TransUnion Live on LinkedIn: Meeting Consumers in the Middle

Andrew Goss/Host:

Well, welcome everybody to another episode of TransUnion Live on LinkedIn. I'm TransUnion's Andrew Goss, and I'm pleased to be joined by my esteemed colleague, Emiko Seale. And I am especially excited to be joined by Joel Rubinson, president and founder of Rubinson Partners, attribution expert at MMA, and dare I say, Movable Middles pioneer. Welcome everyone.

Just a quick note, you might notice some lag on our video today. We're experiencing some technical issues, but don't worry, we should be here and be able to discuss. So, we're here today to discuss a new marketing strategy focused on the Movable Middles, which is explained in a recently published MMA and TransUnion study. But before we get started, I have a quick housekeeping note for the audience. I always start off these conversations with, we welcome your questions related to today's conversation. Just put them in the comments area below. If they're applicable and we don't get to them today, we'll do our best to circle back around with you afterwards.

Now let's get started. First off, I always like to get the lay of the land to start these episodes, so can you two provide an overview of the recently published TransUnion and MMA study I referenced earlier about the segment you're calling the Movable Middles. Specifically, who are the Movable Middles and why did we decide to study them?

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

That's a great question, Andrew. Joel, I mean, you're the Movable Middles pioneer here. Why don't you kick us off and talk about where did the Movable Middles come from?

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

Sure. For many years, more years than I care to mention, I've always felt that there should be a relationship between someone's baseline probability of choosing your brand and their responsiveness to advertising. I mean, you could think of it this way that if you look back over, say, 12 months of shopper data and you saw that someone made 12 category purchases but never bought you, then that would indicate a low baseline probability of buying your brand, but it also means that they've passed on every one of your commercials, every one of your promotions, all your shelf visibility went for naught.

So if they haven't responded in any way over 12 category purchase events, what's the chance that they'll respond to the next message? So I felt there must be a relationship there, and we started working with the TransUnion team on this, working with in particular the data science folks to build synthetic data. And what we noticed was that in our simulated environment, it was always kind of the people who had a mid-range probability of choosing the brand who seemed to be most responsive.

And then when we dug into this further to try and figure out, yeah, why is this? Why do we keep seeing the same finding? Is there some mathematical principle that's going on here? And as we dug into it further, we found there actually is a very solid reason why it would be the people in the middle who were responding. And so, what we did was we said, let's define a segment that we're going to call the Movable Middles as those that have a 20 to 80% probability of choosing your brand. Those below that would be low loyals. Mostly they're close to zero in their chance of buying your brand and they're non-buyers. Those above 80 would be high loyals, and they're probably not going to respond because they have nowhere to go. They're going to buy you anyway, whether you expose them to advertising or not.

So, we said it should be the Movable Middles who were really hyper responsive, and our expectation was that they would be five times more responsive to advertising, not 5%, five times more responsive. And since then, we've done a whole series of experiments, and we've never seen the Movable Middles not be hyper responsive, and the degree to which they outperform non-Movable Middles has ranged from 2 to 23x. So I don't know, Emiko if you want to... Emiko is central to some of the field experiments, so maybe you want to talk about that.

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

Yeah, and let me, just for the audience, play back some of the things you said. Now you're just giving us this great explanation about the moveable middles, and sometimes we think about the loyalty spectrum as you just broke it out, where we have low loyals, and those may be people who have never purchased your brand and always purchased someone else. And then we have high loyals who almost always purchase your brand. And then we have the people in the middle who choose your brand and a few others, and we found a really strong evidence and proof that those people in the middle, and makes intuitive sense, are the ones who can be most influenced by advertising.

And so the work with TU and MMA Global has gone on for a few years now. We were involved in one case study where we looked at a group of Movable Middles for Ally Bank who's been a wonderful research partner. We were able to isolate a group of Movable Middles and compare them to a group of non-Movable Middles and saw a 23x difference in conversion rate over the short-term.

