David Ells: xAPI and How to Create a New Tech Standard

How do you create and establish a new standard in software? How do you use tech to help people learn? And are the people who administer educational systems like universities and corporations in an irreconcilable conflict with students?

In this podcast, David Ells, managing director of Open LMS, discusses his journey in online learning technology, touching on the creation of new standards like the Experience API (xAPI), which tracks comprehensive learning experiences beyond just course completions. David emphasizes the convergence of higher education and corporate training towards skill-based learning and more tailored educational paths to better serve the evolving demands of the job market. He also reflects on the challenges and opportunities in the adoption of new learning technologies, the importance of a flexible approach to education, and his vision for the future of Open LMS in spearheading innovative, modular, and adaptable learning solutions.

Key Takeaways

• David: “EdTech is a great place to be a developer because at the end of the day you're developing people. The end value of the thing that you're working on within education and development is a scaling up of people and an improvement of outcomes for individuals, for their communities, for their organizations.”

• David: “Education is sort of the great leveler of people and economies. And it's something that I am passionate and excited about in terms of, of, you know, what the internet's done to bring knowledge and information to people's fingertips who would otherwise be locked out. “

• David: The Experience API (xAPI) is… “a data format and a communication protocol for exchanging information about events that happen over time that are germane to the learning domain. … It has flexibility to be able to express a lot of event based data sort of ancillary to (learning and education). But it makes particular affordances for the idea of I took this course, I went through this module, I took this test, there's this digital learning elements.”

• Tom: “I love that vision of a resume as being this is certified proof of what I know and what I've done, and it's followed me throughout my education, throughout my career.”

• David: “ I think what's actually happened is that, like so many other things the open web became the corporate web. And so LinkedIn has really kind of become this place where you bring your credentials and those credentials. So for instance, from Coursera or a major MOOC like that you know, basically link back to their site in order to go through that verification step.”

• David: “A standard can define an industry. A standard is a way for people to agree how to how to engage together, you know, how to communicate, how to play together.”

• David: “Facebook was like really just moving into like major mature popularity at that point. And they were very much proponents of activity streams as a data format, which happened to be based off the same core structure. So really the kernel of the idea that became project xAPI was “Let's take activity streams and put it in the learning domain and give it some extra affordances that make it fit for that domain.”

• David: “… a standard without adoption is just some paper, you know. A religion without followers is just some crazy idea. A leader without followers is is no one, right? And so therefore (a standard is) very subject to network effects. Did you get half of the audience? Well, then you're probably going to get the other half because they need to come along for the ride.”

• David: “I think that that breaking up of the degree is an interesting idea that could that could take hold and be part of this, this bridge between academic and corporate…

  Can we talk about things as a mixture of skills in which you have these ancillary, cross-functional elements that are kind of (organized around) … job roles that are relevant in industry. Maybe it's guided by the concept of more individualized specializations that are able to be pieced together because you've got more granular ways to bring together cross departmental material into more coherent packages.”

Transcript

Tom: [00:00:00] How do you create and establish a new standard in software? How do you use tech to help people learn? And are the people who administer educational systems like universities and corporations, in an irreconcilable conflict with students?

These are some of the questions we asked David Ells, Managing Director of open LMS, an open source learning management system provider that helps millions of students get the education they need in both academic and corporate settings. David has been working in online learning for over 15 years, starting as a lead developer at Rustici Software, a pioneer in online learning. David tells us about working with the Department of Defense to develop a new standard to track learning the experience API that has applications well beyond tracking what courses people take online and can be used to create a complete digital record of everything you know and everything you know how to do. And he's our guest on this episode of the Fortune's Path podcast. David, it's great to see you. Thank you so much for coming.

David: [00:01:09] Hey. Thanks, Tom. Much appreciated. Yeah.

Tom: [00:01:12] So I met you when Fortune's Path was just a baby. I was doing work for Rustici, which was a learning company here in Nashville. So talk to me a little bit about how you got involved in learning tech.

David: [00:01:29] By accident. Maybe, as many people do.

You know, I think there's a funny thing that happens in technology. You go and get your computer science degree, and then you almost spin the Wheel of Fortune and end up in, you know, random industry X that happened to need a software developer at that time. And so so my number came up for, for learning and education haphazardly enough because I joined Rustici Software after getting out from MTSU with my master's degree back in way back in 2008.

I picked up so much more than development skills and industry knowledge about about the learning space. But, you know, I'd have to say, all that said to kind of reflect back on it because I've stayed there. Right? I've stayed there now for however, however many intervening years has been 16. What a wonderful place to find yourself, you know, because at the end of the day, I mean, it took me a while to come to a full appreciation of this, but at the end of the day, you're developing people, the end value of the thing that you're working on within education and development is a scaling up of people and an improvement of outcomes for individuals, for their communities, for their organizations. It's a it's a it's a meaningful thing, I think. Education is a very positive indicator for, for for positive outcome. You know.

