Ep. 572 – A Guide to Tracking Research with Emmanuel Probst, the Global Lead of Brand Thought-Leadership & Senior Vice President of Brand Health Tracking at IPSOS

My guest today is Emmanuel Probst, the Global Lead of Brand Thought-Leadership & Senior Vice President of Brand Health Tracking at IPSOS, Author, and Professor at UCLA.

Founded in 1975 and headquartered in Paris France, Ipsos is among the largest global market research and consulting firms. Ipsos has over 18,000 employees and serves more than 5,000 brands. 

Find Emmanuel Online:

Find Jamin Online:

Find Us Online: 


This Episode is Sponsored by:

The Michigan State University’s Master of Science in Marketing Research Program delivers the #1 ranked insights and analytics graduate degree in three formats: 

  • Full-time on campus 
  • Full-time online 
  • Part-time online

NEW FOR 2022: 

If you can’t commit to their full degree program, simply begin with one of their 3-course certificates: Insights Design or Insights Analysis. 

In addition to the certification, all the courses you complete will build toward your graduation.

If you are looking to achieve your full potential, check out MSMU’s programs at: broad.msu.edu/marketing.

HubUX is a research operation platform for private panel management, qualitative automation including video audition questions, and surveys. 

For a limited time, user seats are free. If you’d like to learn more or create your own account, visit hubux.com


Jamin Brazil: Hey, everybody. I’m Jamin. You’re listening to the Happy Market Research Podcast. My guest today is Emmanuel Probst, the Global Leader of Brand Thought-Leadership and Senior Vice President of brand health tracking at Ipsos. He’s also an author and professor at UCLA. Founded in 1975 and headquartered in Paris, France, Ipsos is among the largest global market research and consulting firms. Ipsos has over 18,000 employees and services more than 5,000 brands. Emmanuel, thank you very much for joining me on the Happy Market Research Podcast.


Emmanuel Probst: Jamin, thank you for having me on your show. Really appreciate it. Always great connecting with you and your listeners.


Jamin Brazil: The Michigan State University’s Master of Science in marketing research program delivers the number-one ranked insights and analytics degree in three formats. Full-time on campus, full-time online, and part-time online. New for 2022. If you can’t commit to their full degree program, simply begin with one of their three core certifications, insights, design, or insights analysis. In addition to the certification, all the courses you complete will build towards your graduation. If you’re looking to achieve your full potential, check out MSN news program at broad. msu. edu/marketing. Again, broad. msu. edu/marketing. HubUX is a research operations platform for private panel management, qualitative automation, including video audition questions and surveys. For a limited time, user seats are free. If you’d like to learn more or create your own account, visit hubux.com. This interview is part of a series on market research basics, and our conversation is going to be focused really on your wheelhouse, which is tracking research. You’ve done this for decades. It’s amazing.


Emmanuel Probst: For a number of years, indeed.


Jamin Brazil: Number of years. And for some of the world’s – actually, the world’s largest brands across sectors, which is really interesting. So my first question, and let’s provide some context for the listeners. What business questions does tracking research address?


Emmanuel Probst: So it’s important to keep in mind that tracking research is at the end or towards the end of a research continuum. And what I mean by this is, people first, brands first, need to understand their market positioning, and need to establish their brand identity, and need to understand their audience. What are the demographics and psychographics and attitudes and behavior of people they’re going to market to? And ideally, understand some of the competitive environment. And that’s what I mean by tracking being towards the end of a research continuum because, first, it’s important that brands establish who they are, what they stand for, who do they want to market to, what is their competitive environment. Now, standout business questions that brand tracking address. The main one is how is my brand doing versus its direct competitors? That’s really what 99.9% of our clients want to measure. I want to establish how well my brand is doing versus my four, five direct competitors. And importantly, how am I trending over time? It sounds obvious, but tracking is about tracking. It’s not a point in time study. So what’s colorful, what’s insightful, what’s interesting, what is valuable, is to understand how the brand evolves over time. If it goes up if it goes down, and on what metrics, what territory the brand is trying to own. So, quick pause here, Jamin. But that’s really the main business question that everyone wants to answer through tracking.


Jamin Brazil: Clarification question for you. I really like how you framed it. It’s at the end of the research continuum, but it also spawns research, doesn’t it?


Emmanuel Probst: It does indeed. And what’s compelling, actually, that’s a good segue into the second research question that tracking answers is we want to uncover and understand market trends. We want to understand market shifts and market dynamics. And to your point, Jamin, in term if the tracking instrument is going to prompt some new research questions and some new research projects.