MMA Global has been, and Joel has been spearheading this, again, as the pioneer, have done about almost a dozen case studies similarly where they were able to identify a group of Movable Middles and compare them to a group of non-Movable Middles in a study situation. And that, again, ranged from anywhere from 2.5 to 23 times a difference. So great proof that when we're able to isolate Movable Middles and compare them to others, they are highly responsive to advertising. We found this really exciting at TransUnion. We were very excited in this work.

The next question for us was, well, how do we scale this? We found this in a study environment, how do we help marketers reach their Movable Middles, and the way that we do that as marketers or with audiences? So with TransUnion, we have this great tool set. We have an identity graph that covers almost everyone in the US. We have our measurement capabilities which range from aggregate to identity level measurement tracking, and then we have this audience segmentation framework in the middle that allows us to identify and isolate groups of populations and how they perform in terms of advertising. And so, what we aim to do and what we asked as the next step was: Can we create audiences that are rich in Movable Middles and will those audiences perform better than alternative strategies?

So that was the impetus of our recent research, which is now in this paper that we're talking about today, debunking the short and long-term myth of marketing. And in terms of our research approach, we took a traditional approach where we ran a test campaign, we ran it broadly so that we're reaching a large population, we were able to track, again with our identity capabilities, almost a million people for over 18 months actually. And we were able to see, okay, let's look at different audiences from the millions of the people we were tracking, and let's see if the ones that are rich in Movable Middles actually outperform other audiences. And the way that we were able to build the movable middle audiences were by taking the seed from the survey data and doing lookalike modeling around that to reach a much larger population.

So that was sort of the high-level design of the study. And then we tracked that population over time to measure the performance of the test campaign.

Andrew Goss/Host:

Interesting.

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

Joel, anything to add to that?

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

No. Andrew, you were going to ask another question?

Andrew Goss/Host:

Yeah, well, I was just... What comes to mind when we talk about this is all the work that went in. I mean, I think I should have removed your titles and I'll just call you marketing scientists, right? I mean, it really is interesting what you went through in this, but is there anything else in terms of that first question that I asked, kind of getting the lay of the land of the study in Movable Middles that you want to go into, or shall we move on?

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

I mean, the one thing I would add is that my orientation is towards finding mathematical relationships that then prove out in the real world. Because if you find a mathematical relationship that gives you a theory and that theory points to a particular outcome, then anytime those conditions hold, you should get that outcome if the theory and the math are correct.

So if they're correct, then that means you can have a great confidence that this will always work. So, so far what we've seen... So that's the movable middle is heavily based on math. I mean, I don't want to go into all the crazy stuff that we did, but this was really more than a one-year journey to discover the equations that kind of governed ad responsiveness and to then discover that they were working over and over again for a brand in an audience where it doesn't have much share versus where it does and see that we were able to have a predictable difference in how those two audiences would respond to advertising, that told us we were onto some universal principles.

Now, nothing in marketing works 100% of the time except so far the movable middle theory has, and I'm not promising 100% certainty because there's always something that can happen, but so far we've done 11 field experiments, and in 11 out of 11 cases, 100% of the cases, the movable middle has been hyper-responsive to advertising. So we really think that we're on to something because it's grounded in math that governs immutable relationships it appears.

And then, Emiko who she's pushing it in one direction, I push it in a different direction, but that's the beauty of it. I mean, going in the direction of, well, targeting audiences with high levels of Movable Middles kind of sounds like a performance strategy, like you can goose sales, and that's good, but sometimes you're interested in more buyers because you're interested in building the brand long-term, and that's the direction. So Emiko pushed it experimentally in that direction, and I pushed it mathematically in that direction, and we actually came to the same conclusion. So I don't remember if you let the cat out of the bag yet, Emiko?

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

Not yet. So Andrew-

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

Okay, then I won't say anything about what you found. I'll just make a drum roll on my table.

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

So Andrew, should I go ahead and start peeling into the results? Would that set up by Joel?

Andrew Goss/Host:

Yeah, just stand by for one second because I wanted to remind the audience. I see some more people have joined since we started. Just a quick reminder, if anyone has any applicable questions, we welcome them. Just put them in the comments area below.

