that has used traditional media like television to drive traffic, we have big lingering questions similar to those you raise. However, we've never found a really good way of measuring the constructs and telling the story.

    ReplyDelete
  4. Thanks for the comments, everybody! This kind of work is a lot more stimulating than the measurement of a myriad of online marketing programs!

    ReplyDelete
  5. Brilliant post Kevin. Some of the things which each one us thinks. I will try to see if I can work out a way to overcome this challenge put forward by you.

    ReplyDelete