Jamin Brazil: Perfect. Are there any other business questions that we should highlight here?


Emmanuel Probst: I think those are really the main ones. Now we have a range of key performance indicators. Some of them are very standout, and I’m sure we’ll talk about this later on in this interview. Some of them are more custom depending on clients. But really, let’s simplify and clarify for our listeners today, the point of brand health tracking is to assess, measure the performance of your brand over time versus its competitive set.


Jamin Brazil: Perfect. So let’s move into the next framework or question, which is really what common analytic techniques are used?


Emmanuel Probst: So I’d say two things that are not disruptive. And I mean it this way, number one, we very often rely on the funnel. And as a way to look at how do people go through this consumer journey, if you will, between establishing a need or want and buying the product, and buying the product again, and recommending the product to friends and family. Now, I know a lot of people in our industry criticize the funnel, and rightly so for all its shortcomings. The marketing funnel is not as linear as it appears in textbooks, for example. However, I find it to be a very useful framework because everyone can relate in our industry. You might be a CMO with 20 years’ experience at a large CPG organization, or you might be an intern, or a new hire six months out of school, and you can relate to the funnel. It’s also a great way to communicate with people in our industry to understand how they contribute, where they contribute. So what I mean by this is when you meet someone at a conference, for example, you don’t know what this person is doing. You don’t understand the product and the agency very well. You can always ask, where are you at in the funnel? How do you contribute in this funnel? So long story short, Jamin, the funnel is really a very useful framework. Again, I acknowledge the shortcomings. But it’s a useful framework because it’s a framework we can all relate to. And in terms of the analytical techniques, I think what’s most important is to be consistent with the sampling plan and with the methodology. It’s very important to be consistent over time, remember that we track, so we need different waves month by month, quarter over quarter to be similar, so that we can accurately measure the performance of a brand over time. In short, the analytical techniques. Now, of course, we have advanced analytics, but the core concepts are simple. And I’m saying so in a good way. They need to be simple, relatable, and consistent.


Jamin Brazil: I like how you succinctly put that. Crosstabs, of course, is used consistently when analyzing changes among customer personas or segments over time. Do you see NPS as an analytic technique, or is that more broadly a different type of research?


Emmanuel Probst: So NPS is a very interesting one, and it goes back to my comment on the funnel. Is the funnel perfect? Absolutely not. The funnel is also incomplete, arguably does not thoroughly describe today’s consumer’s journey, yet everyone can relate. NPS is exactly the same thing. Meaning from a methodological called standpoint, NPS, Net Promoter Score is pretty shallow. It’s one number. Detractors and promoters of your brand, and how does that look? But here again, why do people love NPS, and why is NPS relevant? Well, just like what we discussed about the funnel, if you’re a CMO with 20 years’ experience at P&G, you know about NPS. And if you’re a student at UCLA, and you’re 22, and you’ve done one internship in marketing, you also know about NPS. So it is a metric that everyone can relate to. When I worked in customer experience management years ago, I was in particular, I was on Victoria’s Secret, and we would use NPS all the time simply because it’s one thing to discuss brand performance in a boardroom in Columbus, Ohio. It’s more challenging to have this conversation with a 17-years-old that works Saturday only at store level in wherever at a shopping mall. The point is you can teach NPS to anyone. I can teach NPS to a high school student in 10 minutes on a shop floor. So NPS is useful because it is a metric people can relate to no matter what level they are at. However, NPS is shallow because it is just one number. It has no depth, no colors, and it shows no opportunities for the brand and its managers.


Jamin Brazil: Let’s move into the operational considerations. So there’s been a rise in the last three years, especially through COVID, with research ops or research operations. This is specifically around how we actually gather the data. You’ve addressed this already, but I want to call it out as a separate question. Are there any common research operations considerations that need to be taken into account when you’re conducting tracking research?


Emmanuel Probst: Yeah. So I feel that tracking research is one of the riskiest things you can do in research. Let me explain. You need consistency wave over wave. When you deliver an ad hoc project? Well, if things go wrong, you can refill then you can do some weighting and all that. Now, of course, you can do some of this in tracking. But again, you have to stay consistent wave over wave, not only in terms of demographics, and psychographics. But the blend in terms of sample vendors, so that’s why the SKEW you have remains consistent. You need to be consistent in the ways you analyze the data. That’s all to say that research operations in tracking is a beast. I’m not going to wordsmith it. It’s an important, it’s a crucial backbone. It is very high touch, if you will. It needs a lot of attention. So I don’t know if I will describe common research operations. I will set out some best practices. Again, I would say that the sheer size of a tracking program and the need for consistency, it makes it difficult to do it yourself. That’s the way I would put it.