But let's get back to my questions for now. Let's get to that juicy part of the conversation. Let's get to the results of the movable middle study, both short and long-term findings, and then what I'm most excited about, myth busting, what myth did the study dispel? So I'll turn it over to you two to get to that.

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

Okay, great. So I'll start digging into the results. So as I mentioned, we were tracking about a million people over an 18-month period. We had built audiences rich in Movable Middles, and the first question was, can we find audiences that are rich in Movable Middles? The answers was yes. We had this really interesting finding of sort of a younger cohort of young adult singles and families, I think it's what we call them. And then we had another cohort, which was older established adults and retirees. And I mentioned that this was Ally Bank was our partner, but the study was focused on examining responsiveness to advertising by Ally Bank, and we were looking at account openings.

And so again, this very interesting finding. We have this younger cohort of people starting out being more movable than others, and this older, more established retiree cohort. And so then we looked at their performance. First, the case studies that Joel mentioned were really focused on the short-term. So that's where we started. And we expected to see, obviously if... I mean, not obviously, that's why we're doing the research, but our hypothesis was that audiences rich in the Movable Middles will perform higher than those that aren't in the short-term.

So originally, we looked three months after the campaign and we saw, in fact, compared to sort of the broad reach campaign that the Movable Middles did extremely well, I think almost two and a half to three times better than the broad reach campaign. Then we thought, okay, well maybe it is targeting, and broad reach is not, so maybe it's not so surprising to see the Movable Middles performed better in the short-term, so let's look at other types of strategies.

So we looked at about, in the paper, I think we have about half a dozen other strategies. And in fact, the Movable Middles did perform better than the other strategies. And interesting, the demo-based strategy, we looked at an audience modeled around a typical financial services demo, and it did over three times better than that audience in the short-term. So we're very excited about these results because what it meant was that we could actually scale the findings for marketers and reach large groups of the population and really drive impact for marketers.

Now the next question was, well, this great that we see this in the short-term, but what about in the long-term? Are audiences that are rich in today's Movable Middles also going to be affecting people in the future when they're ready to buy or open an account? And so for that, we looked at actually going from 3 months to 18 months post-campaign, and we saw first looking at the number of sales or number of account opens that in fact the movable middle audiences continue to perform at even a higher multiplier versus the other audiences in the long-term. So that was, again, very exciting that we were able to drive both a short-term result and a long-term result with this unified audience strategy around the Movable Middles.

But then the next question was, okay, well, is this long-term result the same people just continuing to convert at higher rates, or are we seeing customer growth in the long-term? And that's really important to brand marketers. There's this thinking and general way of operating that you do have to go, I think everyone agrees, you do have to go bigger for brand because you want to reach more people, but do you have to go as big as everyone is going? And so we're finding a way. So we're asking whether or not a movable middle strategy could actually be an alternate and more effective way for brand marketers to reach that sort of upper funnel audience that they're trying to reach to grow their brand.

And so, when we looked at non-customers, people who weren't customers at the beginning of the test campaign, we found that the audiences rich in Movable Middles did perform better in the long-term. So a little bit of a long-winded way to get there. But in the long-term, movable middle audiences do perform at a much higher rate in terms of driving new customer growth. And I think that, and again, really high multipliers, about three times versus the broad reach and the demo-based strategies.

And so that was huge, a huge validation, and we're really excited about it. And so that was the myth. The myth was that you have to target everyone to grow your customer base, but in fact we're saying you can be much more efficient by targeting the Movable Middles. Joel, do you have anything to add to that?

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

Yeah. So I mean, this might just really reinforce what you said, but basically what I see when I speak to marketers, and we speak to dozens of marketers through my relationship with the MMA, what I see is there's a bifurcation in media strategy that when people think they have a campaign that's oriented to performance, then it's okay to turn on all the science and turn on the targeting and the machine learning, and Movable Middles kind of sounds to them like it falls into that bucket. Maybe better than what they're currently doing for that bucket, but it's still in that bucket.

