The legal and ethical concerns of data-driven marketing in technology

The legal and ethical concerns of data-driven marketing in technology

As our reliance on technology and data systems grows, so do the avenues through which corporations can harvest our information. In this episode, Ananya Bhargava interviews Joseph Ryoo, an Assistant Professor of Marketing at Arizona State University and expert in unstructured data analysis. Dr. Ryoo describes the techniques businesses use to collect and utilize data, the differences between what is legal and what is ethical regarding consumer privacy, and how business students can navigate data-driven marketing in the future.

Transcript

[INTRO MUSIC] 

Ananya Bhargava: Technology is the backbone of society, influencing how we communicate and consume information. As we become more dependent on technology and data systems, the channels through which companies can mine our data keep increasing. With ongoing innovation, it is crucial to find ways to stay informed on what data is collected and hold companies accountable.

I am joined here by Joseph Ryoo, an assistant professor of marketing at Arizona State University. His research primarily deals with unstructured data analysis and how companies can extract valuable information from unstructured data sources such as social media. In this episode, we’ll discuss the utility of data, the methods companies use to acquire and leverage this information, the distinctions between legality and morality in the context of user privacy, and the ways in which future marketers can navigate this evolving field.

Bhargava: Starting off, I know that whenever I have a discussion about data or anything of the sort with my friends, we’re always confused between different terminology that’s used. So could you walk us through the difference between data collection, data mining, and data analytics?

Ryoo: Yeah, I mean, definitely. I think I understand the confusion because there’s a lot of buzzwords being communicated in the industry, artificial intelligence, machine learning. Like you said, data mining, data collection. And I think that’s perfectly understandable. And I think the words that you just described are I think the keywords.

So what’s data collection? Well, data collection is the first step in analyzing anything, because that’s the step where we now are going to be collecting data. So if a company is scraping data from you, for example, what websites you visited, how long you’ve stayed on a particular website, any pieces of information, maybe your demographic information, maybe your address, maybe your education history, any types of information that a party is collecting from you, we can describe that as data collection. It’s not maybe planned in the sense that the more data, the better.

And I think that recent advancements in technology is helping with this. So rather than collecting small pieces of information, I think companies now are collecting as much data as possible. So as much data that they can get their hands on. And this whole process of trying to collect information from their environment, we can kind of describe that as this data collection process.

I think data analysis is where we finish collecting the data. So we have that data that in place. Now we want to extract information. And I think the key idea is that data is not information in the sense that without converting anything, the information that you have is, it may be useful, it may not be useful, it may be noise. And so how we can kind of filter out that noise and collect the information that we need for a specific problem on hand, we can kind of call that data analysis.

And data analysis can be done in many ways. Now you can run mathematical techniques on it.

Maybe you could maybe approach a more qualitative approach, maybe examine data one by one, maybe kind of qualitatively examining, without running any statistical analyses, maybe the insights that you might get as a consumer.

Data mining is kind of a tool, kind of a word that kind of integrates both of these things. So data mining is I think a more relatively more newer word, is a more recent word. So data mining is a bit of both. So you’re kind of collecting data from your environment, large amounts of data at a time. And maybe at the same time you’re analyzing data and saying, is this information valuable? Maybe it’s not valuable, then you discard the data, maybe you go move somewhere else, collect other pieces of data.

So data mining is, I think, incorporates elements of both. So when you see a person that uses this term, maybe this person is referring to both of these things, data collection and data analysis.

Bhargava: That makes sense. And that clears up a lot of my doubts. So I’ve seen quite a bit of your projects and the publications you worked on. And they’re pretty interesting. So my question is, what’s the most interesting project or publication you worked on and what findings did you find fascinating?

Ryoo: So I think the most interesting one that I have done so far is my project on spoiler reviews. So in that research, what I do is I analyze after movies released, the spoiler reviews that people write and post online. I examine whether they have a positive or negative impact for movie studios.