Jamin Brazil: And there’s a lot of risk in doing it yourself. I had one of my largest clients, this goes back in 2005, they launched a tracking study, as soon as they had filed their IPO. It was specifically NPS. And for those that don’t know, Net Promoter Score is an 11 point scale from zero to ten, and it’s shown from left to right. This is pre-smartphone. Well, over the following five years in that tracking study, their NPS continued to drop materially. When you look at their actual survey, this was the point of- We got actually fired from this account. When you looked at the survey on a smartphone, it turns out that you actually had to scroll to the right in order to get into the top three box, which forced respondents basically into that bottom segment. And so they were disheartened because the growth of the iPhone was directly correlated with their decrease in NPS, it was just a technology issue, as opposed to it actually reflecting the score or the populations view of the platform.


Emmanuel Probst: See, and that’s where it gets tricky. And that’s what I mean by high risk versus aspects of harvest surveys, worded harvest surveys displayed on a screen. That’s where there is no wiggle room, if you will. And your example is very valid, albeit painful, of how you can SKEW the results, and frankly, invalidate the results in that case because of research operations that were lacking. And when I say this, I don’t mean to be overly critical because it’s very hard to do something simple. It’s very hard to create surveys that fit on a screen that are short enough to get people’s attention, to address professional respondents and show data quality. All those things, high risk in quantitative research in general, particularly in tracking. In my experience, brand health tracking that and ad impact measurement. Those are the two riskiest research projects one can undertake.


Jamin Brazil: Do you have a favorite case study?


Emmanuel Probst: Of course, what’s frustrating in our industry and like advertising agencies is we cannot disclose, and we cannot talk at length about the measurement we do for our clients for this reason. And so it’s sort of a value of that measurement is very strategic. With that said, I have two things that come to mind. One is for technology brand. So what I mean by technology brand, think first of Amazon, Apple, Facebook, Google, Meta, Twitter, TikTok. Those are technology brands in my book. And what we saw in that measurement is we saw Tesla bubbling up as a technology brand. And I think that’s interesting from a methodological standpoint in market research, also to show what the tool can do, meaning the brand health tracking program, and finally, the world we live in. So the tool enabled to surface that Tesla was considered as a technology brand. And the larger implication is to discover that consumers think of Tesla. Just like them, I think of Meta, and they might think of Google, if you will. Meaning what comes to mind, of course, Tesla is a car but people are so impressed with the technology. And people are also aware of what we call a data exchange, meaning Tesla collects a lot of data while you’re driving. What does Tesla do with this data? Does Tesla monetize the data or only uses the data to make the product better? But long story short, I thought that was very compelling to see Tesla bubbling up as a technology brand.


Jamin Brazil: That’s very interesting and disruptive to your quadrant map slightly.


Emmanuel Probst: Indeed. So that leads to the question, which applies in tech, but not only, what is your competitive set these days? What market does Tesla compete in? So Tesla makes cars. We pretty much all know this by now. However, Tesla also accumulates data. Is there value in the GPS signal, for example, that Tesla collects? Is there value in collecting data about people’s commute to work, or people’s activities, if you will, when they take their kids to soccer, and so on and so forth? And so, I’m going on a tangent here, but the point is to uncover that wealth of data, and what does it lead to for Tesla as a brand? How is Tesla going to position its brand? Is there an opportunity to develop a stronger ecosystem, for example, that’s with your solar panels and your power walls at home that might be linked to your car, and maybe your home thermostats? So this opens a range of opportunities for a brand in isolation, but also for the world we live in.


Jamin Brazil: And as we’ve seen the move to supporting Spotify and Netflix directly in car just continues to blur the lines of where they actually fit from a sector perspective.


Emmanuel Probst: Precisely. And that’s a shortcoming of traditional market, not a shortcoming, but that’s something you could do in traditional market research, that is being challenged with technology brands or brands that involve technology. In other words, in the old world, you could absolutely look at, let’s say, Aquafresh versus Colgate versus Sensodyne. That’s a pretty clear competitive set because they all make toothpaste. And granted, one might whiten your teeth, and one might be better for your gums, and so on and so forth, or at the end of the day, the products are pretty similar. And if anything, they sit on the same shelf. Put it this way, if anything, they sit on the same shelves. But here with Tesla, what are we looking at? Well, it’s a driving machine, it takes you from point A to point B that’s one thing, but does it have the ability to optimize your commute? Does it have the ability to recommend places you can stop at to eat or to recharge your batteries? And as you said, co-branding opportunities? How about a co-branding opportunity with Spotify? How about a co-branding opportunity with Netflix and so on and so forth? So as you said, the lines are blurred. Those are things that tracking tools help and cover and address.


