Here is the secret recipe for getting insights from your marketing research: Bake them in from the beginning. Bake them into the questions you ask. Bake them into the charts and visualizations you create. Bake them into your report headlines.
If you do these things, by the time you finish drafting a straightforward report of findings, all the insights will be there, jumping out at your audience. And it will make them hungry for more. Our feature article in this quarter’s newsletter, Good Research Has Insights Baked In, shows you how.
Other items of interest in this newsletter include:
- Big Fat Margins of Error
- Trim Your Weights at Five
- Using MTurk for Market Research
- Pollsters Disagree on Best Sampling Methods
- Using SMS Texts for Surveys
- How to Fix the “Up Means Good” Survey Bias
- Election Season’s Dumb Statistics
- Why Your Neuro-Marketing Might Be 70% Wrong
- Blast Away—Surveys Are Not SPAM
- Lessons from NYT on Data Dump Research
- Dump Your Landline Sample
- Tips for Sales Teams: Ask a FavorWe are also delighted to share with you:
… which highlights, among other news, our recent work for TransUnion on a survey about big data and financial analytics, plus our recent “9 Habits” article published in Quirk’s.
As always, feel free to reach out with an inquiry or with questions you may have. We would be pleased to consult with you on your next research effort.
The Versta Team
Good Research Has Insights Baked In
Here’s a sobering question for professionals who work in the “insights” industry: How much of your market research really offers insight?
Not just validating data, or routine measurements, or pro forma tracking. Not just percentages of “how many said what” in a customized study. By true insight we mean:
- a direct answer
- to a deeply relevant question
- that was previously unknown, and
- that has immediate implications for what to do next.
What would you say—does one-tenth of your market research achieve a level of true insight? One-quarter? Half?
Surprised by Insight
This question nags at us all the time, as we work so consciously to elevate the value of research beyond just good data to a level of real insight for clients.
How much of it gets there for us? Not all of it, for sure. But at the risk of sounding like a braggart, I want to share a comment we got from a client last month, not because we are so proud (though we are) but because I was surprised and indeed puzzled by it. And since I was more surprised than proud, the comment provoked much thought and discussion among my team.
I’ve spent some time with the report and am really so VERY pleased with the insights – thank you so much. You’re miles ahead of every other supplier I’ve worked with in terms of reporting and analysis. NEVER have I received a report that was this cohesive and clear and insightful. Honestly – you’re great at what you do. Because I don’t have to spend time fixing headers and formatting charts, I can actually pay attention to content – and so I have just a couple thoughts…
Honestly, I was happy. But puzzled, too. What had we done that was so great? It didn’t seem like much. In fact, I was worried that the report was stripped down and too bare. We had big, full- page charts. There was an ongoing storyline moving from page to page, limited to ten or twenty words per page, at most. The executive summary was one page—six simple bullets.
We baked in the insights from the very start. And then “finding” and presenting those insights flowed easily from there.
But insight? Where was the insight? Not wanting to deflate the delight, I didn’t ask. But after our internal brainstorm as to what exactly we did—how we found and reported just the right information that ultimately constituted deep insight for our client—here is the answer we came up with: We baked in the insights from the very start. And then “finding” and presenting those insights flowed easily from there. So easily, in fact, that it felt like we were doing nothing at all.
Insights Are Baked In
Here is what it means that insights are baked in, and how we achieved it:
1 By asking the right questions. We spent a good week or two up front in multiple conversations with the business units about what they wanted to learn. It wasn’t easy because they were after an abstraction. They wanted to know how customers structured their thinking—that is, what mental models customers were using to guide decisions and measure progress.
The trick was to come up with a straightforward and compelling series of survey questions that would capture the possibilities in simple language consumers could understand. The results we got back—the data itself—were eye-opening and surprising. By asking the right questions, there was no need to find insights, or extract them from the data. There they were, staring us in the face. In short, the insights were baked into the questions.
2 By visualizing the data thoughtfully. That’s a fancy way of saying “by picking the right chart” or choosing to display it in a table instead. It seems that not many suppliers are good at this. Data should never, ever be dumped into a standard set of charts. The insights we conveyed resulted from a careful, page-by-page, data-point-by-data-point consideration of the message we wanted to convey on every page, the contrast we wanted to show, and the best chart that would do that.
