Episode 68: Sheila Colclasure Discusses How Data, Digital Responsibility, and Trust Apply to Business Ethics

April 22 , 2020  |  By Matt Waller

Share this via:

Sheila Colclasure is the Global Chief Digital Responsibility and Public Policy Officer at IPG Kinesso. Prior to working at IPG Kinesso, she held multiple positions at Acxiom. Her most recent position was as the Global Chief Data Ethics Officer and Public Policy Executive during her 21-year career with the company.

Through focusing on ethically driven, accountability-based data stewardship and ethical AI, Sheila has deep experience in privacy, data ethics, and operational data governance as a strategic business issue and foundation of trustworthiness in the digital age.

Episode Transcript

[music]

00:06 Matt Waller: Hi, I'm Matt Waller, Dean of the Sam M. Walton College of Business. Welcome to Be EPIC, the podcast where we explore excellence, professionalism, innovation and collegiality, and what those values mean in business, education and your life today.

[music]

00:30 Matt Waller: Today, I have with me Sheila Colclasure, Senior Vice President, Global Digital Responsibility and Public Policy at IPG, Kinesso. You've got tremendous leadership experience in global digital responsibility and ethics and stewardship. I know that you had been at Acxiom for 21 years, and you were, towards the end of your career there, you were Global Chief Data Ethics Officer and Public Policy Executive, and you're still in the same area now, same topic. So, I would like to talk to you about that. As I mentioned to you before, the Walton College has started a Business Integrity Leadership Initiative, and that's a vast topic. But one of the newer topics, actually not so new 'cause you have a lot of experience in it, but it's still new to most companies and there's new things being discovered about it, are the ethical issues and integrity issues around data, artificial intelligence and these kinds of things. So thank you, first of all, for taking time to meet with me about this.

01:45 Sheila Colclasure: Well, I'm delighted to be here, thank you for including me today. And I think the topic of business ethics, holding the ethic conversation, how we operate ethically in business and with data and with our technology is probably the most important thing confronting business today, because we are, as I like to say, in the age of reckoning.

02:08 Matt Waller: You've been involved in this longer than most people, but now it's really, you've got a spotlight on it.

02:14 Sheila Colclasure: Yes.

02:15 Matt Waller: What's changed that caused that to, for it to become a real focus now versus say 10 years ago?

02:21 Sheila Colclasure: I think that we have undergone this rapid innovation, technology innovation, and it's become very mainstream. We've all heard about the Internet of things, where devices are becoming connected. The statistics are staggering. There are today, in 2020, there are 50 billion connected devices in the world. There are 13.2 connected devices per person in the United States, and think about all in our homes that are becoming connected. I've got a smart gate, I've got a smart thermostat, I've got smart cameras outside. We've got so many smart devices. We have the voice-activated in home agents. I've got smart collars on my dogs. Homes are becoming smart. I drive a smart car. Cities are becoming smart. What has brought this mainstream is this rapid innovation, rapid adoption, and it's all driven the expectation by people, that we will have digital experiences that are relevant, convenient, easy, meaningful, valuable to us, and businesses have pivoted to become that and offer that optimized digital experience. The megatrend for 2020 in the business world that I live in is that the user experience, customer experience has become more important than the product or service. And you create those amazing user experiences with data and technology.

03:53 Matt Waller: That's interesting, your point about user experience. There's a growing demand from companies for students that have some expertise in product management. And when I say, product management, I mean, the ability to interface with lots of different functional areas including technology, and create solutions or customer experiences, services, whatever it may be, that solve the problem for a customer. And it involves empathy mapping and a data-driven approach to understanding the customer, and then, using things like design thinking and an agile methodology to be able to create the solution. I mean, nobody was talking about this 10 years ago, but now, we have all these companies coming to us saying, "Would you get students to the point where they are "product ready" the day they graduate?"

