Episode 1 | Data Governance & Retention for Online Learning
Welcome to EduCast 3000. It's the most transformative time in the history of education. So join us as we break down the fourth wall and reflect on what's happening. The good, the bad, and even the chaotic. Here's your hosts, Melissa Lobel and Ryan Lufkin. Hey there and welcome to this episode of the InstructureCast podcast. And we are very excited for this episode. We have a wonderful guest with us today who is a longtime friend of ours. I'm a true expert in the education data field. Kate Valenti, CEO of Unicom. Kate, welcome. Thank you. Great to be here with you both today. We plan to pick your brain on a broad range of topics from challenges institutions face today around data, how to best govern that data across all of the institutions, how do you bring that data together in powerful ways, and what the future holds for data -driven learning design, data -driven everything. That's one of our key topics this year. Great. Perfect. Well, before we jump in, we'd love to have our guests share a little bit about themselves. know you, obviously, Please give us a bit of your professional journey, your background, pets, things like that. Sure. I love those professional stories where people have these really like fortuitous little jumps from here to there. That is not me. If you're waiting for that, that's not the story you're going to get today. So I did not always know that I wanted to be in education technology, but I'm really happy that this is where I've landed. After undergrad, I started my career with Ernst & Young in management consulting, and I was introduced there to enterprise application integration. And that was a really interesting space for me. But I quickly learned that the rest of the firm, which was very road warrior oriented was not for me. And luckily I found Unicon and 21 years later, I am just assuming the role of CEO from our original founders. So I'm moving the company from founder led to founder inspired and trying to make sure that all of those really good core values that we have as a firm stay in balance, but that we also grow with the needs of our customers. I love your background, Kate. And as Ryan alluded to at the beginning, we've known you for quite a while and really have loved working with you over the years. On maybe a little more personal side, we also love to ask our guests, do you have a favorite learning moment? It could be one where you were learning something, one where you were teaching someone else, one where you watched learning. But do you have a sort of a favorite moment in your past around learning that has inspired you or continue to keep you excited about the ed tech space? Yeah, I have lots of them. You know, I was one of those kids who liked school. I was good at school. You'd find me with a scholastic workbook over the summer, you know, doing much to my brother's chagrin, right? Doing my work on the side. That was just enjoyable to me. One of the things I've really appreciated recently is I have two kids, and six, and have been able to watch their learning journey, especially in discovering reading, which is amazing when you kids go from not really being able to interact in the world around them, understanding signs and just the things that we take for granted. And then as soon as they can start to see what those things say, it just opens up this world for them. So that's been really exciting for me as a parent to kind of watch my kids do that. I will say that I think a lot about some moments in my own elementary middle school career where I really felt like I got differentiated instruction and I didn't know that that was what it was at the time. But now looking back, I remember a fifth grade teacher who challenged me to do this whole research project that nobody else in the class was doing, but I was bored and she saw that. And she let me kind of do this entire research project and she like met me where I was with the challenge that I needed in that specific assignment. And I really think back on that as that's one of those moments that she did what she could do. And these weren't digital environments back then. She did what she could do to challenge me and meet me where I was. And that inspires me to continue to help ed tech products and institutions to provide environments where students can have that individual sense of learning and where instructors can actually meet them where they are. I love that story. For you to remember it that way too, from that time in your life, really shows the impact that even some smaller things that instructors, teachers, faculty can do can have such an impact. What a great example. And even though we weren't digital then, it ties so beautifully into our topic because I'm sure there was data she was looking for to see that she needed to differentiate instruction. Even if it was just your behavior in the class or the work that you were doing, she was using data. and maybe not build online data that we all think that to do that differentiation, which is just so cool. I love that. listeners who don't maybe aren't familiar with Unicon, tell us a little bit more about the organization and what you do. Yeah, absolutely. So we're celebrating 31 years this year. So we've been around before EdTech was coined EdTech. We were working in it. I like to talk about us as a technology enablement firm. Technically, we're a consulting and services firm, but our job is to work with institutions and with EdTech. and providers to help accelerate technology, to help create good learning experiences. And we've been doing that across a really broad portfolio for, like I said, three decades now. We have about a hundred people and about 85 of those are client serving and they're all working on projects directly in support of and alongside publishers, ed techs, institutions to help them get their technology in order. So obviously you work with a broad set of education organizations to try to solve challenges with data, but are there some common challenges that are consistent across organizations? Absolutely. I think there are consistent, there are some trends that we see in our customers, which may be different than the trends we see in the market. So there's a nuance there in terms of what people come to a third party to help them to do. There are a lot of institutions out there who are doing really great things with data. They are doing very insightful, deep