E5: The Shifting Shape of Data

As 5G becomes more prominent, proactive marketers are going to have to plan differently, think about data and targeting differently and even partner differently.

In our 5th episode, The Shifting Shape of Data, Jake Moskowitz outlines the significant changes coming to marketing data and talks with agency thought-leaders about the impact of 5G on targeting and ad personalization. Whether they’re in the home, wearables, extended reality or sensors out in the world, all of the new 5G-enabled devices and experiences will increase the amount of data, data inputs, precision and sophistication of data models.

Michael Stich of VML Y&R talks with Jake about how 5G-enabled data will power more personalized, predictive and prescriptive experiences, Lauren Stearley of Starcom Worldwide describes the emerging importance of probabilistic data in the 5G world. Barb Kielhofer of Spark Foundry shares some wise advice about marketing authenticity and talks with Jake about how 5G-improved data will improve ad personalization. Allison Schiff of AdExchanger advises 5G marketers to consider the privacy “underbelly” of marketing programs right from the start.

Guests:

  • Allison Schiff, Senior Editor at AdExchanger
  • Barb Kielhofer, Associate Director of Data Architecture at Spark Foundry
  • Lauren Stearley, Associate Director of Programmatic at Starcom Worldwide
  • Michael Stich, Chief Business Officer at VMLY&R

FIVE, The 5G Podcast for Marketers, is presented by the Emodo Institute and Ericsson Emodo.

Host
Jake Moskowitz, Head of the Emodo Institute at Ericsson Emodo

Producers
Robert Haskitt
Adam Kapel
Jake Moskowitz

Original Episode Art
Jeff Boese

Music
Dyaphonic
Small Town Symphonette


Transcript of Episode 5: The Shifting Shape of Data

“It’s hard for me to speak to you if I don’t know who you are. And I’m not going to speak to you in a very personalized way if I’m not 100 percent sure.”

Let’s talk 5G.

Welcome to Five, the podcast that breaks down 5G for marketers. This is Episode 5: The Shifting Shape of Data. I’m Jake Moskowitz.

Since we began this podcast series, we’ve looked at a variety of 5G-enabled experiences, devices and capabilities that will present marketers with a broad range of new opportunities. We’ve talked with experts about the new network characteristics which make 5G so different from previous mobile generations. All of those things together will have a momentous impact on marketing data. This is our data episode.

It’s clear that in a 5G world we should expect lots of important changes in the field of marketing. When we zero in just the data alone, all signs seem to be pointing to a few specific, significant changes.  Well, technically, it’s not a few. Its…Five.