04:54 Sheila Colclasure: Yes, yep. So the demand has driven this need for agile. Of course, in the Acxiom world, it is a data and tech company, and it's a very mature data and tech company and, of course, started by Arkansans in 1961 and grew organically. I guess at its height, it was $1.4 billion and it's changed and evolved over time. And they were in the midst of a digital pivot at 10 years ago, taking what was a very traditional computer-driven, data-driven processing to support a direct mail engagement to pivot quickly and become digital, to activate the digital channels. You know, 15 years ago, it was email and then they began to add all the other channels for internet connectivity and in-app experiences. So when you work for a data and technology company, which I have for a very long time, this notion of being agile, meeting the demand, working with data, using technology in a very smart way, it's been around for a while. It's mainstream now, right?

06:01 Sheila Colclasure: The idea with Kinesso, Kinesso is a marketing intelligence engine, we are supporting IPG clients with a very agile fleet, cutting-edge set of data and technology. A stack, if you will, that's built on top of the Acxiom expertise and capabilities, and is additive to that for speed to market. And again, we too, use inside of the Kinesso teams, we use the agile method to develop and deploy. I think it's important, the user experience as a megatrend is true. Think about if it's not easy and instant and channel-less, this notion that, I am, as a consumer, recognized when I go to the website and I log in, but when I go to the app, I'm not, and I've gotta go through another experience to log in and get recognized. That is very disruptive, that is friction and people don't like that. The heart of the data ethics notion and this digital responsibility gets down to a design ethos that puts the people at the center. That's where you start when you're trying to design a process that's ethical.

07:14 Sheila Colclasure: It gets you to the right place from what are you designing for, to design valuable experiences for people, and use data in ways that are ethical and responsible. So. In both instances, your commercial objectives and your ethical objectives, you have to put the person at the center and design out from there.

07:36 Matt Waller: Human-centered design.

07:38 Sheila Colclasure: I'm using the parlance, I've tried out the human word, which I love. I'm trying out the people word. At Kinesso, we've deployed a people care system and our designs are people centered. And people covers me as an employee, me as a leader, me as a consumer of goods and services, because we are all people. If you serve people with data, and you serve them with your technology in ways that they believe are just and fair, then comes profit. I don't know if you noticed the group, I don't know if you know of the group, the Business Roundtable out of Washington.

08:18 Matt Waller: Oh yeah, Doug McMillon chairs it.

08:20 Sheila Colclasure: Yes, he does. They changed their remit this last fall, their charter, and many of the CEOs including my CEO at IPG, Michael Roth, signed this new statement, and I'm paraphrasing here, the old remit was corporations exist to create profit for shareholders, to corporations exists now to create value for people. And that's people in their community, people in their professional capacity, people. We are experiencing intense focus on the way companies collect and use data. There's a slew of new laws, new legislation that's been introduced that will regulate and disrupt the flow and the use of data, because in every instance, it hasn't been used ethically and responsibly, so we've broken trust with many of our stakeholders. And we are at an age and a time when we've got to do better, and that gets to the data ethics and digital responsibility mission which supports business ethics.

09:24 Matt Waller: What does it really mean to be digitally responsible or irresponsible for that matter?

[chuckle]

09:29 Sheila Colclasure: Well, there are instances of that, we could talk about companies that have been digitally irresponsible, but let's focus on the responsibility piece. It's really four things, so data ethics, how you use data, is it the use of the data legal because that is a codified social norm, things as a society, we've already decided are right or wrong, we've written them into law, so it's abiding by that. There's other forms of that, such as contract law, promises made on the data. Then the third layer is really this notion of risk assessment or harms assessment, and then the top layer is fairness. So underneath digital responsibility, are you being ethical with the use of data? The next pillar is accountability. Are you accountable for the way your systems work, the way your technology works, the way your artificial intelligence works. Then we have safety. When you activate the data, is it a safe experience for people and for the corporation with the business? And then, the fourth pillar is security. And I began to unpack the ethical pillar, but there are functional things that have to happen under each of the four pillars. So ethics, accountability, safety and security, and those things, if you do that in your enterprise, then you become digitally responsible.