I think the famous example that I looked at in that paper was Avengers. When Avengers Endgame came out, a lot of the actors, I think Chris Evans, he went on YouTube and he was like, don’t spoil the movie for other people. But when I actually examined the data, so I collected the text of the spoiler reviews, and when I related to box office revenue, what I find is that they have a positive impact. And I thought that was very interesting. And I thought I relate a lot of that to my own personal experience.

Movies that I’m not quite sure whether I want to go see the movie or not in theaters, I tend to go online and read the spoiler reviews and see if it’s actually good or not. And I think that’s very interesting because it’s very counterintuitive. And this data analysis process helped me go against that common intuition, reveal the true insight about whether it has a positive effect or not.

Bhargava: That’s really cool. I was just watching an interview with the Brian Reynolds and Hugh Jackman for the New Deadpool movie. And the interviewer asked, where do you draw the line between a good spoiler and a bad spoiler? And they were both stumped by that question. They were like, yeah, sometimes it’s useful, sometimes it’s not. So that was very interesting.

Ryoo: It’s tricky because the actors, I’m pretty sure, the higher ups tell them to just avoid spoilers in general. Maybe they might not need to worry so much because spoilers can have that positive effect.

Bhargava: Yeah, that is interesting. And I’ve seen that in my personal experience too. So how would you sort of apply these sort of methods that you’ve learned to a more marketing or business-based context?

Ryoo: Yeah, so I’ll extend on that spoiler paper a bit further. So in that paper, what I proposed is that, so movie studios, movie marketers, what they should do is they should analyze what people are writing and posting online about these movies. And spoiler is just one small part of what people talk about online.

And what this paper kind of drives home is that a lot of times marketers, they focus on numbers specifically. So how many people have seen the movie, how many theaters are currently screening the movie in the United States and so forth. But that paper suggests that marketers should analyze this unstructured part of the data. They should quantify these spoilers that people are posting online and it helps them project future box office revenue, much more so than just relying on the numbers.

Bhargava: So how else would you say the role of data analytics plays into modern marketing strategies? What can you do with data?

Ryoo: We have a lot of data and I think that’s being enabled by the investments in technology that a lot of the companies are now doing. So oftentimes, just simply your browsing history for Amazon, for example, they now have the capability to record what products you visit, how long you’ve seen that specific product online and all of these things are data.

And what that allows companies to do is to help them make future predictions about what you are going to buy, what you’re going to browse, rather than just relying on vague beliefs or vague opinions of what you’re going to see next. What the data allows us to do is to arrive at a more scientific or more evidence driven prediction.

Bhargava: So how has marketing sort of changed and how do you think it’s going to progress?

Ryoo: So how has marketing changed? Well, I think the key aspects of marketing remains very stable. So for example, we talk about customer touch points. So when you’re dealing with a company, what are the specific points of your journey that you interact with that company? I think it’s easy to think about in examples.

So I’ll give the example of Apple, for example. So what are the points in time that you interact specifically with Apple? And we can think of these various touch points, things like you seeing a commercial on YouTube, that’s a touch point. Maybe you actually physically visiting an Apple store, that’s another touch point. You go to the mall, you enter the store, you interact with the employees. And all the way to the point of purchasing the product, maybe you have a product defect, you go back to the store, you interact with the employee.

I think this marketing process for Apple and how data analysis enables Apple to kind of strengthen this is that whether you have a positive experience or a negative experience throughout these touch points, they have the ability to record what specific detail or aspect that you are happy or unhappy about.

And what that allows them to do is for the next time around, improve on that if they encounter a similar customer to you. And if we can kind of generalize that to thousands and thousands of customers, we can see even though the backbone of marketing, the backbone of the customer touch point hasn’t changed, I think the efficiency and the effectiveness process of that process has changed and improved because now they have a solid foundation, the evidence that they have acquired, which part of that journey they should improve on.