Jamin Brazil: How much should companies expect to spend if they’re going to launch and maintain a tracking project?


Emmanuel Probst: The true answer is it depends. So short answer, tracking goes from 150,000 to 20 million dollars. I’m realizing that giving such a wide range is not helpful to our listeners today. So let me tell you what drives this. Well, first, how often are you going to track? That’s because sometimes I don’t recommend tracking monthly, for example. The needle is not going to move that often. Let’s say you want to try barbecue, for example. If you want to try Green Egg versus Weber grill versus Traeger? Well, it’s unlikely that people’s perception of barbecue brand is going to evolve from one month to the next. What’s most valuable, however, is to do some deep dives at points in time around fourth of July, Memorial Day, when people barbecue a lot. So number one, the cadence. Number two, of course, how much sample do you need? How much sample do you want? Depends if you have one market, one country. It depends also on the penetration of a product. It’s obviously much easier to track toothpaste than barbecue, than it is to track people interested only in self-driving cars, or in autopilots like you have in a Tesla. Now we can do this, but it gets a lot more expensive. That’s because of the sample. And the last main driver of cost, not many of us, but I’m focusing on what’s most important here. The last one is how complex or how straightforward the analytics will be. In other words, you can do cross-tabulations, as you suggested, and you can say, “Hey, there we go, Mr. Client. Here’s your brand versus your five key competitors.” Or you can deliver some insights. So what you add is, Mr. Client, this is with so what, now what, what does this mean to your brand, what you should do next. And you can also add some advanced analytics, meaning you want to understand the drivers. How do people came to the conclusion that Tesla was a technology brand, and here we want to identify the attributes that lead people to believe that Tesla is a technology brand. So that’s an additional layer of complexity, of course, which commands additional funding. The last one is, how many sources of data? What are the different sources of data you want to include? That is a lot of tracking programs rely on surveys. And that’s fine. Now you can add to this new signal, and legislative publications, and social media listening, and CRM data. And all those data sets will make your research instruments stronger, potentially more insightful, yet the data onboarding is expensive. So that’s the answer to what drives the cost of a tracker. And again, that range of $100,000 to 20 million is a very wide range, yet very accurate.


Jamin Brazil: Probably a very similar answer to this question. How long does it take to actually execute tracking research?


Emmanuel Probst: Well, the key benefit of tracking is to track and I keep making things so basic like this, but I sincerely mean it. So how long does it take? Well, depending on the complexity of your program, but obviously, we have tools, and templates, and tweaks, and shortcuts. You can probably get a program up and running in six weeks or less. You want again, the benefit of tracking is to track over time. So you will want to track for several years. I think what’s insightful is to track for two years or three years, ideally for longer because no matter what product you work on, you’re going to want to see a full cycle if you were not aware. Mind you, you will extract some learnings much earlier. You’d extract some learnings from your tracking program at the end of the first month. But to guide your brand strategy over time you really want to stretch it. Ideally, your brand tracking program is just embedded in your brand strategy. And you stay consistent on the methodology and the sample mix and you keep consistent with just tracking on a regular basis. You keep the cadence consistent.


Jamin Brazil: How about setting it up? When you launch a project, a new tracking initiative with a brand, is it about a three-month process to get the initial set of data or is it longer or less?


Emmanuel Probst: I mean, three months is a good reference point, if you will. That said, a lot of people want to see learnings faster so you’re going to extract at an earlier stage. You can extract some tactical insights. Let’s say for example, that you have a specific marketing push, and you just embarked in a co-branding opportunity with a music festival. I’m just going to make things up. Or we have FIFA soccer coming up. So let’s just say for sake of argument that you’re sponsoring FIFA soccer. You will do a deep dive as part of your tracking study to understand the impact of this specific marketing tentpole event. So to your point, to really understand how your brand is trending over time versus competitive sets, you will need several months of data. However, to understand the impact of a specific marketing tentpole activity, or something that might be very technical in nature, you can tease out some learnings much sooner. Long story short, it’s all about the types of learnings you need and you want. You can extract some very valuable sound bites after a few weeks. For the most strategic work, indeed, you’re going to need several months of data.


