When brand tracking actually changes decisions

The preview

Dashboards update in real time, but most brand insights still never change a single decision. Alexander Shmelkov, Digital Product Leader for Brand Pulse at Inter IKEA Systems, rebuilt a 20-year-old brand tracker from the ground up and knows exactly where that gap lives inside a real organization.

He walks through aligning 64 markets around one shared definition of what the brand is for, replacing bloated legacy metrics with category entry points tied to actual growth drivers, and building change management muscle so country teams trust a new number instead of defending the old one. The result is a model for turning tracking data into something marketing and insights teams can act on together, not just report on separately.

The interview

The transcript

[01:18] Alexander Shmelkov: Hi, Steve, thanks for having me.

[01:21] Steve Phillips: I wanted to start, we're going to talk a lot about the transformation you've done at Ikea, which is an amazing case study in moving insights from data to market research, from data to insights. And we'll get into depth on that. But I want to start by asking you, and you've got a varied background in tracking across multiple organizations over the years. Tell me about a little bit about the history of brand tracking. Why do companies want to track what their brand does and how does that impact brand decision making?

[01:51] Alexander Shmelkov: Well, brand tracking is probably one of the most traditional parts of market research, goes back to Nordenbound who established the practice and the premise was the business information available to the company back in the day and still did not cover all the questions and did not answer all the questions like, how can we understand consumer decision making? How can we better understand what our target group is feeling, thinking, and knowing about the brand and about the competition? So these questions were nowhere to be found. And that's how brand tracking appeared as a practice. And to this day, in my opinion, they serve two main purposes, basically. To understand something about memory of people, what they remember about the experience, and get an idea about how they perceive things — how they perceive markets, competitors, the brands — and puts this all together with the business performance.

[02:52] Steve Phillips: That's great. And I know Mark Ritson has talked about how can you have a brand plan if you don't know where your brand is. And brand tracking tells you where your brand is now, but also where it's come from. On a basic level, your marketing teams — I presume they're using it for their quarterly reviews, but also, more importantly, for their annual brand plans. Does it become a key component of their planning process?

[03:17] Alexander Shmelkov: Yes, absolutely. Brand tracking serves two functions. It gives you the retrospective view of where you are. So there is always a bit of a lagging effect to that. But you use this follow-up also to understand what is your next step — to set up goals and to get the idea of how to get to the goals.

[03:38] Steve Phillips: Yeah, yeah, that makes perfect sense. Now, you came into Ikea, you'd had a brand tracking system that had been in place for two decades. What was the old model failing to do for your marketing clients and for the business? And what did you want the new system to do differently to that?

[03:59] Alexander Shmelkov: It's all a simple story. So we started unraveling this question from several dimensions, trying to understand what is working, what is not working, what needs are underserved, where basically gaps are. And very quickly it became clear that there is one common thread which is missing. And this is when we ask the question: so let's stop here and understand, how do — like, what is brand for Ikea and how do we use it to drive the growth? And we immediately start having very nice debates with different parts of the organization. So for me it was a sign that you cannot actually engineer this moment. But this was a revelation that the problem is misalignment. Misalignment on the way we look at the brand. And of course it makes total sense if you think about how different functions and practices are set up. Somebody's looking more into retail topics, somebody looking more into on the marketing side, on product development. So there are a lot of good opinions, understanding, but we needed to have a common foundation. And this is where we realize that this misalignment, misunderstanding of what is the one common purpose that the brand should serve the company, is a key challenge. The other thing, of course, was just because the tracker was so old, there were a lot of obsolete elements, some objectives which were set decades ago when IKEA was not even online, let alone omnichannel brands were still there. And yeah, it was a very big, bloated program which was very slow moving and basically giving a lot of information for us to say, okay, our concept works, but doesn't really show what consumer thinks and feels.

[05:57] Steve Phillips: Yeah, it's a bit like running a technology company. You end up with tech debt — I think with running a brand tracker you end up with research debt and everyone wants to pile in. Did you have to have a discussion internally about how much you could get rid of? How easy was it? Because some people — there might be one part of the organization that is desperate for one KPI and someone else wasn't. How did you manage that process of whittling it down?

