Q&A: Insights Leaders at Suzy Talk Hybrid & Iterative Market Research
Q&A: Insights Leaders at Suzy Talk Hybrid & Iterative Market Research
• • • • • •
Suzy’s Chief Customer Officer Katie Gross sits down with members of Suzy’s Center of Excellence team, including Senior VP of Market Research William Cimarosa, Director of Market Research Laima Widmer, and Director of Market Research Mary Baker, for a virtual chat.
This interview has been edited and condensed. Go here to watch the full interview.
Katie Gross
What does a typical project timeline look like on the Suzy platform and how might that compare to traditional research methodologies?
Laima Widmer
If you think about a normal traditional approach, you're recruiting your quant, you're executing your quant, and then you have to spend time separately recruiting your qual. Or if you're doing an in-person central location testing, you're looking at a 2-3 week timeline easily.
For me, the benefit of Suzy is, as a client, you know when whatever it is you're wanting to explore is going to be ready, whether it's a questionnaire or stimuli assets. So you can recruit a quant audience, save that audience, and you're ready to hit launch as soon as a questionnaire is approved.
You're saving a lot of time upfront, and you can field your project, sometimes in a couple of hours, depending on the population. If it's a general high incidence population, we can field that in a few hours. From there, pull those people out of the quant survey and then have conversations with them the next day.
The access to these people is almost immediate. The variables that may drive shorter or longer timelines are really the sizes of your audience that you're targeting and how niche it is. For me, as someone who's been executing these for years, timelines have always been a big barrier to implementation, and Suzy does truly remove that barrier.
Katie
It's an integrated, seamless transition from the quant into qual.
Laima
I think this is a really important piece too, because I've tinkered a lot with recruiting quant panelists into qualitative, and a lot of us who've been in the industry for a while know that there's a marked difference between the quality of quant panelists and the quality of people that are recruited qualitatively.
What I've been really amazed about with Suzy is the quality of these people when they're in an IDI, or in-depth interview, in the Suzy Live environment. These are good respondents because ultimately you want your qual to be set up around people who are articulate, who can share, think, and communicate well. We've had tremendous success, not only in getting these people into Suzy Live quickly, but the quality is something that is really important to consider.
William Cimarosa
One thing that also helps with the quality is the way you could use the open ends. In the T-shirt company example that we gave, we asked for specifics about what they liked and didn't like throughout that.
When you see how articulate an open-end is, it's telling of how engaged and how good of a respondent they'll be. It is almost like a screening mechanism. When we have such high yields, it allows you to be selective with those because we all know the risks of recruiting qual. You always have a couple of respondents that don't work out. Because we have such a high yield and you're able to use those open ends to explore what it is, the topic areas you want to go into, it can help ensure that you get that high-quality response.
Katie
It's an incredibly engaged panel that we have for sure. How many interviews would you recommend per project and what would that length of the interview be that you would recommend?
Mary Baker
We recommend at least 10-12 per segment. If your sample is homogeneous, then we would stay around there. But if you want to look at individual segments within that, then we would recommend that many per segment to really get a good, broad understanding of opinions and perceptions among those groups.
We typically do 30-minute interviews, but it really depends on the objectives and how many topics you want to cover or how much stim there is to look at. So 30-60 minutes is probably ideal. But all of that is dependent upon what you intend to test, and what your objectives are.
Katie
And for moderation, it can either be you or your own approved moderator. At Suzy, we also have a list of approved moderators who have specialization in certain areas. We also have a back room so that you and your colleagues can take part in any virtual, two-way mirror and converse with them while they are live moderating on Suzy Lives. There’s a lot of fantastic functionality for collaboration.
One question for you, Will: What would you suggest is the right order to conduct hybrid quant and qual?
William
It does depend on the use case and what you need to explore. If you're looking to understand drivers and attitudes that you then need to go measure, it makes sense to open up with qualitative research, find out what is important to consumers and what drives behavior, and then go and measure it. How important are those things that you've discovered in qual? Once you know and you've measured what drives usage, we then recommend that you start with quant.
So in the T-shirt example, we didn't have any specific diagnostics, but we could have put them in from a foundational study. Diagnostics would have been additional questions like, "Seems like it has a good fit. Seems like a premium brand." Typical things that you'd get out of foundational learning, where you already have measured what those drivers are.