Then there's the type of campaign that they might call upper funnel, or they might say is brand-oriented, but what happens is there's nothing wrong with thinking that way. What's wrong in the industry, the mythology is to equate that goal with going broad in your media that you should try and reach everybody with a mouth or everybody with a wallet. No, that is really engineered waste and it's actually a really bad idea, because what we find is that when you target those same audiences that are rich in Movable Middles, as Emiko said, you do much better at bringing new customers to the brand or to the business.

Now, when we first saw this result, we said, how could this be? There's got to be a mistake because if they're Movable Middles, they're probably buyers. But then when we dug into it further, we found the truth of why this must be actually the way that Emiko found it to be from her data.

And one element to think about is when you target audiences that have a higher-than-average concentration of Movable Middles for your brand. Let's say your brand has 15% Movable Middles, and you find an audience with 30% Movable Middles. Well, that would be a great audience to target because they have twice the concentration of Movable Middles.

However, they still have 70% who are non-Movable Middles in that audience. There are still 70% that fall into what you might call a lookalike for Movable Middles, and therefore they are more likely to be interested or receptive to your messaging because even though they're not currently buyers or customers, they are sharing interests or demographics or life stage characteristics that are highly similar to Movable Middles.

So there's a lot more in the math as to why it really has to work this way. But the point is I hope we can get you to think a little bit differently about how to do your media planning, especially when it comes to more brand-oriented media planning.

And one final thought about that is if you're thinking upper funnel and you're thinking brand building, you're probably thinking new customers, I want to bring new customers to the business. So the hard evidence that Emiko found is that actually the most logical way to do that, bring new customers to the business is to conquest, it's to use a media strategy that's focused on non-buyers of your brand, but buyers of brands that you think you're competitive with.

And what we found in every one of these, what we call bake-off analysis, so far is that that is actually the worst performing strategy is to go after inveterate non-buyers of your business because basically they found something else that they like better and it's going to be very hard to move them away. So we really want you to think differently, not only about performance, but even more so about brand building and bringing new customers to the business.

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

And I'll just reinforce some of those points. So you talked about, Joel, I mean, I think this is, again, sort of bringing it back to the practicalities of how we think when we're putting these media plans and marketing strategies together, we often pit sort of brand and performance against each other, or not necessarily against each other, but we think of them as separate streams. And what we're really saying is that you can have a... Yes, from an audience perspective, from the people that you're talking to, you can actually approach it in the most efficient way through the Movable Middles.

And again, one of the ways we think about resolve this short-term versus long-term results dilemma is that those audiences that are, again, the Movable Middles of tomorrow look a lot like your Movable Middles today. So from that perspective, it makes sense why this one audience strategy would work for both. And so hopefully we can give this new thinking, new vehicle to marketers because so far we're seeing a great impact, and it's pretty exciting.

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

Yeah. Can I add one more thing to that? So if you're thinking about marketing effectiveness, advertising effectiveness, it's kind of like a three-legged stool. There's the component that relates to using the right channels and partners for your brand and for what you're trying to accomplish. There's the leg of the stool that relates to creative, and then there's the leg of the stool that relates to targeting, okay? And our in-going hypothesis was there's a lot of research and science relating to those first two legs. How do you get the right channels and partners and ad units, and how do you craft the right message and pre-testing that message? But there wasn't nearly as much science or effort that went into targeting.

So what we're talking about today is targeting, and what we want you to do is kind of decouple in your mind the targeting question from the goal and from the creative and from the ad unit question. We're not saying... So, if you're trying to drive performance versus trying to drive brand predisposition, what we're saying is that the exact same targeting strategy based on Movable Middles will work for both goals. We're not saying use the same partners or ad units. We're not saying that you don't use display for one thing and rely more heavily on video for something else, and we're not saying use the exact same creative message. We're certainly not saying use the same creative message, but what we are saying is the same targeting might actually work for both.