[06:24] Alexander Shmelkov: That was the hardest part of the discussion. Of course it's very hard to give up on something. And some pieces were — they were not wrong, they were just not necessarily a part of the brand effects we want to study. But if you think about, for example, some sustainability topics, it's a big strategy, but it deserves very specialized research. And we kind of actually put everything in one basket. So it took time for us to go across the organization, to form some alliances, to convince people. There was a lot of packaging and repackaging of the story. And while we were doing that, also our understanding was growing, and we also — when we had this revelation that misalignment about the role of the brand and how growth is actually coming to a brand came to us — we realized that we need to bring in a partner in discussion. This is when we started working together with Atomy Boss Institute to create this common foundation, not from a brand tracking perspective, but from the perspective of the business, to set up some ideas on which we can basically — the way I'd describe it — the brand.

[07:52] Steve Phillips: And so you worked with Aaron Bog Best to bring in category entry points as the key element of the tracker, is that right?

[08:00] Alexander Shmelkov: It is certainly a key element of the tracker, and in fact we started introducing the concept into the vision of tracking design even before we partnered with them. But the objective for the partnership with them was broader — to establish a solid understanding of what is marketing, what part brand plays, what evidence do we have about how growth occurs for the brand, how big brands stay big, and so on and so forth. So it was way more than just focusing on a particular framework. And of course category entry points was one part of this discussion, but only one.

[08:48] Steve Phillips: This is an amazing process that you must have gone through — completely rethinking the purpose of the brand tracker, completely rethinking what's in it, having to go around the organization, persuade different departments, presumably different countries, 64 of them, that the new way is the right way and it will provide them more insight. How long did that process take and what did you learn along the way about bringing the organization with you?

[09:16] Alexander Shmelkov: Well, it took us an enormous amount of time. All in all, from when we started to the launch of the fieldwork, it took us about four years — three-something, and it can be different in different organizations. IKEA is a privately owned company. We may not necessarily be pressured by Wall Street to prove our worth on a quarterly basis, but at the same time that makes something different. The key was basically — it was a massive change management process. I would say for me it was a change management school, considering the complexity of the company — that we have a franchise system, that we have 12 retail partners, we are present in so many different countries, and on the franchisor side the global functions also have a very different setup from the majority of other FMCG businesses. For example, I used to work in agency time, and the change management there was basically about showing what we are losing — having no tracker, showing what brands can give the company if we think about it from a certain perspective. So there are a lot of discussions related to that. Like, we need to focus on penetration, because IKEA is probably one of the best embodiments of a mass market business, which is based on volumes, based on mass reach. And kind of starting with these foundational ideas, showing what we're missing out on — having this tracker which doesn't give us a lot of understanding in some places. Like, for example, we know already with the legacy solution that we have enormous awareness. But if we have enormous awareness, why can't we grow faster? And then — by enormous awareness I mean, like, figures like 60, 70% spontaneous. I think, like, it's a great result — what else can you do? But then there's something missing, right? And can we open a dimension to look at the quality of this awareness? And these small discussions kind of piled together and created these big movements, when finally key stakeholders and the organization in general, and management, started believing in the story, and that's how it came to life.

[12:04] Steve Phillips: Yeah, so it's brilliant. You've managed to rethink brand tracking. You've gone through a large matrix organization, pulled people on board, and you talk about, obviously, that's a major part of the role. Then when you actually get to the tool itself — you've talked a lot about the velocity trap, and certainly coming from a Zappi perspective, we set out to make quality market research faster and cheaper than ever before. But agility isn't everything. We know that speed isn't everything. When you talk about the velocity trap, what are you thinking about from a researcher's perspective?

[12:46] Alexander Shmelkov: Yeah, I think about a couple of things. On one hand, there is a certain speed of decision making which we cannot actually increase, because at the end of the day decisions are taken by humans, and decisions involve some kind of thinking. So we need to give humans time to think, even if we use the best technology, if our workflows are completely AI based, or it's not a machine taking the decision. So secondly, some things are not changing that fast. So when we do some big investments in product development, in prices, in many different things, there is a certain lagging effect of how quickly it will sink into consumer minds and how quickly we can register it with a tracker. Of course, if you, I don't know, put a banner sale in front of the store, you will probably see the same day a spike of transactions, because people see the banner, they start coming to the store. But when you try to understand things like perceptions — right, do people perceive the brand as more affordable, do people perceive the brand better in some aspects over the competitors, and so on and so forth — this will take time. The same will happen with memory metrics, and so on and so forth. So some things, granted, are moving slower, like perception — it's a more stable thing. People are not jumping from one category entry point to another on a daily basis as well, so they're also quite stable. So there are a number of elements which we include normally in the tracker which will not move very fast. But some things, of course, are changing and more prone to change, and it's worth allowing organizations to see them in a more dynamic fashion. But again, there is a certain reasonable level — you do not want to look at change in some awareness or consideration figures on a daily basis, because it will just paralyze your decision making and basically represent more statistical noise rather than a real market change.