Bhargava: Yeah, so that’s an interesting, usually when we think of data-driven marketing, the first thought is usually like targeted marketing, right? Because we use all that consumer data to make it more personalized, segment markets, things like that. But it’s interesting hearing that it can also be used for feedback.

You mentioned forecasting. Are there any other uses that the general public usually misses?

Ryoo: Maybe kind of like very surprising ways that people use data. I think one is your geolocation that I’ve very recently read about. So what your phone does is, for example, if you have an iPhone and you visit a store, what your iPhone does is it kind of interacts with the store and stores that data of what stores you’ve actually visited.

And what this allows the mall to do is to track which stores you visit and in what order you visit. And this is very surprising because it helps the mall and I think the brands in the mall to kind of coordinate their marketing efforts.

If they know which stores you’re going to be visiting, they’ll probably predict what future stores you’re also going to visit. So if you visit Apple first, maybe you’re also likely to visit maybe a store for Apple phone accessories and so on.

And so what you see is that a lot of these stores are just sharing data. And I think that’s very surprising because you would expect that they wouldn’t share data that you would expect these companies to compete against each other. But there’s this rise of companies that sell customer data to other companies. And so I think that’s very surprising in that the data that you share with one company is actually shared with other companies. And that this is, in one sense, very fascinating because it helps marketers to kind of enrich themselves with regards to customer insights.

But at the other hand, it’s kind of worrying in that, do you actually know how they’re using your data and to whom they’re sharing your data?

Bhargava: Yeah, so that does sound concerning. And so how do we sort of make sure and be aware of how that data is used?

Ryoo: Well, I think legally, do companies have that term of service that nobody reads before you sign up for a subscription or something? They present you a wall of text, ask you to read it, whether you agree, whether you consent or not.

And that’s usually the legal protection that they have. So if you later on find that you’re not very happy about the data that they’re collecting, how they’re using the data, well, what they’re going to fire back is that you agree to it in the first place. And so I think the easiest answer that we can kind of give to ourselves is that, read the terms of service. But on the other hand, we have to evaluate whether that’s fair, because often times a lot of these companies, they’re very essential for our social life.

For example, if you fail to consent to some of the conditions that Apple gives you, are you really going to be able to use your iPhone? And so I think there’s that moral dilemma as well, not just that legal aspect, but do we need to hold these companies up to that standard of, even though it’s legal, should you actually treat us customers this way? And I think this is a very recent issue.

And I think because there’s been a lot of data breaches very recently, I’m sure you’ve heard some of them where hackers, they steal company data, or maybe these companies, they lose track of the data and it leaks to the public. And I think because a lot of these issues are just becoming public, that these kind of problems, these moral dilemmas are being now discussed openly in public. And so we’re not very sure of how companies, what their position is moving forward. But I think at the very least, because we’re having these discussions that we’ll reach somewhere that’s not here, where the legal protection is all the thing we have.

Bhargava: So then how do you see this progressing and what do you think needs to be done? So obviously data is like a valuable asset when it comes to marketing especially. And we shouldn’t completely give it up. I think some data collection is justified and it can really help improve customer satisfaction and things like that.

But then there are also the issues you mentioned. So how do you see this going and what would you like to see happen?

Ryoo: Well, first, I think technology is advancing very rapidly. And so I think we’ll see more and more avenues for companies to understand us. I think when the pandemic happened, person to person, in-person meetings were not possible. I think everyone was on Zoom.

But I think from that, a lot of companies were developing virtual ways for us to interact. Oftentimes they had a separate platform where you had your own avatars and you get to walk up to different people and start one-to-one conversations. And this was all happening virtually.

I think what’s interesting about this is that the way that we behave, things that we talk about online, that’s going to be a new source of information for these companies. So the companies only had one aspect, maybe this publicly facing side that the companies had access to, how we broadcast ourselves to the public, what we purchase, what we talk about with regards to the product to our friends, to our family, but how we behave virtually.

Maybe that’s a more intimate side of things. Maybe you’re online, maybe you have a different persona, maybe you have a different personality. Maybe you’re sharing things that you wouldn’t otherwise talk about in the public. Once these kinds of information become available to companies.