analysis. and informing what decisions they're making with data. The trend we're seeing in our customers right now around data is, I think I can fairly sum up with the essence of, we have data, we know we need to be using it, but we're not prepared to do that. And that preparation could be technical, it could be we just don't have the infrastructure to do it, it could be cultural, we just don't have the policy in place, we don't have common agreements on how to use It could be that all of sudden the environment just feels too complex. And so they know that they don't have their house in order. They know they should be doing more with what they have, but they need help kind of getting that house in order. And so that's a theme that we are seeing right now kind of institution as well as ed tech. You talked a little bit about this. We have the pleasure of being on a panel together recently and on that common agreements piece, you've talked about some of the customers that come to you looking for help. one of the challenges they face is everybody wants to use data differently in the institution or has different outcomes or trying to, and they're even thinking and defining data differently. What's driving that perhaps? And then how have you started to coach institutions on, you know, trying to get past that even first step of getting some alignment in the institution? I think what drives that we're at this place in the conversation is that historically The people who own the data were the people who use the data. So I'm the registrar, I own my data and I use my data and nobody else is coming for it. And so you can have a little bit of a walled silo in that case, but we live in a very different world now. We see all around us opportunities to have better insights through data that is joined between systems. And insights when we have people join their brains together to look at information and to look at that data and try to glean insights out of it. So we're moving from a, was okay to be siloed into, we have to think of it as our data, not my data, right? We have to think about data as a shared asset and what are the policies and decisions and cultural guides that we're going to agree to? in order to make best use of that, in order to leverage that data. So it's a cultural shift, I think. Yeah. We see that sometimes where there's a lack of consistent cross -organizational policy or even a misunderstanding of what the policy is. How do you help organizations try to organize and get on the same page across the board? Yeah, where would people start, maybe even, as they're trying to put that together? Yeah, so this this one. I've been thinking about this a lot recently. I think a decade ago when people decided that they needed to do something around data, they just threw technology at it or like, OK, somebody said we need to get better insights. And so we're going to do a data warehouse, the data lake, whatever we were calling it at the time. And that's going to fix all our problems. And it didn't work because it's not just a technology problem. And I'm a little bit worried that data governance. could go the way of the technology silver bullet, like, if we just get a data governance structure in place, it's gonna solve all our problems. And don't get me wrong, I'm a big proponent of data governance. But your question is sort of where do we begin in that conversation? And I think that it's critical that organizations come together around an actual initiative that they are trying to inform with data. So instead of saying, we're going to go do data governance. And this is from our own consulting. We work with customers who've said, we need your help building a data governance program. And it doesn't work when you build it for data governance sake. So where we begin is find an initiative that needs to be data informed and build your governance structures to support that. you have an enrollment issue or you have a student success initiative that you want to inform with data. Build a data governance structure that allows you to make decisions within the context of that initiative because that then aligns your stewards, your owners, your data council, if you have one, around an actual set of decisions, then sort of in this nebulous like we're kind of a council for data, but we're not quite sure what kind of decisions we're making. So that's our recommendation is begin with an actual institutional initiative that has funding, that has support, right? That has an actual measurable outcome and then build your structures to move that forward. And what that does for you is makes the second one easier, makes the third one easier because then you've already created some process around tackling something that's very specific that you can tie down. And then when you go to the next one, you kind of have a structure for doing it again. So that is a huge aha moment for me. I love that. And I even think about our own experience at Instructure and our listeners know I share some of the inside baseball too. You know, I think at one point in our history, we tried to do exactly what you're suggesting not to. Let's set up a data governance council or approach. And we got some work done, but there was a lot of conversation and I'm not sure it solved anything real. And then we've recently in the last couple of years had to really focus. We've always been focused on privacy and security, but we've really taken that up a notch as an organization. And that became the project for us to then think not only externally about how we're supporting institutions, but internally as well. And I don't think I ever like put two and two together, that it was, we needed the context of a project in order to get the kind of data governance practices we needed. And I love the piece you said about actually attaching a budget, making sure it's funded because I think a lot of those kind of die in the conceptual phase because they just, there's no money for it. It's a great idea. We need to do it. We didn't plan for it. We have no money, right? Exactly. And it tends to be something where you've got lots of different groups from different areas in the organization. And so it becomes expensive to just continue to have conversations over and over again without like an actual remit. What are we here to do and to solve? So the budget's important. Yeah. Well, and speaking of that remit piece, one of the things is we're talking to