Tom: [00:03:07] It's really interesting. So you see your role as someone who develops other people.

David: [00:03:14] Absolutely. Yeah. I mean, at the end of all that, that's that's sort of the meaning that all the nuts and bolts work finally connects to. Right? Education is sort of the great leveler of people and economies. And it's something that I am passionate and excited about in terms of, of, you know, what the internet's done to bring knowledge and information to people's fingertips who would otherwise be locked out. According to some, some traditional structures that sort of were a little more selective and aristocratic. Yeah. Yeah.

Tom: [00:03:53] I mean, it still helps to be born rich. And there's a lot about our success, which in my opinion is mostly luck. Who our parents are, what country we're born in, what gender we are. Lots of things that still have a major impact on our outcomes are outside of our control. And so I'll just ask as a hiring. So you hire people in your role or do you not do that anymore? You still look. You still hire people?

David: [00:04:24] Sure. Yeah. I'm still still, still part of hiring processes. Have have have been part of those through this year.

Tom: [00:04:32] The certifications mean something to you. Like if you get resumes and someone is, say, Microsoft certified or whatever, does that have an impact on your decision making?

David: [00:04:41] Slight. I it's certainly not a, not a primary factor, but without a doubt it's going to be experience. And, you know, it's I think we'll touch on this at some point, but learning is experiential. I think application of knowledge is the way that people really learn and develop. So information alone is just kind of part of the part of the equation, right? Getting to the point in which you are using that, acting that out, so demonstrable experience and success, you know, in a role or a similar role is is the most leading indicator. And then of course, how you how you show up in interviews, what your tone is. You know, what what some of your sort of values and principles are that shine through in, in some of those conversations.

Tom: [00:05:35] So I'm glad you mentioned experience because that brings up the X API. And I'm going to try to describe what this thing is. Well, I won't even bother. I won't even bother because you're expert in it and I'm an amateur. Tell everybody what is the xAPI?

David: [00:05:51] The Experience API. It's it's short for the experience API. And fundamentally, it is a it is a data format and a communication protocol for exchanging information about events that happen over time that are sort of germane to the learning domain. So it, it it has been created within that domain. It has flexibility to be able to express a lot of event based data sort of ancillary to that domain. But it makes particular affordances for the idea of I took this course, I went through this module, I took this test, there's this digital learning elements. There are also other modalities, like if I open this app and I spent 20 minutes on it and I finished these three levels of Duolingo and or I've completed, you know, the the fourth test on this flight simulator. But but fundamentally, it's about collecting all that information that's happening in a temporal fashion, right? All this stuff that's accumulating over time in order to build a picture over time of what learning development is happening in an individual level, at an aggregate level. People are using it in ways to study how people engage with content. So there's almost a little bit of content analytics or Google Analytics type type type factor to it. But really, yeah, it's, it's, it's blobs of JSON that are about actor verb object people experiencing things where a lot of those, those events are going to be learning learning based.

Tom: [00:07:37] So what I loved about, first of all, when Xapi came out, I was a giant skeptic. I was like, who's who's going to care about moving to a new standard? We've already got one that basically collects a big blob of data that you can't see or can't do anything with. You just care about the passing score.

David: [00:07:55] I would argue it's been proven half right.

Tom: [00:08:00] I was half right. Not bad. Half right. Yeah, well, I wish I was completely wrong because I, like you, mentioned a moment ago about the importance of experience in a job interview. And the xAPI, to me, is a way to capture experience. What have I done? Is that a reasonable description?

David: [00:08:19] Yeah. It's fair. I mean, there has always been a vision associated with xAPI, even from the very beginning of when we thought about it. That was the concept of in particular data mobility is really important. So, so in the world of SCORM one of the key limitations that motivated the research that went into xAPI, which was supposed to be the successor to SCORM. One of them was it only fits inside of digital learning within a web browser context. So SCORM itself is kind of defined as a JavaScript API and therefore is confined to that environment.

It turns out that that environment has been really versatile and lives with us on our phones, and has actually kind of been able to stand the test of time as a learning modality. The other thing is that Scorm, though it has a data format, has no communication protocol to get that information out in any kind of standard way. And so, you know, one of the things that we were trying to solve with Xapi was data mobility. And one of the very first things that strikes you about data mobility is let me take my data with me. And so there was always this concept of kind of a personal data locker, almost more of a structured LinkedIn collection of achievements, Certifications. Learning events that I could take with me. Badges that I could take with me and present as verified evidence of. Of things I've learned and have done to develop myself.