On the one hand, it could be very exciting. They could be developing new types of services, products, maybe they could be helping how they recommend specific things to us. But on the other hand, it’s kind of worrying in the sense that the degree of privacy that we have moving forward in time is declining very rapidly.

And so again, it goes back to the issue of, I don’t think terms of services that we consent to in the beginning is not sufficient anymore. And that’s because as we use a specific company, use a specific product, use a specific service, more and more the degree that we share our information to has to increase. Maybe you weren’t sharing so much at first, but as you use, say your iPhone more and more, you’re being pressured to share more and more information.

And so this moral dilemma, again, legally there might not be an issue, but this moral dilemma of, is this correct? Are the companies on the wrong of doing things? I think these philosophical issues of morality, I think this is very interesting. And I think this is something that has not been talked about so much.

My research talks about what methods you can use to analyze data, what methods you can use to analyze unstructured data. And you see a lot of these researchers moving towards this front, machine learning, the amount of data that we can process. But I think what’s getting less attention, but what deserves more attention now is on liberal arts side of things.

The philosophical discussions that we’re not having is our privacy important. Is it valuable for us? If it’s intrinsically valuable, then it should be respected by companies. But because we don’t have these answers to these philosophical discussions, what you’re seeing are people like me that are making advancements in data analysis without any of these constraints that are telling us, hey, maybe this is not very ethical.

Maybe this is kind of out of bounds. Maybe this is something that you should get permission from the customers before you analyze anything. And so I think there should be this balance of, again, the philosophy side of things and the data analysis, the analytical side of things. But I think that balance is not there currently.

And so I think what’s going to be interesting is that we will see now the other side moving up. We’ll now see people becoming more knowledgeable. We’ll see people rising up, voicing their issues. And then hopefully we’ll be able to have those kind of discussions, kind of put a check and balance to the other side of things, the analytical side of things.

Bhargava: And do you think that if we do have those discussions and there is a solid philosophical and moral foundation for data collection, do you think that companies, other than appeasing the consumers, have any incentive to sort of curb their data collection? Because there’s obviously a lot of financial incentive in collecting data. So do you think that just consumer satisfaction is enough for companies to pivot?

Ryoo: So I guess it boils down to it should we be placing restrictions on companies. And at least we would like to think that companies, they have the freedom to conduct their business that they want, right? Because we have that freedom in this country. But complete freedom is not a good thing.

And so we do have the lawmakers that kind of set the boundaries of how competition should take place. And so what drives lawmakers? And I would like to think, I would like to be optimistic in thinking that it’s us that kind of sets things for lawmakers. And if we do have these discussions, if we do pressure them, if we do vote correctly, I think we should kind of have this trickle down effect where we influence the lawmakers and the lawmakers should then influence the companies.

But if we are satisfied, so companies are collecting a lot of data from us, they’re doing it in an unrestrained manner. And we as customers are satisfied. If that’s the case, if you’re not voicing our opinions, if you’re happy with the things, the way things are, if you’re not having our philosophical discussions of whether this is morally right or wrong, maybe the companies don’t have that incentive.

Again, if the companies are seeing that their customers are happy, why should they restrain anything that hinders how they operate their business? So I think it ultimately boils down to us. Are we happy with the status quo? Are we worried that companies are collecting more data?

These are not very clear questions. I myself am kind of split on some of these issues. But if we don’t have these discussions, maybe things are going to drag for a bit longer than we would like to. And so I think just becoming knowledgeable, I think that’s the very first step. And we’re becoming knowledgeable because of unfortunate events, again, because of the breakdown of data systems with Microsoft, because of data breaches, we’re becoming more knowledgeable because of these things.

So hopefully we’re going to be moving towards the right direction.

[Outro Music]

This is the marketing edition of We Mean Business sponsored by the Reynolds Center at Arizona State University.

Originally Appeared Here