institutions, particularly about their learning management data or their learning data as a whole, they're pack rats. mean, there are institutions that have kept, you know, they've been our customer for 10 years, have all of that data live. In addition to they've got access to the data and an LMS that they might have been using before that. And it's starting to become, I think, one of these convening moments for institutions around. What do we do? Why do you think that is? Why are we keeping data forever? And what are some of the challenges behind doing that? Like being pack rats, particularly of learning data. You have a remote control in your house that you are pretty sure you don't own the device that that remote control goes to, right? I have some in this drawer right here. Why do we have iPhone cords from like three revisions ago? Because we might need it someday. Because it harder to sort through it and make decisions about what to do with it than it is to just keep it. We might need this someday. Our historical models could use it. IR's been saying they wanted it, Purely, we might need it someday. And that's risky, right? That's risky. And I don't want to catastrophize at all. I recently had a conversation with the college president, and I asked her, what keeps you up at night with regard to data? And she said, a breach. And so the more data you have, especially data that you're not using that you don't need, the more exposure you have during a breach. It costs money to keep that information. Even if you're in a SaaS platform, mean, you know SaaS platforms have to charge for the space that people are using. You're paying for it one way or another. You've got regulatory issues. You've got country by country, sometimes state by state. Guidelines around how long we can keep data and what we can use it for. GDPR and the California Privacy Rights Act both have items in them that are in the spirit of data minimization. We can only keep that information for why we collected it and for the time that we're actually using it. So you've got a lot of risks for just letting it sit. But I do think it's mostly just because people think it's harder to sort through it. education is so unique with the guidelines and the regulations that we face. from the obviously with Instructure and Canvas, we tend to be kind of focused on the teaching and learning side of the house. But this actually extends across an institution well beyond just student PII. I mean, this is administrative data. There's a lot of data. So how do you get everyone on the same page? if you have a data, some sort of centralized data function that has governance within their remit, that's a good place to start because those people can be conveners for the different areas of the organization that need to be involved in that conversation. And it's starting with who your data executive is and how that function has to look at data retention, data deletion policies within the context of governance. Then you've got connections out into like your CFO's office. So finance needs to be able to inform cost benefit for keeping or purging. You've got legal. Sometimes it's depending on who your learners are and what countries they come from and what states they come from. You've got specific guidelines and laws that you have to abide by. Those are two others. And then you've got folks from IT because as we know, it's not always easy to purge data from a system. There are products, some products that make it easy and some products that it's harder to actually do data deletion in a true way. And so you need people from those systems to come together to help to say, okay, if we're going to do this, this is tactically actually how we would get this done. Yeah. Well, and thinking about that in your comment earlier about start with a project. There are going to be listeners listening to this podcast thinking, my gosh, this is my institution. We need to be thinking about this. We need to be doing this. What are some of the questions that perhaps that group could start with? Obviously, of course, related to the project that they might kick off. But what are some things they need to be thinking about at the outset as they are starting to put in not only data governance, but data management and the practices around sort of emptying out the junk drawer of all of those remote controls and old cables. Yeah. So I'll take that into spots. think data governance and then data deletion or retention specifically. So in terms of data governance, the term is sort of universally hated. People don't like the term data governance, but we don't have a better term for it. But it sounds very much like it's about security. Data governance is really just about getting your data into the state that it can be useful. And what are all of the structures, the people, the tools that need to align in order to do that. So when you're thinking about data governance, you're thinking about first culture, like this is a shared asset within the organization. And so you're thinking about what are the organizational needs that you have that are going to be required in order to align people around that. And usually you have a champion, you have a data executive who can kind of set that as a theme or as a culture to say, we are committed to data as a shared asset. So think that's critical sort of at the top holistic level. More tactically, data governance explores things like knowing what you have. You'd be amazed at the data catalog processes. Data catalog actually as an ed tech product category is really, really trending right now. And people are trying to figure out what is it that we have and where and how do we trust it? So knowing what you have, being able to trust what you have, making sure it persists across the different ways that you might use it. Those are sort of all of those quality and like lineage concerns. So how do we set our culture? How do we know what we have? How do we trust what we have? And then how do we create processes and those connections that enable us to actually do really impactful things with that information once we sort of got that groundwork laid. On the retention side, I think the questions that these groups need to ask are things like, it's really a cost benefit, right? So what can I do policy wise to make sure that I'm covered from a regulatory legal perspective? It's about if I start to purge, am I actually going to see a cost reduction? Am I going to