  1. The first is the amount of marketing data. When content and interactions become immediate with 5G, consumers will engage more, do more, consume more. Ericsson estimates that mobile data use in the US. will triple in just the next three years. Plus, like we discussed in Episode 2, 5G will bring about an explosion of new devices, most of which will be sensors which will be providing a constant flow of data into the cloud. We’re entering an era in which the amount of data available will grow exponentially. More data can be more useful and valuable, but only if data vendors, partners and platforms and marketers create smart, new, 5G-ready data models, models that shape massive amounts of data into impressively specific opportunities to meet customers at well-defined stages of their journey. 
  2. That leads us to #2. 5G will elevate the value of probabilistic data. Today, in marketing circles, the term “probabilistic” is generally considered with synonymous with “weak,” “inaccurate,” and “guesswork.” But in a 5G world, the explosion of consumer data is likely to redefine that perspective completely. The better the data inputs, the better the output of those models. Deterministic data is likely to become secondary to vastly improved probabilistic data. And here’s the thing: when it becomes highly accurate and highly actionable, we won’t call it probabilistic anymore. We’ll call it AI. 
  3. The accuracy of location data from some vendors will improve dramatically. 5G uses higher frequencies, which means more dense networks of towers, or in other words, more cell antennas. So measuring where a device is between antennas will mean 5G cellular location data will achieve a level of scale, accuracy and precision never before possible. It’ll bust well beyond the limitations of GPS data, such as lack of accuracy indoors or inability to know which floor of a building a device is on, or a signal bouncing off a crowded street of high rises. That’s because cell tower data will supplement its current accuracy with far better precision. The CEO of AT&T said at CES that 5G could track location down to the millimeter.
  4. Expect lots of new sources of targeting data. As the number of consumer touch points grows, so, too, will the number of new data sources. We’re not just talking about new apps; we’re talking about new devices, many of which will gradually or eventually connect directly to the 5G network. This is a really important point. It means appliance companies, security companies, insurance companies and many others will hold the keys to new, valuable consumer insights and data. Mobile carriers will transport a lot more data than ever before because 5G means devices will send some data into the cellular network, as compared to today when most peripheral devices use WiFi and Bluetooth. To use or apply that information, marketers will need to forge new relationships, lean on new programmatic solutions, and develop new data sourcing strategies.
  5. Marketers will access data differently. Better access to high-quality data can help marketers connect with consumers on a more personal level. In the right-now world of 5G, targeting decisions will be make in real time. We tend to think of programmatic advertising as a real-time process. We even refer to inventory auctions as real-time bidding, but anyone who browses content in “real time” today knows it takes three to four seconds for pages to load. 5G will reveal that “real time” has been a misnomer in the high-latency world of 4G. Some vendors will embrace 5G by focusing more on real-time AI algorithms, building solutions around those algorithms, and moving those solutions as close to the consumer as possible. That will require ad platforms to implement edge-cloud computing infrastructure. That shift to the edge will enable lower programmatic latency and that will enable the true real-time data-driven decisioning required to keep up with 5G. It will also enable more efficient and effective optimization.

It can be super interesting and exciting to speculate about marketing shifts and opportunities in the 5G future, but as sure as there are changes ahead, not all of them are positive. Here are a few of the potential data-specific downsides.

Marketing leaders are going to need to update their marketing and advertising tech specs. Tech and data companies will need to incorporate key technologies that can handle and leverage the data that springs from 5G. When considering new solutions, think “automation, AI, virtualization, and edge-computing.” That’s a bigger change than many companies currently realize or appreciate.

And 5G doesn’t eliminate bad actors. It may actually invite more. The growth in overall traffic and usage will lead to a significant opportunity to improve targeting and personalization, and some data sets will become exceptionally rich and accurate as a result. That’s a pretty fair prediction. However, there will be plenty that don’t. Expect to see a rise of data vendors, poor-quality data, and misleading 5G data claims. Whether they’re in the home, wearables, extended reality, or beacons and sensors out in the world, all of the new 5G-enabled devices and experiences will increase the amount of data, data inputs, accuracy, and advanced data models. Forward-thinking marketers are going to have to plan differently.

Michael:

I agree. In fact, I think if we’re worth our salt as marketers, we’re going to understand these new formats and these new formats of data. But then, to your point, just as significantly, what to do with that data.

Jake:

Michael Stich is the Chief Business Officer at VML-YNR.

Michael:

And I think that machine learning is one where we are entering into a new phase where we’re not only looking back, we’re looking forward. We’re not only doing our own pattern recognition, we’re also getting some machines to do pattern recognition alongside us. I think that’s really exciting. That really dramatically accelerates not only what we know but what we might now in, perhaps, non-obvious ways. It will also help us with what might happen next, so moving more toward predictive and prescriptive experiences. It’s super exciting for me. As a marketer, there’s a whole lot that can be done there. But I would argue that it’s going to become both an opportunity and, increasingly, and obligation to get out in front of it.

Jake:

I’m lucky to be in a situation where I get to go around to agencies and talk about 5G, and one of the things that comes up is what to do about it. What should an agency be thinking about in terms of getting ready for 5G and that’s one of the things we talk about. Rethink who you work with and how those companies are built, what their infrastructures are like, and the companies that are AI-forward, that have access to the best input data are the best at the data science to built algorithms based on those inputs, and then have the technology to put those algorithms at the front of the product so that the product is based on the decision-making in those AI models. Those are the companies that are well-positioned to thrive in a 5G universe.