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

Absolutely. That's a great... Thanks for adding that, Joel. That's so important. We're not saying that there aren't different strategies, but that the audience component is something that can be considered through the Movable Middles together. So hopefully that's something for marketers to take back to their planning teams and start integrating, because again, we're just very excited about this work and encourage people to take a look at the white paper, and yeah, start joining the conversation. We've been having a lot of good ones lately. So yeah.

Andrew Goss/Host:

That's a great segue, Emiko, for my last question, my question, but before we get to that, I wanted to remind everyone, and I see at least one coming through in the chat right now in the audience that if you have any applicable questions, please put them in the comments area. We'll probably get to them right after we answer this next question, so stand by on that.

But for the last question I had, I want to tackle an area I imagine many attendees are waiting for a little bit more tactical here. While the findings are very intriguing and certainly for a Movable Middles marketing approach, that's a mouthful, how can brands implement this Movable Middles approach?

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

Yeah, that is a great question. And again, I'm going to refer back to the white paper because it's there documented for everyone to refer to after this, but it starts with that being able to identify Movable Middles seed audience that will depend on the data available. But as you know, there is a lot of data available for marketers to use for this. So we would find that data source, build a seed audience, then we build audiences around that seed audience, so lookalike modeling, activate those audiences and then measure if we have the right measurement framework in place. So those are the basic steps. Joel, anything to add to that?

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

Yeah, right. I mean, I would just reinforce that the white paper is actually footprints in the snow for anybody who would want to replicate this analysis. Whether you are reanalyzing old data where you saved ad serving and conversion data and you want to, in this case, put it against the TransUnion identity graph, or if you are considering a campaign where you might build a testing element into that campaign, the white paper basically shows you how to do it.

Now, the only element that is not inherent in what TransUnion offers is classifying people in terms of whether they're Movable Middles or not. So there is one more piece that has to be brought in, which is mentioned in the white paper. In the case of the Ally study, we used survey data. You can use other sources also, but we do have a survey instrument. I mean, it's more than a questionnaire and it's more than surveys because these were carefully constructed questions in a certain order in a questionnaire that we've now used on four studies.

So it really is more of an instrument, I would call it, where it will allow us to score the survey respondents 5 to 10,000 or even more survey respondents on whether or not are they movable middle, low loyals, high loyals, or maybe even non-category buyers. And by classifying them into those four groups and then onboarding those IDs in every one of those segments by onboarding that to the TransUnion identity graph, that's where the implementation part of it takes place.

But basically the white paper does give you footprints in the snow of how to do it, and Emiko would be, I guess, the right person even to contact to really get started with TransUnion. Or if you have a TransUnion rep who you're working with, you would contact them. But we are really hoping to start establishing a broad number of marketers and a broad number of sectors who now have kicked the tires on this media theory because we really think it makes marketing work better.

And if it makes marketing work better, everyone comes out ahead. The ad tech partners come out ahead because they provided tremendous value for their offering. Marketing inside of a marketing organization comes out ahead because they can show to the CFO, "Look, marketing really works. Look what we were able to drive." Consumers come out ahead because when you get greater response to your advertising through better targeting, that's really a surrogate for the concept of relevance. You are delivering messages to people who find that advertising more relevant and that's why they respond. Irrelevant advertising adds very little value to someone's life. Relevant advertising can add tremendous value to someone's life. So you're really doing the greater good when you make marketing work better.

Andrew Goss/Host:

Great. Anything else on that point before we move on? We've got at least one viewer question coming in. Okay. We'll move on then. And apologies for the pronunciation here. I'm going to try my best. Kalpana Sehwani, hopefully I got that correct, at InMarket asks, how do you define performance? And this is important, right? Is it sales lift or ad engagement or something else?

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

Emiko, you want to take that or you want me to take that?

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

Sure. We've used performance a few different ways here. So at least I guess in one sense we were talking about brand and performance creative, and then we're also talking about business performance. So I'll start with the first one. When we think about brand versus performance creative or campaigns, we generally are talking about driving for performance short-term performance, whereas brand strategy is more focused on the long-term. So when we sometimes refer to that brand versus performance dilemma, we're talking about long-term growth of brand versus driving short-term gains with our performance objectives. When we were measuring general business performance for the study, we were looking at account openings over time. So that was our business outcome metric that we were using for the study.

