Turn Your Data Sideways, and 5 More of Katie’s Secret Tips on Finding Insight

Over the past quarter I’ve been asked to review probably two dozen reports, ranging from media analysis reports to brand tracking studies to internal employee surveys. What they all have in common is a startling lack of information from which the client can draw conclusions.

There’s been nothing wrong with the data itself. There’s plenty of it, in a variety of flavors, shapes, colors, and sizes. But in the end, I always have to ask, “So what?”

The answer isn’t more data. The answer is asking the right questions and getting the right cross-sections of the data.

I always manage to wring some marketing insight from these data sets. Here’s how I do it:

Katie Paine’ 6 Secret Tips on Finding Marketing Insight in Your Data

1. Connect your data to goals and corporate priorities.

First off, make sure you are addressing a business problem, not just claiming credit for a specific activity. The reason for your report should be to solve a problem or provide an opportunity for improvement. Research is resource intensive, so use those resources wisely by focusing on your corporate priorities, not just your activities.

2. You’ll always get more “gold stars” with marketing insight than you will with simple numbers.

Next, remember that, no matter how good your numbers are, what people want to hear—what they need to hear—are insights and recommendations. They want the story the data tells. So, if you have limited time, focus on the insights, not the numbers.

3. Turn things sideways.

Often, the key to finding marketing insight is to look at the data from a new angle. Literally.

One of my most memorable moments of measurement was standing in a conference room of a very large and well-known consumer products company that is renowned for its marketing prowess. The room was packed with marketing analysts and research pros. We were looking at results of a worldwide customer satisfaction study that included thousands of data points. The results had been summarized in a table organized by geographic region. It looked something like this:

The discussion among the experts in the room was focused on the totals along the bottom of the table, especially what was going wrong in Ireland and Spain. They came to a quick consensus that the company’s declining sales were due to the bad economic climate in both countries.

I saw something quite different. Instead of looking at the totals of the columns, I looked at the figures across the table. By doing so it became immediately obvious that their biggest problem was that they were perceived as a company that was hard to do business with, everywhere in the world! I’ve added color coding to illustrate the point:

Both insights were valid interpretations of the data. But only mine was actually useful, since there was little that the marketing team could do about the state of the economy in those two countries. They could, however, do something about the overall perception of the company as hard to do business with.

They applauded my genius and wisdom (naturally 😉 ). But, in fact, it was mostly that they were looking at the vertical columns, and I was looking at horizontal rows. This might seem like an incredibly simple difference, but too frequently flipping things on their sides is all it takes to reveal important insights.

4. Focus on the right data points.

Most reports these days are very long on data and short on marketing insight. Part of the problem is just the vast number of data points offered up by most measurement platforms. Unfortunately, they are frequently the wrong data points. The right data is often buried in a haystack of distracting numbers.

When you feel overwhelmed by too much data, the solution is often to remind yourself of just why you’ve collected the data in the first place. Research frequently starts with a specific question or problem that needs to be solved. Your insight(s) from the data from that research should show how that problem can be solved. So focus on your original question, and set aside data that doesn’t bear on it.

5. Cure Your Cross-tab Deficit Disorder.

All the data in the world can’t deliver marketing insight if it isn’t segmented correctly. Sometimes the problem is that the client doesn’t know which cross-tabs he/she needs, or what to ask for. Other times the provider of the data is trying too hard to prove or disprove their own beliefs. The solution is to cure your Cross-tab Deficit Disorder: use cross-tabs that will reveal answers.

Make sure that you get cross-tabs not just by gender and age, which are fairly standard, but also by length of time with the company, and/or extent of familiarity with the brand. Also look at what media sources/type they read, and anything that you think might have influenced a key data point.

Case in point: We were doing a study of attitudes and awareness among the populace in upstate New York. In general, the news was pretty good. So good that it was hard to figure out what recommendations I could possibly make to improve the client’s program.

But one of my standard questions is always, “Do women feel differently than men?” A quick look at our cross-tabs on one particular question turned up the startling fact that women were twice as negative about my client’s initiatives as were men. Turns out women hated the initiatives, but also that the client’s messages hadn’t reached women yet. My recommendation was to ramp up communications to younger women, who were most likely to get their news from media.

I used a similar exercise to find insight in an employee engagement study. At first all the numbers looked pretty similar, until I analyzed them by length of time with the company. Those that had been there the longest were the least on board with the company’s priorities. The newbies knew all the messages and priorities by heart, because they’d just gone through orientation. If the company hadn’t identified the problem, there was a good chance they might have lost a lot of disgruntled long-time employees, and all that institutional knowledge would walk out the door.  

6. Develop a standard set of questions that you always ask. 

Here are the standard questions I always ask when looking at a data set. You’d do well to ask them, too:

  • Do women feel differently than men?
  • Did seeing different types of media influence opinions differently?
  • Did older people see things differently than younger?
  • Does length of time with the company make a difference in the responses?
  • Does physical location make a difference in the responses?
  • Does familiarity with the brand impact attitude?
  • Do younger people feel differently because they have different experiences with the brand or because they get their news from different sources?
  • Is there any seasonality to the data?
  • Does the form/type of communication make a difference?
  • Is there any difference in response by level of education, ethnicity, or gender?

For media analysis I always ask these questions:

  • What topics or subjects caused were associated with the worst coverage?
  • Who was quoted in the worst coverage?
  • When and why did coverage go up substantially?
  • What was the competition doing at the same time?
  • Is there any correlation between Google Trend data and media coverage?
  • Is there any correlation between website traffic and media coverage?
  • What messages failed to break through?
  • Who was the most effective spokesperson?
  • What, if anything, made the client do better or worse than the competition?
  • How visible was the CEO compared to their peers in the industry?

Not every question is appropriate for every data set, but by starting with good questions, you’ll find insight much faster.

Those are my tips, now go out and find your own marketing insight! ∞

Thanks for the photo up top to Pexels from Pixabay

4/5 (1)

Please rate this

Shopping Cart
Scroll to Top