EdTech and the ‘empathy-based’ technology vision with Kirk Gray of McGraw-Hill Education
Kirk Gray relishes his role as VP at McGraw-Hill Education. In this episode, Ledge talks to Kirk about his fascinating perspective on technology and education.
An EdTech veteran of 18 years, Gray has worked for a variety of startups before moving into engineering leadership. Having experienced the dawn of EdTech, Gray goes into detail about the early years and how the industry has evolved.
We hear how the rise of empathy-based technology promotes teachers getting the best out of their students, working with educators instead of against them. Kirk also talks through some of the difficulties of working in EdTech, specifically the ethical barriers surrounding data collection.
Ledge: Kirk, great to have you here. Thanks so much for joining us.
Kirk: Absolutely. Thank you, Ledge, for having me. Appreciate it.
Ledge: Would you just give a two or three minute background of you and your work so the listeners can get to know you a little bit?
Kirk: Sure. Kirk Gray. I live right outside of Denver. I’m currently a VP of Engineering at McGraw-Hill Education.
I’ve been at this software thing for going on 18 years now. I’ve managed to mostly stay in EdTech throughout my whole career, with stops at various startups and consulting shops along the way.
Started doing good old Java Enterprise development back when that was a big thing, and moved into more distributed systems, architecture things along the way. At some point, realized that I could really make more of an impact. As an engineer you get that, “Oh, I wish we could affect this. Why can’t I change this? Why can’t have the ideas instead of just implementing?”
One day I said, well, I’ve got to make that leap down to management. So, six or seven years ago I started doing the leadership thing and I have discovered that I really love it. I try to stay close to the technology, but it feels really great to be able to make a team, really excel at one, and bring other people up and along with you.
That’s what I’ve been doing the past few years and I definitely love the EdTech space. I love the feeling you get at the end of the day. The worst thing that happened is that maybe a teacher had a little better time teaching and a student might have learned more.
It’s been something that I can’t get away from, and don’t really want to.
Ledge: I’m looking at your background. You’ve worked for like every major player in the big… Enterprise EdTech is your resume.
Tell us some stories there, how that’s evolved. Eighteen years is a long time. I did a little work in the space myself. I know some of the super funded VC EdTech stories are not the full story of education technology.
I’d love to hear your thoughts.
Kirk: I think, everybody who experienced the dawn of EdTech when I was in college was like, “Wow, you can get your book but on CD.” That’s where it started. I think it kind of stalled there for a long time and there was a lot of the talking about how we’re going to disrupt the industry and we’re going to replace teachers with robots. That was this this really exciting thing.
What it comes down to in the end is, education is a pretty well established thing and there are a lot of people that really good at it. What I look at when I think of what good EdTech is, is how do we scale teachers? How do we make it so that teachers can do what’s really important, which is work side by side with students?
To understand who they need to work with and what they need to work on. For students, give them the tools to get the boring stuff out of the way and let them really focus on learning and doing.
My kids are Montessori students, and I think one of the great things about that is there’s so much self-paced work and we can really personalize what the curriculum is to the student for where they are that day in that given subject. I think that EdTech, when done right, can make that happen, not just in Montessori but in any classroom. That’s where a lot of my focus has been.
At Pearson, we were always chasing the student profiling. What is a student and what does it mean to be a successful student? Then, in this class what do successful students do and how can we nudge people to really achieve and get the right content in front of them? That’s the same thing we’re trying to do at McGraw-Hill, and that’s the real promise. It’s not to like replace teachers or replace the system.
When you hear, and we’re going to make these robo-schools… It’s great that people are interested and I think it’s great that money is flowing into this rather than how to build the next best Tinder, but we should also be realistic about what’s the best impact that we can have with our money?
Ledge: I’m interested to know if this is the kind of space where you would really hope to create intelligent technologies that are data driven.
Obviously, you can look at it and say, wow, this is a trove of data for ML and deep learning and AI and all these things, and yet I know from my own experience that the data input in education and the collection of any kind of useful anything was really the hard part. Where in other industries – you talk of healthcare, you talk to e-commerce or what have you – it’s easy to accumulate terabytes of data to learn from. In education it’s not.
I wonder, how do you guys handle that at the top level?
Kirk: Especially when you’re looking at K-12, you have myriad privacy concerns which are valid. I don’t know if you’re familiar with what happened with, but they had $100 million and they were going to record all these test scores and all this user activity and information, and they were going to share it.
It was a really admirable, laudable mission, right? The minute people started realizing, wow, this is so cool. How did you do this? They were, “Oh, it’s because we have all your data,” and they were like “Ah!” All of a sudden they were gone, in a year. $100 million. Poof. Gone.
We struggle with that all the time. You have to certainly, because you can ask an 18-year-old with an email address, “Is it okay if I collect all this and here’s what I’m going to give you?” It’s important to be upfront with what you’re doing. But in case, well, you just have to say we understand that we’re going to have to anonymize everything. We’re going to have to really make sure that there’s nothing super concerned about PII. We’re constantly vetting our solutions to make sure that we’re not doing anything that could go back to a person. Then you do your best.
There’s still a lot to learn if you look at data in a broad sense. There’s so much we can do. One of the great things about being a bigger company McGraw-Hill is that, we touch so many aspects of the student and teaching lives just by virtue of the surface space of our solutions, that we can really get a great picture.