Jamin Brazil: Tricky question for me to ask you considering who you work for, so I apologize ahead of time, but are there any DIY solutions that you’re aware of?


Emmanuel Probst: You know, Jamin, I’m actually very comfortable answering this question because I don’t sell the dream. I’m going to be very honest. Sometimes you want the involvement of a big agency. Sometimes you might not need that. So let me articulate. Begin with the end in mind, meaning, what decision are you making as a brand? How important is this decision for you? Because if you’re deciding on a $200 million media investment for the year, or marketing activation for over a year, that’s a very different deal than if you’re a startup, and you’re just getting your feet wet, and you’re just trying to understand your competitive landscape. So what is the size of the price, and what is the risk for your brand? And also, a very important question is, are you willing to take responsibility for the outputs? Because the upside of working with an agency, as opposed to working on your own is you transfer the responsibility, and you transfer the risk onto the agency. Meaning when things go right, you leverage the agency as your stamp of approval, and you say, “Ipsos said that blank.” And that carries some credibility for you in your organization. Conversely, if things go wrong, you blame your vendor. The major downside of doing things yourself, in my experience, is now you have to take responsibility for it. Meaning if things go wrong because you messed up your sampling plan, you’re on your own. So the honest, ethical or humble answer to your question is there is room in the marketplace for do-it-yourself tools, do it yourself solutions, whether it’s in tracking or in any other aspect of research. And I actually commend those firms in developing strong technology offerings. They have tools that are intuitive, that you can easily access online, you don’t need tons of training, and they serve a purpose. However, if the program has more visibility, the brand is more mature, throws out big decisions, and you need to transfer responsibility, of course, all this implies that you have the bandwidth. So when you want to do it yourself, that implies that you have 20 hours a week to dedicate to your program. Well, you want a big agency when you’re going to need the reach, also when you’re going to need global capabilities and I think most importantly when you’re going to need a stamp of approval.


Jamin Brazil: Well said across the board. My last question. Who are the go-to partners that companies should consider working with? Obviously, Ipsos and congratulations by the way of being the most innovative company, according to Career Report in market research. So obviously Ipsos is in that mix, but who are some other companies that we should be thinking about?


Emmanuel Probst: So, of course, that question, it becomes cheeky as we would say.


Jamin Brazil: Very cheeky. We see Nielsen, of course, has had a position in the market relative to media tracking, I think, historically.


Emmanuel Probst: Well, with no disrespect for Nielsen, I believe they are being very challenged.


Jamin Brazil: You’re seeing that of course.


Emmanuel Probst: From what I read recently. Look, and here again, I give you a humble, honest answer. I think it’s about the expertise, the vertical expertise of the team you work with. I live in Los Angeles. Here, you have smaller shops that do very well at tracking movies, for example. That’s a very niche industry. Or you have firms that specialize in tracking high-end luxury brands. Those are very niche, very tricky industries. I think at the end of the day is when you meet with your vendor, are you talking what’s most important? Are you talking with experts in their field? If you’re a DTC brand, if you’re in CPG, if you’re in tech, if you’re in pharma, you want to talk with people that have tons of experience in the field because they will be able- See, that’s where the value added is. They will be able to read those results in light of the competitive environment, the category, what they know about your market. So to me, it’s about what is the methodological and the vertical expertise of the team you work with, number one. And number two, it’s about this chemistry. Again, sounds very obvious, but it is so important that you align. I was going to say you get along, but you have a similar vision for the industry and for the program between the agency and the brand.


Jamin Brazil: Outstanding answer. I love the rubric that you’ve provided, and I think that’s exactly how I would think about it if I was commissioning a tracking study. Our guest today has been Emmanuel Probst, the Global Lead of Brand Thought-Leadership and Senior Vice President of brand health tracking at Ipsos, author and professor at UCLA. Emmanuel, thank you for joining us on the Happy Market Research Podcast today.


Emmanuel Probst: Jamin, thank you so much again for having me on your show. Thank you to all our listeners. Really always a pleasure connecting with you.


Jamin Brazil: It’s such an honor having you on the show and we all appreciate it greatly. For you tuning in through audio only, there will be a blog post associated with this particular episode. That will be hosted at the happymr.com website. And you can find a link to that in the show notes. Additionally, you’ll have links to Emmanuel’s LinkedIn, etcetera, etcetera in the show notes as well. So with that, I hope you have a great rest of your day.