Michael:

That’s well-put. I think in general, the advertising agency and, for that matter, IT services model is already in a period of reinvention. I think what you just described is a good set of reasons why it must reinvent. It’s not only partnerships, but it’s also the kinds of services that marketers will desire and pay for in order for them to move towards it. My opinion is, there’s already so much new experimentation around new models for brands and advertisers in terms of what business they’re in and looking for new revenue streams. Just as significantly, so many new models on the agency side to try to think about new ways to monetize.

Let’s think about what kinds of capabilities 5G enables. 5G enables very rich video communications. 5G enables the ability to have very flexible, dynamic experiences as it relates to services to you. 5G also provides a real accelerant to personalized experiences so the content that you see is very uniquely tailored for you in very relevant ways.

Jake:

Lauren Stearly is Associate Director of Programmatic at StarComm Worldwide.

Lauren:

In terms of the long-term effects, I think data collection is pretty much going to be the biggest thing that is going to change, looking at behavioral patterns, creating those models, getting them into the platforms. I think we’ve seen the industry change a lot over the past several years and the companies that are able to thrive are the ones that are able to respond to the changes in the industry very quickly and be agile and create what marketers are looking for and make that available to them.

I think the buy side is typically the side of the industry that is catching up in terms of technology. I think that will change because all the data will be purchased through the buy side. In the past, with header tag and everything else, the sell side has been dictating the technology updates and the trends in the industry, and the buy side has been struggling to catch. I’m hoping to see the buy side sort of shine in this case, and I’m hoping to see people hire the right data scientists and get this off the ground. Because when the network is plugged in, they should have this ready to go.

Jake:

Since we are in the early days of 5G, and it’s going to be a roll out, a ramp up, let’s talk for a moment about when does it matter. When do we know that it’s time for StarComm to make significant investments? How much scale is enough to matter?

Lauren:

that’s an interesting question. In our industry, we usually wait around for there to be what we call a statistically significant data set. I think 5G will redefine the definition of a statistically significant data set. What are the behavioral patterns that we’re looking at for a consumer? How many times do they have to repeat a specific action in order for it to be significant? These are the questions we have to answer before we can even create a model. Then once we do that, we have to collect the data over a given period of time, and then we have to put it in a segment and send it to the DSP so that people like me can buy it.

That’s a good question. Again, it’s another thing that 5G will redefine for marketing.

Jake:

You mention that 5G could redefine what the right size is. What is it today?

Lauren:

Today we have big data, still. We have a lot of data and we have data scientists who have put it into segments and created models off it, but what I don’t think we have is a set of data points that we’re looking at. I think we’re maybe looking at one or two or three, but we’re not looking at behavioral patterns over a longer period of time necessarily. Maybe we’re looking at people who go in and out of a specific geo-fence, but once they’re out, we’re not really tracking them anymore. We just said, “Okay, this person came within the fence. Put them in this audience.” I think we have a lot of data now, and I don’t want to say we’ve been limited with what we can do to it, but we haven’t been able to create these stories that will evolve once 5G is part of the picture.

Jake:

As 5G takes hold, do you see it having an impact on how we view probabilistic versus deterministic?

Lauren:

Yeah, of course. The probabilistic will definitely be characterized as the data sets that we’re going to define through the 5G update. We’re going to be looking at the behavioral patterns of people and how they go about their day and how much time they spend at each place, which place they go to, where they commute, and those sort of things. While deterministic data will still be important, probabilistic data will become much more valuable than it was in the past when we were looking at people’s browsing patterns online.

Jake:

We misuse the word deterministic in the ad industry. We use it only to mean we didn’t change it, whereas the word is actually supposed to mean, “This data is true.” Here’s an example: Imagine I download a silly mobile game and it asks me to register my age and gender. Am I going to be honest? Are you honest in situations like these? And yet, if a marketer is using data directly from a registration like this, that data today is considered deterministic simply because it is unchanged from its source. The fact that this data is registration data means it’s even more highly regarded.

In the ad industry, a deterministic designation doesn’t require any proof, so probabilistic isn’t at much of a disadvantage, and it will only get better with 5G.

Barb Kielhofer is Associate Director of Data Architecture at Spark Foundry.