get to a point where this cost me less because It's not zero cost to implement policy, right? And then there's an opportunity cost in there, right? So what's the opportunity cost if I do purge historical data? Am I doing anything to cut off my analytics or my modeling? And I'm not a researcher, I'm not an academic, but I can imagine that at some point in the further back you go, the less relevant that data becomes actually in what you're trying to do just programs change, expectations change, and so it's hard to line that data up at some point. So that's a snapshot of some things that that group would need to tackle. Yeah. So we talk a little bit about data becoming more more important as far as data -driven decision -making, continuous improvement across an organization. And I love what you mentioned earlier where just getting all the data into a data lake or into the same spot isn't the end all be all, right? That can actually, without the right approach in place, that can be very difficult. What do you say to institutions to make sure that they can organize the data to ensure they can actually leverage it in powerful ways? It's useful once they actually consolidate it in the same spot. Yeah, and that hits directly to, think, the spirit of data governance is getting your house in order so that you can leverage data to be useful. So thinking about those items I talked through in aligning yourselves around those decisions and those touch points that need to come together in order to be able have agreement of how something's gonna be used, have processes for pulling different systems together that maybe we haven't crossed in the past. So there's lots of like tactical thoughts about how we're going to be in agreement that we're going to do this as an organization. But I'll go all the way back to like the best way that an organization can ensure that they are aligning themselves. to leverage data is to do that within the scope of an actual initiative that they need to inform. So again, not data governance for data governance sake, but if we say we have all of this information, and I think this is where a lot of institutions start, is but we have all this data and we need to figure out what to do with all these systems. We need that. And the reality is that if you pick an initiative and you say, are going to everything together we can to help us to understand enrollment decline and do something about it. And you actually build your governance structure to inform that specific need that is much more valuable than trying to do the, got all of the stuff from all the systems into one place. Like who cares, right? If you don't have the right decisions or processes to make use of that, the technology piece doesn't matter. So again, it's like, don't want to sound like a broken record, but going back to being initiative focused, I think is the best way for institutions to kind of get that crash course and understanding what decisions that they need to make in order to utilize that data to be helpful to them. Yeah, I'd almost say the way you've described that, and it's so perfect, it's backwards design for data -driven decision making, right? So for all of our educators, instructional designers, right? It's what do we want to solve for and how do we work words to get at the data we need for that and how should we be maintaining and using that? It's such a, I think it's so counterintuitive. I hear all the time. I know you do too, Ryan, the same thing you mentioned. We have all this data. We can do so much with this. I can, you know, and you get hung up there as opposed to what are we trying to achieve and how does that data get us there? Absolutely. Yeah. So I have to ask, and Ryan's probably going to chuckle at me because he's usually the one that asks this. Where does AI fit in all of this? It's the big topic, right? And it's got to play a role or maybe not play a role in this. Yeah, I was going to say, how long did it take us to get to AI? Exactly. So it used to be pandemic, right? You counted the minutes until COVID was mentioned. Now it's said COVID was a postcode? Unprecedented, I think, was the buzzword we used to love to listen for. So I think AI informs this conversation. I don't think it changes this conversation, but I think it does inform the conversation. When we're thinking about data programs, we're thinking holistically. We're thinking about all of the things that need to come together to build a really robust, rigorous, secure, equitable program. so AI is great, but it's just in the technology component. It does not replace the human in the loop when we talk about data programs. And for me, there are so many very critical components of data programs that are supremely human. Like we had a predictive modeling a decade ago and lots of came on the scene over, you know, like 2015, 2017. And they went out of my black box, here are the students that are struggling. And faculty were like, I'm not gonna do anything with that information, because I don't know how you got there. So I think we're going to see AI in two places. I think we'll see AI in our products, which we already do. So we have data storage products, we have data integration products that do really clever things with AI, and we'll just continue to see more of Like how can we do the things we need to do in ways that we use AI to help us to offload processing? So I think that will continue. I think we'll see AI in our processes in terms of productivity support. So if I'm looking at all of this data and I can set AI off on finding some patterns for me, so then I can look at that and say, all right, what do I think about this? Like that's a great use of AI is to offload the pieces of it that we don't need a person doing that. let the machines do that. But then ultimately, like I said, data programs come down to these like supremely human aspects, which are I need to take that information, I need to decide how I feel about that information. I have a friend who calls himself a data therapist, right? Like, how do I feel about what I'm getting back from this? And then do something with it. And that's human. And so I think AI is involved in our tooling and involved in our processes. that really feeds then the human in the loop to go into actually create the impact with that information. Yeah, well, and I think one of the things I love that you talked about the whole COVID, I think pre -COVID, is oftentimes it was sensitive