Tom: [00:09:59] I love that vision of a resume as being this is certified proof of what I know and what I've done, and it's followed me throughout my education, throughout my career. And you can go to as a deeper level of detail as you want. And as you say, it's totally portable. So I bring this data locker with me. And it can be investigated by any, you know, tool that you use to analyze data and come out and say, wow, David's a perfect profile given his history, given his activities, given his interests, the things that he reads, just, you know, for fun. Because all of that makes up what you know. Yeah. It's a fascinating.

David: [00:10:52] It's a fascinating space. There was for a little while, there was a pretty good take up of a group called the Digital Learning Consortium, which was IBM led. And they pulled together people from some pretty major names, Microsoft and LinkedIn and some, some of the big guys, some of the small guys we were invited into a couple of those meetings and started to develop some thinking in terms of verifiable portable credentials in which you could be issued something that had cryptographic signatures so that you knew that it came from whatever that institution was, and, and you could take it around and, and have that verified against third parties who wanted to want to know. Oh, okay. Yeah. You did you did do that. You did get that degree. You did get those credentials. I think what's actually happened is that, like so many other things the open web became the corporate web. And so LinkedIn has really kind of become this place where you bring your credentials and those credentials. So for instance, from Coursera or a major MOOC like that you know, basically link back to their site in order to go through that verification step. So if I go earn a certificate for a specialization on Coursera as long as I pay for it, they will give me and pass. They will give me a verified certificate which they host on Coursera. You can then go around and say, yeah, I earned it. And the way to know that that's true is basically to click through the link, go to the domain, which is, you know, owned by Coursera and say look, it says David L's name there. And he's earned the specialization in data engineering or whatever it might be. And the one really the one place to go to go parade that around is is going to be LinkedIn.

Tom: [00:12:46] I mean that's there's a lot to unpack in that story. I love what you said about how the open web became the corporate web. We could talk about that in a moment, but I want to talk a little bit about the establishment of a standard. So probably should have done this at the beginning of the conversation for people who haven't been hanging around learning tech for 30 years, but or 20 years. So a learning management system, I'll let you describe what is a learning management system and what's it for? And then why is this standard important for it?

David: [00:13:18] Yeah. The basics of a learning management system are are pretty quick and easy, which is it's a platform to deliver learning content to learners and track their progress against it. Everybody's interacted with one at some point along the way in an educational context. If you've ever used something like Moodle or Blackboard or Canvas in a college setting. And or you've probably used one in a corporate setting to do something as, as banal as as compliance training. Compliance training. Absolutely. Yeah, that's that's sort of, you know, it's funny. It's compliance training is sort of the the heartbeat of the e-learning industry.

Tom: [00:13:59] What everyone hates.

David: [00:14:02] It's kept it alive all this time for us to figure out how to do more interesting things with it. But it's still, it's it's still table stakes.

Learning management systems, you know, you've got concepts such as administrators, educators or teachers leaders. You've got students. You would have a course concept, you assign people into it. People are enrolled in courses. There could be modules that are involved with that course. And within those modules there's going to be content of all different types. There could be just written content, there could be links out to other things, there could be videos. But that's a learning management system, is to bring all those things together into a relatively sane platform to, to, to, to, to execute that, that activity.

Tom: [00:14:50] And then you have the standard you mentioned earlier, SCORM. Why is that important from to for a learning management system.

David: [00:14:59] So my dear friend and and mentor and just a great person in, in my life and career Mike Rustici, who hired me into Rustici Software back in 2008. He had a great saying, which was, you know, a standard can define an industry. A standard is a way for people to agree how to how to engage together, you know, how to communicate, how to play together. Think specifically about it's a pretty good analogy is the DVD standard, which is actually getting to be kind of a bad analogy because it's getting to be so old. It's like me talking about blockbuster or something. But, you know, think of a DVD standard, which was a way for all these content producers and all these DVD players to have interoperability to all come together and be able to work with each other so that you didn't have. Oh, well, Toshiba is only play universal DVDs and Samsung's only play XYZ kind of of of content. It's the very same thing for Scorm. It gave a way for content providers and learning management systems to play together so that content providers could provide one output that you could import and plug and play into any number of, of learning management systems that had been created, and vice versa.

Tom: [00:16:29] So you did Rustici develop the original SCORM standard or was he involved in that or. So what was the brainstorm to say? We need a new standard. And then why would someone try to establish that?