[15:18] Steve Phillips: And it's really interesting, obviously, when you're coming into creating this new tracking environment compared to when — certainly when I started my career, but then I'm a little older than you — brand tracking was almost the key way that marketers would see where their brand was and what the momentum was. Obviously now there are other forms of data that are helping with that. You've got much better sales data, probably instantaneously. You'll have social media data, you'll have reviews data being analyzed for you. How do you see the role of those different data streams interacting with the tracker? How do you see them complementing each other, or how do you think about those multiple data streams?

[16:03] Alexander Shmelkov: We started our journey with the premise that we need to in-house brand tracking. And by in-house, I mean normally that you will find easy access to the data which can help you arrive at the same conclusions or answer the same questions. And of course we know that it's not easy to in-house a survey, because you need to have a partner running it for you, especially at a scale like that. So we were looking at different sources to see if they actually are good enough to replace it. And we found out that it's not like that. In some industries it probably can be easier — I'm thinking about different digital platforms of different sorts. But still, when it comes to brand understanding, right — so the only way to gain brand understanding is either to listen to real-life conversations where people discuss it, or ask them directly. And the problem is, of course, people do not go online to discuss how they drank a beverage or bought a couch. They do it sometimes, but not on a daily basis. And the completeness of this data, the representativeness of this data — although it's a massive data set, of course it's a good question. So we decided, for ourselves, for our situation, that survey is still an important component, but we would be stupid to ignore other data sets. So we built the tracker in a way that can help us complement it with different data, internal and external. We have data — we're hosting data in our native data platform, so we have access to it on an ongoing basis and can combine it with other data sources and do the analysis which we need to do. So that basically was part of the vision.

[18:13] Steve Phillips: Great. And so we've talked a bit about what you set up and why, and it's an incredibly complicated thing, not only to set up but to persuade people what you're doing. So huge congratulations for doing that — it's an amazing job. But tell me about how you landed it. So you've now got this new system, it's much more up to date, it's much more fluid, it's answering the current questions that the brand has. You're integrating it with other data streams. You've got multiple departments in very many countries. How are you getting them to not just have access to that brand tracking data, but to make decisions based on that tracking data?

[18:57] Alexander Shmelkov: That's the toughest point. And our vision is that we need to move away from the mindset that we're here just to democratize data or generate insights and then leave it aligned on the table. So basically our role as an insight function is to activate. So activation is the most important part. And this means, essentially, that we need to have people who can be close to the business, but also not just the business-partner element of it, but also the whole delivery chain of different people, starting from the data set and going all the way to the business, where we can transform this vast knowledge into insight pieces, into thought pieces for some specific requests, for standardized topics, but also as expertise in a broader sense, in broader practices. And move it, move it to the business. It requires a lot of human effort. So basically this core process requires several different competencies which we need to put in accordance with each other. But the key, in my opinion, is to have the right people with the right competencies who can take the results, the findings, and basically translate them to the business in the way of — so now what? Now what are we doing with that?

[20:35] Steve Phillips: So give me some examples. I mean, you've got stores all over the world, you're based in Europe. How do you get — now that you've implemented this great, shiny new tracker, much more improved, much more readily available — how do you persuade the team in Brazil to take action based on this new data, this new insight that you can deliver?

[21:05] Alexander Shmelkov: Several, several ways. First of all, Global Brand Track is part of our franchise offer. So we own the brand. We also own the measurement system. We used to deploy our legacy solution, Brand Capital. We phased it out, replaced it with Brand Pulse, with a kind of more convenient access. We had dashboards, reports, and better cadence, and so on and so forth. And of course we translated this knowledge through our insight metrics to all the countries. So countries have access to that, and from certain point of time they lost access to the legacy solution.

