It’s already a couple months into the New Year, but it’s never too late to discuss emerging or already popular trends in market research.
While you’re probably familiar with the buzz words (big data, user generated content, mobile, etc.), we take a closer look at why these topics are consistently at the top of the industry’s list and how you can unlock their potential.
The problem: market research (MR) firms have the technical know-how to collect and report big data, but lack the creative ability and tools to synthesize it into real-time, actionable results or a strategy.
Oh yes, big data: the amorphous, over-used concept and catch-all phrase that just won’t go away, and for good reason. David Wilson blogged about the importance of mining big data and how it can give you a competitive edge and lead to greater market share.
While it’s easy to sing the praises of big data, it’s critical that we also recognize that there is a huge variety of data generated at an incredibly fast rate everyday (or as GreenBook likes to put it, we’re faced with the 3 V’s: velocity, variety and volume).
A shortage of skilled personnel, lack of inter-departmental collaboration, and inefficient use of big data tools are identified as some of the key obstacles that loom over MR professionals eager to harness the power of big data.
With the odds stacked against MR firms, what’s to be done? Here’s some advice on how MR can use the skills already at its disposal to work with big data:
- Why? Remember, MR is ultimately about figuring out the “why?” behind consumer behavior and trends. While big data and its accompanying tools can answer all the other W’s (who, what, when, where), MR must step up to the plate and create effective, impactful business solutions from the analytical information at its disposal.
- The “human element.” Emotional intelligence is something that big data tools lack and MR excels in. Extracting workable insights from a cross tabulation of thousands of respondents requires a more in-depth look at what’s making masses of people tick in the first place.
- The “kitchen sink” approach. Enter a large load of data (i.e. independent variables) into a predictive model and then see which data elements—these should include behaviors, transactions, attitudes, and demographics—are most predictive of your outcome (i.e. dependent variables, such as satisfaction, net promoter score, or likelihood to join, spend, upgrade, decline, defect, etc.). This process yields a useful predictive model and the identification of the most critical data elements among your infinite ocean of data.
This leads us to our next trend of predictive analytics, which goes hand in hand with big data.
The problem: there is a lot of ‘noise’ to be contended with in data. How can market researchers sift through loads of qualitative, quantitative, and big data in order to find patterns or causality that can solve common MR problems?
We all want less churn and more customer retention. That sentiment is something most market research firms have in common, but the problem rests in sifting through the noise to find the valuable insights that qualitative, quantitative, and big data can provide.
In order to target both at-risk and highly profitable customers, it’s necessary to create and apply predictive models. Here are some basic steps to take when considering a predictive analytics model:
- Cleanse. Consolidate and organize your data so it is informative. This is a critical first step in the modeling process that requires time and careful attention. Otherwise, bad data and excessive noise will only create bad predictions.
- Criteria. Now that your data is cleaned up, it’s time to question it and apply criteria in order to create predictors and show behavioral patterns. Ask questions that will aid in revealing what data points are relevant to understanding the customer or business objectives.
- Compare. There may be several models worth evaluating and comparing—there is no one size fits all. The only way to find the ones that best suit your needs is to test their performance with different types of data.
- Consistency. Finally, be sure to consistently check your model(s) of choice’s level of accuracy, validity, and calculation of error.
The problem: with people constantly attached to their mobile devices, it is critical for market researchers to reach these savvy consumers in the most effective and appropriate way possible.
According to a consumer survey by The NPD Group, 37 percent of PC users are turning to their smart devices to surf the web and access apps like Facebook. The world is showing us its ever-increasing interest in mobile technology, and it’s time for market research to make the most of this shifting behavior. The question is how.
There are many elements to consider when going mobile: privacy issues, consumer patience, and available mobile methodologies (i.e. the diary, online panel, or simple survey). From our own research, we’ve discovered the following trends that work:
- Shorter, simpler, text-based surveys
- Image or video uploads
- Survey applications for download
- Location based surveys
Keep in mind that while going mobile is a tempting offer in the competitive marketplace, don’t do it without justification: mobile research or surveys should support business goals or objectives. To further explore the dos and don’ts of mobile research, check out Rick Reed’s blog post.