Andrew Goss/Host:

Okay. Anything else on that that you want to touch on, Joel?

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

I think Emiko brought up a really interesting way of approaching the question. When people think about brand versus performance marketing, brand advertising or advertising that builds positive brand perceptions should also deliver performance. But the short-term effect of getting more people to be positive to your brand is already subsumed, the economic value is already subsumed in the short-term sales gains that you see, the incremental sales that you see in any kind of test versus control experiment or MTA analysis.

So what we're saying is that the short-term stuff, the sales lift stuff is kind of what we mean when we talk about performance. But when we talk about brand, we're still talking about performance. It's just that it's the part of the performance that hasn't occurred yet, but where that favorability, that incremental favorability that you are creating will pay off as an annuity over time for that brand.

Now, we've analyzed that as part of the MMA brand as performance research program, and TransUnion has been one of the partners for those studies. And what we find is that when you build favorability to a brand, those consumers are three to five times more likely to convert and become customers over time.

And the second thing we find is that favorability has a very long half-life. They might not remember the ad for that long, but they remember how they feel about the brand. So if the ad changed the way they feel or think about the brand, that feeling is likely to stay with that consumer. Well, the average half-life is well over a year. So the advertising that you do today can actually have a multi-year benefit for the brand in terms of sales performance. So when we use the word performance, we mean short-term. But that's not to say that brand building and brand advertising doesn't have a performance dimension. It's just, it's the long-term part of it that you haven't quite seen yet, so you have to be patient.

Andrew Goss/Host:

Okay. It looks like we don't have any other viewer questions at least now. I guess I'll say, Emiko and Joel, any last-minute thoughts, kind of a gotcha question here, but anything else you want to hit on before we close things out?

Emiko Seale/Sr. Director Marketing Solutions, TransUnion:

I mean, I'll just say that this is a really exciting area for us, and I think there's huge potential for marketers. I would encourage anyone tuning in today to reach out to us and join the conversation as I mentioned before. We really feel like we're just scratching the surface to what we found here.

And I just want to mention again what a great partner Ally Bank has been in all of this. They have a great team that we've been working with for the past few years, and they've been really leaning in. Joel has mentioned that he's worked with almost a dozen brands, and we're definitely getting some momentum around this. And we want people to come in, be interested, ask tough questions. It just pushes us harder to really uncover more value for marketers and for the consumers that are receiving the marketing.

So I just, again, encourage people to reach out and lean in if this is catching their interest at all. But again, we found it to be a really lucrative area in terms of the richness of insights. And again, we feel like we're just getting started, so.

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

And the only thing I would add is if you're already working with TransUnion, it would be really easy to kick the tires on the movable middle theory. If you're a member of the MMA, it's also a lot easier to kick the tires because as representative of the MMA, basically I provide free direction on how to fast track the analysis. And if you happen to be a member of the MMA and the customer of TransUnion, then it's really easy and kind of shame on you if you don't kick the tires on this. So that's what I would say that this really has the potential to be transformative, so what are you waiting for?

Andrew Goss/Host:

Well, thank you, Joel. Thanks for the plug, but most importantly, thank you both of you, Emiko and Joel, for these great insights, really valuable stuff and eye-opening conversation. When I'm talking to people across the enterprise at TransUnion, I'm never surprised to see all these insights that we have.

And a special thank you to the audience for joining us. Again, if we didn't get to your question today or if you didn't ask your question yet, please put it in the chat here and we'll do our best to get back to you soon. Now, to download the findings from our recent study that we discussed today and both Joel and Emiko talked about, I'll be putting the web address, transunion.com/movable-middles, and a QR code up on the screen momentarily. You can also find it by going to the LinkedIn Live event page description tab. And until next time, we'll see you on TransUnion Live.

Joel Rubinson/President, Rubinson Partners and Sr. Research Advisor, MMA:

Thank you.