Ten Surefire Ways to Wrestle Insight from that Pile of Data on Your Desk

The Paine of Measurement

Because research without insight is just trivia.

by Katie Delahaye Paine

Okay, I recycled that subtitle from an ad campaign we ran back in the '90s, but it's even more true now than it was back then. Between the automated monitoring systems, the Web analytics, Google analytics and a hundred other data points, today's communications managers are awash in data, but starved for insight. The market is hungry for standards, but truly malnourished when it comes to insight and figuring out what to do with all this data.

So, I thought I'd provide a fast lesson in what to do with a pile of data when it's staring you in the face. Here are ten straightahead ways to gain insight from your data:

1. Look for failures first
The most valuable lessons come from discovering what is not working. For instance, when the podcast you put up on YouTube has only been downloaded once. By your mother. Perhaps you need to look at the content that is being downloaded to see what's missing.

2. Tie the data back to what is most meaningful to you
If you're working at a non-profit, correlate it to online contributions, or to new volunteer signups, or to the percent increase in donors per mailing. If you're working in high tech, tie it to qualified leads, or requests for more information. If you're in the education marketplace, tie to application requests, or alumni participation.

3. Make sure your metrics accurately compare programs
It's great if you're measuring two different product launches by the degree to which the conversations about the new products are positive and position the product accurately. However, if you spent $10,000 on one and $500,000 on the other, you should definitely be looking at a cost per message figure to compare the two. Otherwise you're comparing apples to Beef Wellington and no one will believe you.

4. Don't show any data you can't make decisions with
Data not only has to be accurate, it needs to be relevant as well. So, for example, if you're comparing your spokespeople and thought leaders and the degree to which they are quoted compared to the competition, make sure that the VP who quit last week isn't on the chart. She's not going to be figuring into any of your future plans, so don't waste your presentation on her. (Unless, perhaps, she turns out to have been three times as effective as any of your other spokespeople. Then you have a valuable benchmark to work with, and maybe your spokespeople have some lessons to learn.)

Similarly, if you are tracking specific reporters or bloggers, make sure they are actually assigned to your beat. Many times bloggers will cover one story intensely, but then will never write about your industry again. So be careful who you call "The most influential blogger this quarter!"

5. Understand the "So What?"
What is the data trying to tell you? What opportunities is it trying to present to you? So what if you've classified all your conversations into the 27 different categories we recommend, now what? Now you figure out what types of conversations you need to pay attention to. (Read more in this article: "5 Reasons to Love 27 Types of Conversation and 19 Types of Video.")

6. Use data to make your point
Your data can be the clincher for your argument. If you are having trouble getting your senior management to show up at meetings with prominent bloggers, study what the competition is doing. Are they out there more prominently in key blogs? Are they on Twitter? Nothing like a little "His is bigger than yours" to get the juices flowing.

7. Give it context
Is my 20% negative horrible? (Not if you're Dell and you used to be at 80%.) How do you compare to others in the industry? If you're in the defense industry, 20% negatives isn't too bad. But in packaged goods or higher education, you'd be out of business.

KDPaine & Partners has recently studied social media use (social bookmarking sites, Facebook, external blogs, institutional blogs, and YouTube) in academic institutions. Below is a chart with some benchmarks you can refer to for general context. (We gathered the data over a 30-day period, September-November 2007, and included all references to five major universities. To ensure comparability and a manageable data set, content related to athletics was not included.)

8. Look for seasonal trends
We always present data on a 13-month rolling calendar so we know whether the slump in coverage is due to reporters being busy with political campaigns, or maybe to the fact that the agency or PR team has taken the summer off.

9. It's all about the conclusions
If you are telling me what the data means, tell me something useful. Don't tell me that I achieved 300 million opportunities to see -- tell me what got me there. And then tell me how I can get to 400 million.

10. Test, measure and test again
Measurement is an ongoing process. You're not going to learn everything the first time out. So when something doesn't work, recognize the opportunity to tweak it. Measure it again to see if the tweak worked. If it didn't, try something else.

Wishing you large measures of success,