to talk about data and there was a of would jump to the misuse of data or things like that. And I think one of the silver linings post -COVID really has been this acceptance of we can't always see the students sitting in front of us. We need to be able to paint a picture of them using data, right? And then we can extend that to our organization and really use data in more powerful ways. So I love that I hope that we're able to have these conversations more openly. And, you know, we've seen that much evolution in just four or five years. What does the next one to three years hold for an institution looking to use data more effectively? Yeah, I think that's going to vary wildly by institution. Like I said, at the very beginning, there are some schools who are doing this, doing it really well, you know, being able to surface great insights. And then there are some who are just trying to get started. And then there are those all in between. I think we're going to see more hiring of data leaders. I think we're going to see more executive positions that are over the data strategy. I also think that we're going to see more operational uses of data to make decisions like what programs are profitable and earning for learners in the marketplace, what they should be based on inputs. You know, we've got these new gainful employment, financial value, transparency rules coming out. And so I think the data is going to be key in informing the operation in those kinds of aspects. But I think it's going to vary wildly from institution by institution and those that will kind of come out ahead will be those who are able to kind of get themselves arranged around the shared asset. Like we're all in on using this data to do something impactful for the institution. Let's get ourselves together in agreement on that. And then the other pieces can follow. Those are just decisions about how those are tactical decisions. But I think the schools that really come together and see that there's immense power in the data that they have. And as long as they can unlock it, I think those will be the ones that really make strides in the next couple of years. I like that because institutions can find themselves anywhere on that spectrum. And I think underlying that is institutions will do more and they will be more thoughtful about the work that they do. And again, if our listeners get any like huge aha moment, mine was align it to a project and have that be at least where you start or to help you move along that trajectory. I'm curious your thoughts because again, you're a deep expert in this space. Unicom does a lot of work in this space. What didn't Ryan and I ask or what did we talk about that maybe we should be thinking about when we think about data governments, data management, the future of data? I think we covered a lot of our key topics. Like if we made a best practices list, I could sort of rattle them off. I think we covered a lot of Maybe I can share something that listeners, folks who are interested in this topic might find interest in it as a takeaway, something tangible that they could look at and something that I'm excited about that we're involved in. There's a group in EDUCAUSE called the Student Success Analytics Workgroup. And we have one of the leaders of that work group is one of our own, Dr. Tasha Dannenbring. And last year, the work group produced a framework and a rubric. around the maturity of student success analytics initiatives. And so I know it's student success analytics, it's kind of a specific slice, but there are a lot of components to this rubric that could be more broadly applicable. And essentially what the rubric does is it gives institutions the opportunity to sort of self assess where they are in their analytics program. So sub that in your data program. And it also gives you the view of what could be next as you grow. And so, Melissa, as you said, like, I can find myself on this trajectory from just starting to mature, but there's always room for growth. And I'm an operational executive at heart. I've been an operator a lot longer than I've been a CEO. And so having that roadmap to be able to say, this is what we're doing now. This is where we want to go and kind of seeing what schools who are doing it at the highest, most mature levels, what they're doing. how they relate, how they're doing things that are mission driven, how they're doing things that have data equity in mind, how they're doing things that have like a do no harm approach, which is big when we come to predictive modeling, right? So all of these considerations, this rubric sort of anticipates and you can look at it this spectrum and kind of see where you are and then see where you can go. And maybe we can put that in the show notes so people can link out to that. Yeah, we'll drop that in the show notes for sure. really excited about that and other models and evaluations like that that can help schools to really benchmark where they are and then where they could be going. I love that. We'll make sure to include that in the show notes. And that could be even a, I know we asked earlier, where do they start? Maybe that's a first start for institutions. Take a look at the rubric and start to figure out where are you? Even if you're an individual listening to this and you're thinking, I wanna go try to influence my organization or start having these conversations, this could even be itself a great conversation starter. Absolutely, yep, that's a great suggestion. Excellent, okay, thank you so much for joining us. I'm sure we'll have you back on at some point in future. You're an amazing guest and I'm sure we have more to talk about. so much for having me. Yeah, this has been wonderful and we'll make sure to link some more information about your practice as well, particularly the work Unicon does in case I have a feeling there'll be some listeners very curious about the additional work that you're all doing and where you might be able to help. Thank you. Absolutely. Thanks for listening to this episode of EDUCAST 3000. Don't forget to like, subscribe, and drop us a review on your favorite podcast player so you don't miss an episode. If you have a topic you'd like us to explore more, please email us at InstructureCast at Instructure .com or you can drop us a line on any of the socials. You can find more contact info in the show notes. Thanks for listening and we'll catch you on the next episode of EDUCAST 3000.