Providing a visual display of data requires thoughtful deliberation of all options, even options that seem to violate our usual best practices, like avoiding pie charts. That deliberation has to happen as we are looking at the results and after we see how the data are falling so that the chart (or table) we pick is telling the right story unambiguously and with power. In short, the insights were baked into the charts.
3 By writing storylines. The worst kind of report offers up a chart, and then puts charts into words with a bullet point. “Men are more likely to give higher satisfaction scores than women.” If that is the point to be made, and if the chart has been built correctly, it will be obvious. So why say it?
In contrast, the best kind of report sums up the key take-away, which the chart merely displays as evidence. It does not report data, it provides a narrative that interprets data. It tells the reader what the findings mean, and then offers up data points or visualizations as proof that the interpretation is correct. “Men like our products more than women do,” for example, is the story in the data, and that is what needs to be said.
Indeed, one of things we did in our report—almost without thinking, as it has become baked into our approach—was to write headlines that tell a story. We write those headlines without numbers and without reference to data at all (see our video on Making Research Matter), and then adding numbers or data back in only if they help. In short, the insights were baked into the headlines.
Avoiding the Data Dump
Ultimately, the report we delivered for this client was straightforward and relatively easy to produce. We didn’t squeeze our brains or use special analytics tools to extract difficult insights. We just laid it all out, and with the insights baked into each thoughtful step, it achieved exactly what our client needed.
In contrast, consider this quote from a client-side researcher highlighted in the just-released 2016 Corporate Researcher Report from Quirk’s:
[Our pain point is] turning vendor reports into something truly insightful. Vendors are still too much into the mentality of delivering the data and less about what it means and the implications for the business.
Is it possible, as a client, to extract, edit, and revise data-dump reports into something truly insightful? Sure, but it’s tough. There is little to leverage except lots of overwhelming data. It will need a painstaking overhaul of how the questions are framed, how the tables and charts are constructed, and how the slide titles and the narrative are laid out. Chances are you will be better off starting over from scratch.
The secret recipe for good research is to incorporate insights into every step of the research and reporting process. Ask the right questions. Display data thoughtfully. Write storylines that highlight what the data mean. If you do these things, by the time the report is done, insights will be there, baked right in. The final product, full of true insights, will make you and your business partners hungry for more.
Stories from the Versta Blog
Here are several recent posts from the Versta Research Blog. Click on any headline to read more.
Margin of error statements are usually misleading because they account for only one type of error. Here are five other critical sources of error they ignore.
One tracking poll this season did not pay close attention to its statistical weighting, which skewed results via an over-weighted, outlying Trump supporter.
More and more academic research is crowd-sourcing sample from places like MTurk. What gives, and should you consider it for your own market research?
This recent exchange among the nation’s top survey experts shows the lingering debate about accepting online non-probability samples. But we’ve come a long way!
SMS text messages are new territory for market research, but so far the method shows promise. Here is recent data on how they boost response rates and data quality.
There is an important bias in our culture such that survey items higher on a page are evaluated more positively than others. Here’s the scoop and how to fix it.
An odd joy of presidential elections is citing all the ridiculous statistics journalists offer to predict who will win. Here are some doozies.
There is a software flaw in most fMRI analyses that suggests significant brain activity where there is none. Up to 70% of significant results are likely to be wrong.
Survey invitations, even to total strangers, are not considered spam. But there are good reasons why you should follow anti-spam protocols anyway.
NYT journalists are adopting a failed strategy of the market research industry: dumping data into articles without writing a story. Here’s how to do better.
New research shows that cell phones are no longer a threat to phone surveys, but a superior alternative. You are better offer eliminating landline calls entirely.
New research shows that asking a favor of a buyer when negotiating a deal enhances feelings of reciprocity and trust, making a successful sale more likely.
Versta Research in the News
Versta Research documented the struggle that lenders face in keeping up with their data. Findings are highlighted in an interactive infographic, with an e-book available for download.
MORE VERSTA NEWSLETTERS