Thursday, November 17, 2005

5 Magic Metrics

I recently had a conversation with the CMO of a leading global technology company in which they described for me their desire to construct a marketing dashboard focused on the “5 most important metrics” for their business. When I asked her what those might be, she quickly began to list the possibilities – around 10 of them as I recall – before she stopped and admitted that the task might not be so easy.

The more we talked, the more clear it became to her that getting down to the 5 Magic Metrics would take a diligent effort of experimentation and elimination, perhaps starting with 40 or 50. The appropriate metaphor was the old story of “I’m sorry this letter is so long, I didn’t have time to write a shorter one.” The risk of jumping too fast to the logical 5 is that you might select the wrong ones and achieve the wrong goals in a very efficient and effective way.

If you want to get down to the right 5 (or 4 or 6 or however many) metrics that really forecast success, you owe it to yourself (and your CEO) to undertake a thorough exploration of the 30 or 40 hypotheses that would emerge from a cross-functional assessment of “what really drives the business.” You’d probably not be surprised at the lack of consensus within even the best-managed companies on which 30 to even start with.

From there, it takes a bit of effort to acquire the data to test each of them for diagnostic and predictive ability, or to develop a proxy approach for the inevitable majority of metrics where the data doesn’t exist. Not that it can’t be done quickly (read: a few months), but it does require a deliberate effort.

So whenever I get the CMO request for “the 5 magic metrics,” I agree with them that it’s a great idea to strive for simplicity and to align your marketing measurement framework or dashboard to reinforce the company’s specific goals. But I also advise them to be careful about how they issue that direction to their teams, lest they create the impression that they’re only interested in simplicity (which might be interpreted as superficiality), or they send the message that speed is more important than accuracy. They’re both important.

Start with a hypothesis on what the 5 key metrics might be on the highest level on your dashboard, but don’t sacrifice the real insight derived from exploring the broader spectrum of options and validating your hypotheses. The difference will be measured in the credibility and longevity of your measurement plan.

No comments: