If I were to tell you it’s 51oF outside, how would you dress? Is that cold or is it warm? Would you bundle up or put on short sleeves?
I mean, 51oF in Finland is a heat wave and Finlanders will wear lightweight clothing. But if it’s 51oF in Florida, people are breaking out the parkas and gloves.
It’s all a matter of context. And the same thing applies to performance metrics.
I recently attended a big-name webinar loaded with data points. I was overwhelmed with graphs showing me things like the number of click-throughs by daypart by media channel. I listened eagerly as the presenter shared what he perceived as the optimal number of emails to send in a series. Slide after slide after slide.
After about 5 minutes of this I thought, phooey. All this data was actually pointless. The presenter had provided no context. I had no idea of the circumstances surrounding the results he was reporting.
Hearing that tweets were opened and retweeted more frequently on weekends than on Wednesdays did me no good since I did not know who was tweeting what to whom and why. Ditto for every other media metric shared.
As an example, you’d expect that people who had previously bought Phillies fan apparel would respond in amazing numbers to an email offering 4 free tickets to the next Phillies doubleheader. If that result is part of the reported figures but unsegmented, it weakens my ability to use the aggregated response rates to project the performance of my email campaign to mechanical engineers in the utility business about a new heat dispersion technology we’ve developed.
Wow. That was a really long sentence. Here’s the key point – To make proper use of outside results, we need to compare “apples-to-apples.”
What I’d like to see next time are at least some sort of data source descriptions. At a minimum, separate B2B from B2C. Perhaps they could split it further into sports, retail, financial, coupon, event, and other offers. Then I could remove values for efforts that I know would skew results when trying to extrapolate response for a sci-tech effort.
Remember, data without context is just a bunch of numbers. Don’t be dazzled by sheer number of numbers. Make sure you understand where metrics come from, how they were developed, their timeframe, their biases if any, and so on. Only then will you be able to determine if they can be applied to your situation. And only then will the data become actionable information that can help you improve your results.