You being in data architecture, do you think that 5G will make consumers have even higher expectations of personalization and do you think 5G will enable marketers to be more personalized.

Barb:

Yes, it will definitely enable marketers to be more personalized. If nothing else, if we’re talking about consistent tracking. That’s our big issue right now. It’s hard for me to speak to you if I don’t know who you are, and I’m not going to speak to you in a very personalized way if I’m not 100 percent sure. It looks really off, and I think when you’re tracking somebody and you’re using something very specific. Specifically, I’m thinking of the Facebook ads that use your profile information. They say your name and where you’re from and they might have pithy things about it. It already you kind of the creeps. If that’s wrong, it’s really going to set you off and be odd. One, it’s creepy that you’re using information, and two, it’s wrong.

You really have to be very dead-on to have it work and and work properly. If there’s the insurance there that it’s going to be accurate and good, it will become more personalized, and to what level it can become more personalized will increase.

As the consumer, it’s hard for me to step into their world. I know I like things that are more contextually relevant to me, but I’m also looking for it. But I don’t want my time wasted. My time is very precious, and I think all of our time is very precious in this day and age.

Jake:

Except mine.

Barb:

Except for yours, of course. Why would you waste somebody’s time with something that is completely off base with who they are and what they need in their lives. You might, on the off chance, be showing them something they never even knew they thought they wanted. But the chance of that, I feel, is very thin. I think people are very grounded in who they are these days. We want to see things that reflect our lives and we need, and if you have to see advertising, I think you should want it to be very specific to who you are. If it’s going to be creative and interesting and you’re being shown something very exciting, if it is something that personally excites you, that’s just going to get you that much more jazzed up.

My marketer brain tells me that people should want this and want personalization. I don’t know whether or not people realize it’s happening. I think with the awareness of data people are starting to have their eyes open a lot more and are starting to pick out things and tease things out in a way that they’ve never done before. My hope is that, as we become more personalized and people recognize it, they’re not so creeped out by it. They feel good that the brand is trying to speak to them and not waste their time and make sure that it fits in their world and feels authentic to who they are. If we’re able to track them appropriately, it can be authentic to who they are. The scary part is where there are those giant misses where you’re really trying to be very relevant, and either your data is off, or your environment’s off, and it’s not landing. I think that’s almost the worst part. It’s less iffy when it’s really dead-on.

Jake:

It’s funny. Before you went back to it, I was going to go back to that point you made because I think it’s a really important point. You started really talking my language when you said, If I’m not really confident that the data is right, I’m not going to do personalized marketing even if I have the technology available to me. That’s a really important point.

The Emodo Institute published and article a few months ago on exactly that point. It basically says we can talk about personalized marketing until the cows come home because we have cross-device databases and we have dynamic creative technology, and we have dynamic copy. But let’s not forget the data that you’re personalizing based on. If that stuff is wrong, then it is way worse to do personalized marketing than it is to do non-personalized marketing.

Barb:

Agreed. It just comes off as very false. Like I said before, if anything doesn’t work in advertising, it’s being false. Authentic emotions, authentic behaviors, authentic creatives, authentic experiences, these are what people are going for in their day to day lives, and when you miss the mark on that it shows up so fast and furious. And there is no worse way to turn off a potential consumer than showing them something that totally misses the mark, is totally false, and not on the pulse. Our job is to continually to be on the pulse.

When you mess that up, I think you do lose people in a bad way. They know they’re being marketed to. That’s even worse. You hear people go, “Oh, those freaking ads that just keep following me. That thing that I was going to buy and now I can’t stop seeing.” They know it. They don’t know it in a way that they’re longing for that item, but when it’s authentic to them or they are engaged by it, or they are able to engage with it, it’s much more interesting to them. Personalization really does allow for all of those things if you can get it right.

Jake:

It’s clear that in a 5G world, people and businesses will access data in ways that they’ve never been able to before. That should raise new ethical considerations for advertisers and market researchers. The data may be available, but should you use it?