The goal setting process is globally set when it comes to rent goals set on the Brand Pulse. At the same time, we know that some countries sometimes used to use local trackers as well, alongside with the global trackers. Because known problem with the global tracker, you cannot actually satisfy every situation and every team, because countries are different, cultures are different. So we need to find a balance between the cultural specificities and the global consistency. And sometimes country just want to have additional angle which is not a part of it, and that's fine.

So we have this situation. We know that in most cases these trackers are used to complement, because they focused on different topics, like, for example, more specific on communication, on different audiences. Whenever there is a competition, we of course need to have a dialogue with the country, because you cannot use different source for planning, right? So it should be consistent. We should keep building consistency across the whole value chain. So that kind of dialogue, and that, of course, requires also a lot of change management and adoption. Because the first reaction when people see something new, especially new tracker, is like, "Oh, everything is different." Help me understand how and why did

[23:23] Steve Phillips: you spend a lot of time on, on airplanes for the first year or so?

[23:27] Alexander Shmelkov: Well, we spent some, but of course with, with digital capabilities, it's much easier

[23:33] Steve Phillips: to have these dialects.

[23:34] Alexander Shmelkov: But yeah, it was, it was a big part of the change management, because you can see a trend break as a something which says that the new tracker is not worth it, or poses a big question, or you can see a trend break as a learning moment. We took the second route, and let's acknowledge why the change occurred. Let's acknowledge that we're not measuring the temperature here, we're measuring something relative to something. And yeah, that was our approach, to help teams embrace the difference as a kind of normal situation, not as an oddity.

[24:16] Steve Phillips: Brilliant. And of course we haven't talked much about AI. We've been talking about people and persuasion and behavior change. But during this process, obviously AI has been getting better and better, or more and more usable for these types of things, particularly for analysis dissemination. As you were going through this process, how did you think about the role of AI? How did you integrate it into your systems, and how did you make sure that you didn't lose sight of actually needing the people on the ground, in the markets, in the stores, to actually use it and make behavior change?

[24:56] Alexander Shmelkov: I'm laughing because there were two moments along the way which was like that. The second one was with AI. The first one was with the dashboard. The day we started developing the dashboard, I opened LinkedIn and I see the picture: Dashboards are dead.

[25:16] Steve Phillips: Yes, yes.

[25:17] Alexander Shmelkov: I was like, okay, what are we doing here? The same was with AI. When AI exploded, of course everybody were super hyped, but okay, how can we use it right now, and what does it mean for the whole operations, for the whole concept of the tracker? Can we do it differently? We took pragmatic approach. So in my opinion, AI, as any technology, is a huge amplifier of good and bad practices. If we had bad data, if we do not know what to do with data, if we do not have any frameworks in place, AI won't help us to be smarter. Because you need to help AI contextualize what you work at. At the same time, when AI appeared, until to this day it's still in the early phase of the development. We didn't want to have hub back solution, so we started some small limited tests where it can be used. So now recently we launched AI in our backend and the data, because it creates a lot of productivity there. It's super helpful for data engineers. But this means that we also had to go through all the data, all the notebooks, prepare everything so artificial intelligence can work with that efficiently and provide a real efficiency. Of course, the main push was to implement AI in the end of delivery change. Let's use it to summarize and make and derive insights from the data. I would be careful there, because in my opinion this is still a human domain. AI can conclude quite solidly on the situation, and in most cases, especially if you prepared it right, if you contextualize information for it, it's doing a good job. But I still see signs that it often miss some kind of human aspect of what is hidden in insights. And it's very much like Mark Pritchard was talking, accounts from a CMO perspective, that yes, AI is a good backbone, but it won't build the brand for you, because it's missing this human element, and we are here to bring it. So the same goes to good insights. So we use it as a backbone for the backend. Now we're testing if it can be a help also for insights leaders and insight specialists in their day to day job. But not so much yet as a kind of silver bullets to create, to get inside. Yeah, exactly.