You then test your ideas to see how they're performing. You look and see which ones are rising to the top and which ones are going to the bottom, from a statistical standpoint. Because we have the open ends tied to those answers. And you can see the performance differences, you can then dive into the comments. Dive in if one is underperforming or overperforming, dive in there and see what those responses were. Save that audience and invite them into a conversation, so you can learn the specifics of what's driving that.
Using that T-shirt example, once we understood that T-shirt design and the fit was more important than the animal itself, we were then able to iterate very quickly and go back and test it again.
One thing that I want to point out is we kept a benchmark, so we knew what was performing the best. We also kept the top performers from the previous round to make sure that we'd actually made a difference. From there, we got success. We statistically outperformed both the control and the top-performing shirt from the first design, but we could have gone in and dug deeper. We could have found out from the likers and the dislikers what it was that was driving their opinions and kept iterating and refining.
If you need to explore and then go measure, start with qual and get a measurement of what's driving behavior. Once you understand what that behavior is, use that to drive your basic questionnaire guide for your qualitative feedback loop.
Katie
Yeah, that's a really good point. And what happens if the narrative from both the qual and the quant come out slightly differently? Sometimes the qual story does not always match the quant. How do you marry the two? And I guess in part, that is iteration.
William
There's nothing subjective about this. This isn't a randomized selection of audiences. We're observing the actual answers. So, if there are two different storylines, we're going to go in and identify what's driving those answers. If something is underperforming, we're going to actually select the respondents that indicated that this wasn't performing up to their standards and learn why.
You should guide qual around what the numbers tell you. We don't need to marry them if we can actually use the numbers to then go and find those specific respondents.
For example, ‘You rated this low or poorly on a seven-point scale, tell me a little bit why.’ There's no need to line things up then. I've actually observed this respondent make that answer. I can now go and find out what it was that was driving it to iterate. That's the beauty of the open-end. You can look at and see why they answered the question the way they did, so you can go and learn more. You're marrying it by actually asking the questions about the answers you observed them actually submitting on the quant panel.
Katie
A couple more tactical questions: How does Suzy work with clients on creating discussion guides?
Mary
Typically, if there's a hired moderator, they will typically take the reins on the discussion guide, but we also provide content recommendations and flow recommendations if clients choose to develop a guide on their own. Another important component is that you're really looking to uncover the results from quantitative studies.
To get under that hood, you'll include some of the same topics to explore with the respondents. Some of the content actually comes from the previous quantitative research, but it really depends on the objectives and what you're showing, and absolutely we can provide guidance on that.
Katie
That is great. Laima, one question I had for you: You’ve worked on the client-side, on the market research agency side, you've seen your fair share of platforms. In your expert opinion, why do you think Suzy has a specific benefit when it comes to these quant-qual hybrid methodologies?
Laima
I can't say enough times that the core benefit for me, putting my client hat on, is access and speed. We are all moving at a thousand miles a minute, and we, as marketers, as clients, as brand managers, don't have the luxury of time to build into our development cycles long research projects. Those days are gone.
Part of what Suzy does is give clients the ability to access consumer feedback really quickly, so they can always layer in the voice of the customer. That voice of the customer is really feeding a lot of the things that you touch and do really quickly, very simply. And I can't stress enough the power of the combination of metrics of hardcore quant numbers with the qualitative depth of what those numbers need.
That means you can show your leadership teams a bar chart with numbers, but then you could hit a highlight reel where people are talking about what those numbers mean. That just brings a lot of things to life in a very compelling way to help move decision-making, help move prioritization, help move to action a lot faster. For me, it is really just the integration, the speed, the time, and the economy. Suzy offers an incredibly cost-efficient way of doing this.
William
Yeah, I think that's a key point. Having personally been on both sides as a client and as a vendor, we shouldn't undersell the importance of the model. You're not having to screen each time that you run a survey. We build an owned panel around your screener, which gives you that quick access.
You're not paying per complete. You're paying by the question. And while that seems like a subtle nuance, it really does allow you to be more agile in terms of speed. When you're not having to recruit and incentivize each time that you want to do a quantitative study, it really allows you to have a quicker turnaround. We were able to run that study where we had one day of quant, one day of qual, refinement, and then go back into the field. It's very easy to do because we're not screening the respondents each individual time. We're building that panel and we're targeting as we need to and as we see fit.
Katie
Yeah, absolutely. It's a significantly faster way of conducting research.
For the full Q&A, visit Suzy.com.