"Prediction Markets was one of the more interesting topics I heard discussed and presented, because it has the potential to really change the way we think about traditional approaches to quantitative market research."
I just attended The Quirk’s Event — the always insightful annual market research conference—held in Irvine, California. I met a bunch of interesting people, learned a lot, and thought I’d share my observations about some of the key market research trends discussed at the conference, including:
- Prediction markets
- Agile research
- Online communities
- Consumer behavior
- New tools
- Image sorting for quant surveys
- DIY choice modeling
- Gamifying surveys
This was one of the more interesting topics I heard discussed and presented, because it has the potential to really change the way we think about traditional approaches to quantitative market research. Based on the belief that a group of people, or a “crowd” as they call it, can accurately predict other people’s reaction and behavior to a new product, claim, ad, etc, This methodology suggests that you don’t have to talk to your target market to get accurate research results. Instead of recruiting a target market — which can lead to higher costs and extended recruiting time, depending on the criteria — researchers can recruit a more general audience. On the surface, it seems like a cheaper, faster option than the traditional approach.
How does it work? The framing of the questions is changed from first person to third person — instead of “would you buy,” it’s “would other people (who buy these kinds of products) buy” this product. They also use an interesting survey approach where people place bets on the ideas or concepts they think will be most appealing, and those that do the best job predicting the winner get a bonus monetary incentive.
My take: Those presenting this technique all claimed validation studies exist, but none of them actually shared any details. So I’m still a bit skeptical. In virtually all quantitative research surveys, it’s so common to see differences within subgroups such as age or gender. If you have a non-representative mix of these groups, you’re going to get a different and inaccurate result compared to the result from a representative sample. And beyond demographic mix, these methods recruit people that don’t even necessarily buy the products being tested. I’m not buying into this approach just yet. But I am intrigued and plan to explore it further.
Making market research faster, cheaper, and more efficient is one of the major trends in our industry. The “Predictive Markets” approach I described above was adapted for market research for just these reasons. Other approaches are also popping up as well—all are being described as “agile” market research. Let’s face it: Traditional market research can take a lot of time to set up, execute, and develop the reports.
How does it work? The agile market research techniques I’ve encountered generally aren’t creating new methodologies; they’re simply taking the tried-and-true approaches and making them quicker. They could be short surveys, qualitative message boards, or online focus groups, etc. One thing common thread for “agile research” is they tend to be shorter and more standardized than traditional custom research. What you lose in flexibility, you gain in speed and cost efficiency. The more custom the study is, the more time spent discussing and debating which questions to include, how to word them, etc. among all the stakeholders; and, on the back end, the vendor has to take more time to develop the reports. All this adds time and money.
My take: The agile approach is not meant to replace all research studies, but it’s great for those cases when you need results fast and don’t have a big budget. A good example is concept testing, and I’m in the process of testing an automated platform on a new project right now.
This is an approach that’s gaining momentum. With an online community, a company recruits its own consumer panel for market research.
How does it work? There’s a set up and ongoing maintenance cost, but, once you have it set up, each study you do is generally cheaper and faster compared to independent custom studies. Rather than the traditional approach where you spend your whole budget on a few big studies spread out across the year, you can conduct research much faster and more frequently with an online community.
The result: You’ll get insights when you need them and leverage a platform that can adapt to your research needs as they vary throughout the year. Some of the online platforms being developed are also offering a wide range of tools beyond just surveys, including message boards, online focus groups, diaries, etc.
My take: There’s a downside. Those willing to join the panel are going to be inherently biased toward your brand or company. So you’re not going to want to conduct any kind of research where you need a truly representative sample, such as branding research. In general, there’s going to be acquiescence bias for anything you’re testing — new products especially. The other potential problem I’ve heard from two clients is sometimes companies that develop the panel management platforms aren’t necessarily the best researchers. However, despite the caveats, I think the benefits are strong enough that as long as you know what you’re doing (or have good guidance) and understand the limitations, they can be a great approach. I’m actually in the process of developing a partnership with an online community company to offer this to our clients.
Learning from Behavioral Economics and Cognitive Psychology
This topic alone deserves its own blog so I’m going to just give you the overview for now.
The field of market research is starting to learn from cognitive psychologists to better understand how the brain works, particularly on the topic of how we make decisions and the biases that can affect them. Smart market researchers will learn and understand these principles which will have an impact on what types of research they do and how they execute it.
I saw several new tools that I thought had some promise. Here are a few worth mentioning:
- Image sorting for quant surveys
Image association has been a common technique in qualitative research for years, but it has been challenging to adopt for quantitative research. A company call Protobrand has developed a platform that can be integrated into quantitative surveys to do just that. What is also very cool about their platform is how it summarizes and visualizes the results in an interactive tool. Think of a word cloud but for images.
- DIY choice modelling
Taking DIY research to the next level, a company called “aytm” has developed a platform for DIY choice modelling. You can conduct consumer choice models such as max-diff, choice-based-conjoint, and other techniques yourself. A little scary if you ask me, but in the right hands could be really useful.
- Gamifying surveys
Work is being done in the industry to make surveys more interesting and fun to take. A company called “Datagame” had an interesting platform designed to do just that. Their tools are primarily for preference type methodologies like max-diff, and were very slick. My personal opinion is the types of questions they were gamifying were already pretty easy. I want to see a company simplify the hard stuff – like rating several brands on a battery of attributes. However, I did like their tool and the direction the industry is going.
Although this review of Quirk’s was not meant to be exhaustive, hopefully it gives you a sense for some of the hot topics in the industry.
The evolution of market research is exciting. While some new techniques and methodologies may just be flashes in the pan, there are definitely innovative approaches surfacing that provide a real benefit, and they help researchers in our continuing quest to make market research more predictive and actionable.