Barb:

That’s always the big question. What’s you’re agency’s perspective on it? What’s also your client’s perspective on it? Any partners you’re utilizing, what are their perspectives on it? Any of those could be three different answers. Whether you legally have the right to use it or not, any one of those partners or stakeholders say yes or no can completely change your way moving forward. I feel like that’s constantly evolving as the market constantly evolves.

We’ve even had conversations on specific targeting and beyond what’s legal, what do we think we have a moral right to do? Some of those conversations are now coming into play which is even a step beyond. Legally we have the right to use this data, we know, but should we use this data? It’s getting more and more complex. I think those conversations are going to continue to happen. Nowadays the first question everyone’s asking is, “Do we own this data, yes or no? Do we have a right to use this data, yes or no? Have people opted in, yes or no?” That’s really a first and foremost question beyond anything that we’re talking about.

Jake:

Do we have the moral right to use this data? That’s a great question.

Barb:

That’s the one that no one can answer. Or we’ll get many, many different opinions on it.

Jake:

Exactly. It probably depends on the entity, the circumstances. It probably changes over time.

Barb:

Yeah, and what you’re using it for. You could have the same data set, you might have the same target, and be using it for campaign X and it will feel authentic and morally okay, and not get the “icks” about it. And then you could have the same exact data for a different campaign and just feel like you’re exploiting people. I don’t think anyone likes to have that feeling. It’s not even just the data itself, its the context. It’s really, really nuanced.

Jake:

Obviously, there’s a lot of talk about how Millennials are so much more open and they’re so much more willing to share. I wonder how much of that is simply a function of age, and as Millennials get older and get into mid-career and stuff like that, will they change their mind about all the stuff they used to be so open about. Or is it built in to who they are because they were raised in that way in that environment and they will always have a different perspective? It’s interesting to me to think about how it will all end up.

Barb:

It’s possible. I think part of it now it’s so ingrained in our technology you would have to give up a lot of the personal comforts you have right now. Are people who have been born and raised to those comforts want to give them up? I know what it’s like to live without the internet. Could I live without the internet, now? Probably not, but I know what that was like. If you are growing up with a tablet in your hand from the age of three, I think it’s a very different experience.

If you think about even post-Millennial, kids who are being born now that are going to be able to tell their appliances what to do and be able to control everything without even moving a finger, will they be able to give that up once it’s been given to them, knowing that there’s a lot of trade-offs behind all of that. Every advantage we’re given, we’re giving something away. A lot of the tech is free for a reason. There was a trade-off to that and you have to understand those trade-offs. It’s partially about being more open and more sharing, and partially about it’s just about not wanting to give up all the good things just to have that level of privacy.

Jake:

Before we go, 5G means lots more data, and mobile operators, search engines and social networks becoming even smarter. Of course, with all the new digital consumer experiences, more data, accuracy and touch points, 5G will have an impact on privacy, too. Allison Shiff is a senior editor at Ad Exchanger.

Allison:

We really do need to think about the privacy implications of whatever you happen to be doing at the beginning because after that, everything is going to speed up. If you just jump in without thinking about how you might gather consent, or if consent is really even really the mechanism that you need in a particular case. It is going to be difficult to disconnect in a world where everything can be connected to a really fast connection, and not just at home. The convenience of that and the ability to watch a video wherever you are, all the entertainment opportunities. There are a lot of benefits for consumers, but if you’re going to do something cool, you always have to look at the other side and say, “When I flip this over and look at the underbelly, is this going to fly? Is it kosher?” And that is definitely something to do at the beginning.

Jake:

On the next episode of Five, could 5G act as the privacy tipping point? Will consumers get to have a lot more control over how that data is used and what they get in exchange for providing it to marketers? We’re going to talk about the privacy initiatives, advantages and concerns that are emerging as we enter the 5G era. Tune in next time for an engaging conversation with an expert in the privacy policies of over 60 countries.

Thanks for joining us.

The Five podcast is presented by Ericsson Emodo and the Emodo Institute and features original music by Diaphonic and the Small Town Symphonette. This episode was produced by Robert Haskitt, Adam Kapel, and me. I’m Jake Moskowitz.