10:55 Matt Waller: I wonder, those four things, how aware are leaders of corporations and organizations? I wonder if very many of them know they should be paying attention to those four.

11:06 Sheila Colclasure: I think, until now, we've talked about a lot of that work in the security function and in the privacy function, and many times the privacy function gets pulled into the legal swim lane, which is a fine place. But what has happened, as we talked about a moment ago, we're at the intersection of business data and technology. So the way a business engages its audience, connects to its audience, plans its products and services, is all data-driven now, right? And you can only do that with technology. So we have this intersection. It means that the way businesses use data is fundamental to business strategy, and that's where all the decisions around data ethics and digital responsibility have to be made in the business strategy.

12:00 Matt Waller: So you're working for a marketing intelligence company, and a lot of the examples I've seen in the news of issues, have been around marketing, and so I can see why they would want someone in your position for this kind of a company. Tell me a little bit about the intelligence engine of IPG.

12:27 Sheila Colclasure: The intelligence engine is Kinesso, and it is meant to and designed to bring the data and technology functions together to fuel all of our clients' marketing and advertising engagements. Marketing is the advertising of culture. You are right, there have been some misdeeds and misconduct. I would argue that the instances of bad acts or inadvertent acts are few but potent, but by and large what brands want to do when they market, is they want to have an authentic, durable, trust-based relationship with their customers. I think that there's been some instances where marketers have operated without restraint.

13:11 Sheila Colclasure: So what we're doing right now is we're trying to really change the narrative and the thinking. There's a difference between fair, persuasive communication and unfair manipulation. So you've got to design your intention, your commitment, your process, your oversight. And your ability to let people participate with transparency and choice has to be ethically sound and has to be fundamental to your business strategy.

13:45 Matt Waller: Sheila, you mentioned this Privacy by Design concept, and that it's so important. What does that mean?

13:51 Sheila Colclasure: Privacy by Design gets at the idea that instead of bolting on privacy things that are good for people, you actually build your products that way from the design to the development, all the way through deployment. So it becomes, again, central to your design ethos. There's a discipline that has developed over, I'd say, 15 years. In the early days, of course, there was, Ann Cavoukian, which was at the time, the British Columbia Data Protection, official talked about that as an idea. There are others that have talked about it as privacy engineering. But it gets down to when you design your product solution service functionality, you ask the questions in the design phase. You ask, "Is this legal, is it just, or does it create harm? And if it potentially could, have we mitigated or eliminated that possibility?" And then, the fairness test. Are we delivering capabilities impacting consequence to people that they believe is fair to them, that they like, than those that don't have the opportunity to know about the data use and exercise control. I choose to opt out, I choose to have access to all the data that you've collected about me and that you're using about me. So you design that functionality as well into the product.

15:23 Matt Waller: So Sheila, why is all of this so important?

15:27 Sheila Colclasure: Trust is the fundamental of every human relationship. It's the basis of my relationship with my children, of my relationship with you, of you with your spouse, of me with the educational institution I graduated from, of me with my employer, of all of us with the brands we choose. The real takeaway on all of this is, at Kinesso, we're designing our solutions this way, because the people that work at Kinesso love it. Intrinsically, they wanna work for a place that they trust, they believe is doing the right thing, and they can see has a methodology to get there, that is sound and results in good outcomes for people. Our clients love it because they want a partner that's delivering data and technology that they can trust. So we have developed expertise inside the company, and we are designing our products this way because our clients trust us to do it and we want to be trustworthy.

16:31 Matt Waller: That's so true, and it's been true before technology, it's just at this point, we need a new way of thinking about it, because it is different. One aspect of this, you mentioned AI has become really popular again, and I know, I've even heard of stories of where artificial intelligence is really propagated by us and things like that, where there's been two people with the same income power, except one's a man and one's a woman, and a man gets five times the amount of credit when they're applying for credit, things like this happen. And a lot of that is because the algorithms are looking at historical data and projecting forward, not taking into account things the way they should. So I'd like to know what is meant by ethical use of artificial intelligence, and where's this technology going from your perspective?