[28:19] Steve Phillips: This has been an amazing story. What a transformation. It's taken a long time. You've had to take a huge organization with you. As we know, brand tracking is at the center of brand. It's about where you are, it's about where you want to go, and helps you understand that. Having gone through this translation, how have you thought about measuring the success? You've obviously made this major change. Can you measure the change in the organization over that period of time? Have you, did you have KPIs that you were thinking, we want to transform over the change

[28:55] Alexander Shmelkov: we do have. We had a lot of discussions about that, starting from the extreme views, like, can we do ROI on insights? And of course we can't. But it's also worth discussing why. So now we have some satisfaction studies which goes across the network, which help us to understand more on the practical side how people use. How do you feel about the solution? What are pros and cons? What are gaps in many ways, in the delivery, in the support, in some more kind of substantial, fundamental elements. We of course keep a close connection with a community, with the site community, and with the business, having regular metrics calls where we are kind of receiving a lot of information about how the product being used, and then, of course, observing and supporting business in dissemination of Brand Pulse results into business. And when you see CEO, and CEO talking from the big stage, addressing Ikea and talking about the plans and challenges, when the Brand Pulse is one of the biggest contributors to creating the inside behind their presentation, it's, of course, a great recognition. So it's a combination of these qualitative factors with more focused user satisfaction and usability research.

[30:43] Steve Phillips: That must have been a fabulous moment. Congratulations.

[30:46] Alexander Shmelkov: Yeah, it's of course nice.

[30:50] Steve Phillips: Right, now let's move on to our lightning round, where I'm going to ask you a few quick hitting consumer related questions. Are you ready? Excellent. Okay, first one is a bit of a softball, but faster data or better interpretation?

[31:06] Alexander Shmelkov: Bad interpretation. Cost doesn't mean good value is in interpretation of what you derive. Of course it's a combination of value, cost, and speed, but value, for me, is the main pillar.

[31:24] Steve Phillips: Brilliant. Next. What's one brand tracking KPI you think gets more attention than it deserves?

[31:33] Alexander Shmelkov: Oh my God, there are so many. I would go with purchase intent, because it had nothing to do with what people were going to be doing in a year, at best. It shows you the way they stand now with your brands. But at the same time, we also use very complicated approach to ask these questions, with all the scales. Great for analysis, but not necessarily great for respondents. And I think that this metric is extremely overbearing, although convenient.

[32:07] Steve Phillips: Yeah, no, I'd agree with both of those. And what's one skill every leader working with consumer data needs that has nothing to do with methodology?

[32:16] Alexander Shmelkov: Soft skills in general, persuasion. Be able to persuade, in different channels, people with different beliefs, and sometimes with opposite beliefs.

[32:29] Steve Phillips: Next, what's one sign that an insight has actually influenced a decision?

[32:35] Alexander Shmelkov: When you see that conclusions and recommendations are taken by the business, and the action is taken on this basis and on this premise, and, of course, the result of the action.

[32:51] Steve Phillips: And five years from now, what's one thing about brand tracking you think will look fundamentally different?

[32:57] Alexander Shmelkov: I think that technology will further change how we do trackers. There's probably going to be more impact from Insights, I don't know, probably even from synthetic audience, which is rapidly growing area now, that will definitely change the way we access data. I also think that maybe in five years from now it will become even more complicated to do traditional service.

[33:28] Steve Phillips: Yeah, yeah, very likely. Before we go, we have one last question we ask all our guests. What's the one thing that marketing and Insights teams need to do together in the next 12 months to help your company win with consumers and grow their business?

[33:44] Alexander Shmelkov: Well, in the time like that, I think we all need to get together, step back, and look again at our value propositions, because this is the fundamental challenge for every business, not only Ikea, for so many industries. I do think that this is one of the hardest challenge, and one of the most questions to address.

[34:10] Steve Phillips: Yeah, when things are moving rapidly, taking the time to sit back and think about it, it can be incredibly valuable. But quite often people miss it. I think that's a great answer. That wraps up this episode of the Inside Insights podcast. Thanks to Alexander, director of Global Brand Tracking at Ikea, for joining us. If you'd like to contact Alexander, you can find a link to his LinkedIn profile in our show notes, or at InsideInsightsPod.com. If you haven't subscribed yet and want a regular stream of insights and Insights knowledge in your podcast feed, hit that subscribe button in the podcast app, or follow us on YouTube. Okay, that's all for today. See you on the next episode of Inside Insights.