Yes, it’s a great slogan. (Thanks to Gavin O'Malley and Pointroll.) I overheard it at OMMA, a conference that primarily focused on using analytics to measure online marketing. It was full of data wonks and stat geeks who cheerfully claimed credit for all kinds of stuff without taking into account any of the contribution from PR and social media.
Which, of course, got me thinking. The reason these guys (and gals) have all the power and credibility in their organizations is because they do use data to achieve their ends. PR people, on the other hand, have a tendency to let the data push them around. You’ve heard PR’s feeble data wrangling skills expressed at meetings, as in these real-life quotes I've had to deal with:
- “I really don't believe that PR data, I just know that direct mail works.”
- Or, “There’s really no data that shows that social media/PR was effective.”
- Better yet, “You really can’t put any data behind that relationship/reputation stuff.”
And why do we let this happen? Because we aren’t familiar with statistical analysis -- or are afraid of it, or don’t know how to do it, or haven’t thought of using it. We let everyone else throw their data at us and we just duck under the table. Then we can’t figure out why we aren’t getting a seat at the table.
So how do we turn this around? Make data your bitch:
Use data to achieve your goals. Don't let it push you around.
Take control of your data and make it work for you. Don't let it just arrive in waves and pour over your desktop.
Understand your data’s strengths and weaknesses. Leverage them so that you win in the end.
Don’t be afraid of data. Look it full in the face and say, “Bring it on!”
Here are ten easy steps to dominating your data.
1. Trust your data.
Make sure you’re getting data you can have confidence in. If you aren’t sure of your data, you won't use it like the incredibly powerful tool it is.
- If you’re using an outside vendor, ask to see their inter-coder reliability scores. Ask them for their confidence levels and do periodic checks to make sure the data is coded the way it should be.
- If you’re doing the data analysis yourself, ask a colleague -- or better yet a graduate student at a local university -- to check it for you.
2. Know your data.
Know it backwards and forwards, inside out and upside down. Take the time to learn the details. If you’re presenting, get prepared by having a friend or colleague ask you the really tough questions ahead of time. Have the answers at your fingertips.
3. Draw conclusions.
Don’t just say, “There was a big spike in June,” or, “The numbers went up 10%.” Provide the “So what?” Like this, for instance: "The numbers went up 10%, which was significant not just for the increase, but because we accomplished it with half the budget. Therefore we recommend…”
4. Tease out the really interesting stuff.
Here's an example: We were recently reporting on a thought leadership program in which one organization consistently dominated the industry. Normally, we look at the comparative share of thought leader quotes for each organization. In this case we noticed that the dominating institution didn’t just have the greatest share of quotes, they also had the highest number of individuals that were quoted. We dug into the data to discover that it was the depth and breadth of their program that enabled them to dominate in every area. If their leading expert on swine flu was busy with another interview, they had 4 or 5 others waiting in the wings all with equally fancy titles. The lesson here is that if we hadn't really dug into the data, we never would have discovered what the competitive advantage was.
5. Don’t just do the standard comparison.
Everyone always compares data month-on-month or year-on-year. We recommend looking at data over a thirteen month window to spot long term trends.
6. Take a good, hard look at the bad stuff.
You’ll learn a lot more from your failures (and look like a hero for stopping a dog of a program) than you will from fixating on a small improvement in performance. Even more interesting: Look at what the competition is doing right. Where are they beating you? Finding out what customers like about the competition will yield much more insight than just listening to them complain about your own company.
7. Beat the bushes.
Run correlations on anything you find interesting. So it doesn’t correlate, no harm done. Move on. Although, sometimes you can learn more from what doesn’t correlate than what does.
8. If something doesn’t make sense, pounce.
Now that you really trust your data, trust yourself when something looks a little off. Get down into the dirt and figure out why. Nine times out of ten, it's not the data that wrong, it’s the program.
9. Increase the depth and breadth of your data.
Beg, borrow and steal data from throughout your organization. Get it from online sources. Take your Market Research department to lunch and see what they’ve got. Track down your competitive intelligence gurus and take them to a ball game or dinner. (They’re the ones who really have the data. And the budgets.) Walk into your corporate library with a batch of fresh homemade chocolate chip cookies and find out what they have access to.