17:29 Sheila Colclasure: Yes, it's been around a very long time, but it is certainly the future, absolutely is the future. And I think ethical AI, if we don't plan for it now, then we are all in trouble as a human race. I think AI has the potential to solve many of our greatest challenges as humans, everything from how we're going to address our climate changes to how we're gonna make sure our food supplies are safe and secure and distributed appropriately, to eradication of disease. Artificial intelligence is actually a continuum of computer algorithms. In my world, in the marketing world, you have very simple analytics that are like segmentation tools. And then you go up the continuum, and you progress through more sophisticated analytics all the way through machine learning, deep neural nets. And then you have different flavors of AI, it's on the sophistication end. You have narrow AI that's a very specific agent solving a very specific business concern, managing a very specific process. You have general AI, that is much broader and is self-learning perhaps.

18:25 Sheila Colclasure: And it's about the way these things are designed and deployed. AI has to be trained with data and data sets, and it all boils down, even the ethics, boils down to our human values. If you consider this with all these connected devices we've talked about, they both generate data on a continuous basis and they consume data. And the little agents, either narrow AI or broader AI, all this computer science that captures the data and makes it useful and informs my consumer step in my journey. It's as big as our imagination. But if you think about infinite data, it's no longer about the data, it becomes very quickly about how you train the AI. And in your example, the historical data sets that we need to train with, we need to think for a moment, are we training the data with balanced data? If our systems of credit already skew to males, then in the train of the AI, one of the questions, and the programming needs to account for the historical skew in data, alright?

19:45 Sheila Colclasure: This is very prevalent in AI that's being developed in the healthcare space to solve disease. Historically all of the medical protocols, the medical trials were by and large done on men of a certain ethnicity. If we only train our AI and it's program to say, whatever outcome this historical data yields is the answer, as opposed to there's already skew in the data, it doesn't have enough of these other categories of people, either by age or demographic or gender or ethnicity, then you're reproducing the skew, the bias in the AI. So this is a developing field but it's very, very important because the development is happening right now, and AI ethics is perhaps more important than data ethics, and we need to really start talking and designing that way now.

20:42 Matt Waller: What do you see are the most significant challenges with digital privacy?

20:48 Sheila Colclasure: I think one of the most significant challenges in digital privacy really is this notion of check the box versus operational accountability. The operational accountability is very, very difficult. Check the box policy, privacy policy stuff is easy. I said what I wanted to do, we've all agreed, I wrote it down, I made you sign it, we too agree, but translating that into the way the systems work, the designs work, and having sufficient monitoring and oversight of the way the systems sign is the challenge.

21:28 Sheila Colclasure: I'm reminded of, when I worked at Acxiom, and of course was the Global Data Ethics Chief, one of our very largest clients, the business buyers wanted to move ahead with some very robust digital capabilities. Their legal compliance, privacy team said no. And we went up and sat down in a room, Chatham House Rules, with my whole team, the business team of this client and their legal compliance, privacy people. And when they walked into the room, the legal compliance people sat on one side, the business people sat on the other. You could tell they were at odds. So we worked to the day and we charted all the issues. The takeaway was this: The legal compliance, privacy people had to be brought in to the design phase, and ask their questions about when you design this project or this solution capability, what is your intended outcome? Because we see this could happen, how are you designing to prevent bad things? How are you designing to enable the good things? It all goes to trust, trust by the privacy team, trust by the customer, trust by the people that are doing the work, and trust in the corporate brand.

22:48 Matt Waller: Thanks for listening to today's episode of the Be EPIC podcast from the Walton College. You can find us on Google SoundCloud, iTunes, or look for us wherever you find your podcasts. Be sure to subscribe and rate us. You can find current and past episodes by searching Be EPIC podcast, one word, that's B-E-E-P-I-C podcast. And now, be epic.

[music]