Where, if you’re one of these startups, you’re touching this one little sliver. It’s great that you have this awesome ML capability but you don’t have any testing, any training data. Where’s your training data?
That’s a really a tough thing. You probably know from your experience selling into EdTech, there’s a high barrier. Getting to be in the circle of trust with the other EdTech vendors is a high bar.
That’s one of the reasons it’s nice to be at a place like McGraw-Hill where you’ve crossed those hurdles and now it’s really about actually solving the technical problems versus the business problems, I think.
Ledge: Walk us through… The engineering organization, obviously, of a company like that is huge. You sit on top of tons of different initiatives.
How do you think about those? Where does the leading edge meet just having to maintain a bunch of legacy code? How do you think about all that?
Kirk: The key is understanding what you want to be good at. Especially, you have companies Netflix and everyone is like, “Oh, I want to be like Netflix and I want to make chaos monkeys, and I want to be cutting edge.”
Then you have to understand, is that what people really need from us? I think what they really need from us is the thing that solves the teacher’s problem today. Frankly, a lot of our users are not highly demanding. What we really try to focus on is the empathy of, not like is this is whizbangest thing, but is this the least number of clicks?
When I worked at Schoolrunner, it was a startup, we’d go sit in classrooms in New Orleans. You’d watch the teacher use the system and you’d say, “I’m wasting 22 minutes of their day,” That’s a lot of their day when you think about what we’re asking them to do. Especially when you think of some of the Teach for America folks that are down there, and they’re working 80-90 hours a week.
How can I, as a designer of software solutions, come up with the most empathy based solution and not the most, “Oh, does it have …?” It should be 16 years old as long as the face of it, what they touch really works well.
We have to keep that in mind when we make tradeoffs. Instead of, is it the greatest thing that I’m so proud to speak on at a conference?
Ledge: Yet I’m sure you’re doing a whole bunch of stuff on the cutting edge as well.
What’s the exciting thing? What’s next?
Kirk: Like I mentioned before, we’re really trying to harness that data and build those pipelines such that we can merge all those different streams of data into something meaningful. I think we’re pretty close to doing something exciting thin there. We’re working with technologies like Spark, so we’re on the cutting edge with a lot of that.
We’re doing all the cloud things that you would expect, and we’re trying to use the right things where possible. We’re pretty aggressive about trying to rebuild our user interfaces to feel really desktop-ey, especially when we deliver content to the point of, “Oh, look my book on CD.” We really want it to feel immersive. We want to have it feel like you’re in the content. That requires a fair bit of strong UX work and UI engineering.
It’s really where we focus a lot of our effort today.
Ledge: How do you think about hiring for an organization this big? You have to have certain heuristics.
We’re obviously in the business of trying to figure out who are the very best engineers in the world. We’re trying to find and hire them, and make sure that they check all the boxes.
How do you guys do that? What are your heuristics at the big scale?
Kirk: I think there’s something to be said for hiring someone who wants to work at a startup. Isn’t always going to work at a place where you’ve got this many moving parts.
I think we’re lucky because we really are a mission driven company, and that there are a lot of people… Everybody has been through education. Everyone’s had an education. A lot of people have been to college too. They all know what they loved and what they didn’t love so much about it.
So people have this they have this investment already in what we do, most people want to feel good about what they do, so I think we have a leg up in recruiting.
One of the interesting things is, with a lot of the coding boot camps, if you look at the people coming out of those boot camps a lot of them are ex-teachers or people that were in education. It’s a really interesting overlap there, because you have these people that have lived it and now they have the skills to change it and they’re really excited about that. That’s a pretty interesting place that we can look to hire.
But really it’s just that we want people who can think about not just what’s cool for them in their resume but what’s great for kids and teachers. Being able to pick … the heart while also providing really high quality solutions.
We need things to work. It’s not something where we can AB test the thing and try it out, because this is fairly high stakes for these kids and teachers. That quality mindset is really important.
Ledge: So I can see – the listeners can’t see but I can – that you’re at home right now, and I imagine that you guys have some kind of opinions about remote work. Often that’s controversial. Some people like to have an office and some people are okay with vast remote workforces.
How do you think about that in your engineering org?
Kirk: It’s interesting. Here at McGraw-Hill, we have so many offices all over the country that on any given project we’re all kind of remote. So, whether I’m in my basement in Highlands Ranch, Colorado, or if I’m at the office in LA or Boston, I’m probably on a call with somebody from Seattle, or something that.
We’re friendly to remote work, obviously, but when you can co-link your teams I think it’s great. It’s nice. We try to be mindful and intentional about our team design so we don’t have the one outlier. Those are the … of remote work, where you have six people in an office and then that poor sap on the phone that’s like, “Can you tell me what you’re writing on the whiteboard?”
We try to do that, but really we’d be silly, especially in the job market today, to not hire the best person for the job where they are. If they need to travel a little bit to build that camaraderie and the shared understanding then let’s do that, and make sure that they’re the most productive person that we can find on the market. That’s the most important.
Ledge: Well, Kirk, great insights. Totally appreciate it. Great to have you on.
Kirk: Yeah, it’s great to be on. Thanks so much.