David: [00:16:47] Core base of Scorm, Scorm 1.2, which is still sort of the the the the main lowest common denominator of e-learning standards today was already in existence by the time Rustici Software sprouted into existence. And it it had been developed by a group called ADL Advanced Distributed Learning, which is the learning and development department for the US Department of Defense. That the context in which that standard got put together there gave it its initial legs to stand on because it got stamped into some rule sets about what you had to do to engage with the Department of Defense from a learning standpoint. And so it had this, you can call it kind of artificial demand that was set in place, but because of its relative scope and simplicity but effectiveness, it started to get traction out into the industry.

Tom: [00:17:50] So to get to Mike's point about a standard can define an industry. So the Department of Defense is an enormous buyer of e-learning. Early adopter, enormous market for e-learning. And they go to their vendors and they say do it this way. And then that begins to take legs outside of defense. It becomes a standard that anybody who's developing content starts to say, well, you know, we need to follow this because this enormous company has said, yeah.

David: [00:18:20] And, you know I'm not sure if we'll touch on this separately, but when I think of what really makes a standard successful or not, it's adoption, right? I mean, you know, a standard without adoption is just some paper, you know. A religion without followers is just some crazy idea. That's right. You know, a leader without followers is is no one, right? And so it's it's it's. Yeah, it's the practical manifestation of people following along with that standard. And so therefore it's very subject to, you know, network effects. Did you get, you know, half of the audience? Well, then you're probably going to get the other half because they need to come along for the ride. But that initial setting in place, finding a way to establish that core demand, that's not going to go away in order to give yourself a chance of that adoption was what made Scorm, I think, so successful. Because there were a lot of other options that could have taken hold instead.

Tom: [00:19:22] So what was the the idea at rest or what? Talk to me about like at Rustici, the story around we see the limitations of Scorm 1.2 and then Scorm 2.0, which I don't think anybody did anything with.

David: [00:19:38] Yeah. Yeah.

Tom: [00:19:39] And then like, you know what? This is really not good. Not what we need. And then how did you all go about developing a new standard.

David: [00:19:48] Yeah. So, so at that time and it's kind of weird to rewind to some of these years. You have to be.

Tom: [00:19:55] How long ago is this?

David: [00:19:56] This would have been around 2009 into 2010 that things started to really get motivated and moving on, on thinking through what a successor might look like. And it's interesting to rewind to those years because you have to be of a certain age to remember them. What was going on? Yeah. You know, mobile devices were really kind of just taking hold at that point. There were a lot of people who still didn't have them. And they were still quite emergent. So were there for like mobile apps that went along with that. And you may remember that, you know, all the initial mobile apps were natively built. So they, you know, since then we've moved to a place where so many things are built in sort of a web format, even if you don't know it, they're sort of being built with HTML and CSS and then just delivered as though they were they were native on your phone. But at that time that created this, this big awareness of an environment that was not the browser. And that was one of the things that that made people go, wait, Scorm can't go to this world. It can't go into these native apps because it has to be in a browser. And then, you know, some of the other things that were going on was bandwidth was expanding significantly. We haven't really experienced that again for a while now, right? Like, we don't think really about bandwidth and the next step change of bandwidth and what that's going to mean. But at that time, we were moving into a lot more video delivery than we ever had.

David: [00:21:45] And all that was like much more possible than it had ever been. And that created another one of these instances of, wait, there's going to be video platforms that do not really conform to this idea of import this package content into this learning management system, and off you go. And so there were there were dynamics at work that were that were really pointing towards some new modalities and some new opportunities for the way that people were going to engage learning. And I think it was that sort of inflection point that made ADL and the rest of us go, wait, maybe we should be thinking about what we need to put into place for this next version of of the future. And so during that time, I think, I think they still have this program, but it was it was a little it was a little more healthy back then, which was they were putting out contracts, research contracts, paid research contracts out into industry to kind of do co research and to bring those results back in to ADL, which then they could then use to develop and move forward with their initiatives. And so one of those was indeed for successor to score him and Rusty Software, with its relative place in the industry and credibility, won that research contract. And so we worked on that in a sense sort of commercially against this research contract to, to develop this next standard. And boy, what fun it was.

Tom: [00:23:27] Well, let's, let's talk a little bit about that. So the for the anybody who's gotten lost, I'll say you want to own a standard or define a standard because it allows you to define that standard in a way that's beneficial to you. We see this continuously in tech where things like Bluetooth is a standard. So how that gets defined has direct benefits for manufacturers, for tech companies because you get to say what it is like. A classic example was you talked about DVDs. I remember I'm old enough to remember VHS and beta tapes. And this this story gets told a lot because there was one way of playing a videotape or one method of creating a videotape and playing it back. Betamax, which was high quality. And then there was VHS, and no one knows. I don't know, someone somewhere knows, but I don't know why VHS won, but it one it was a much worse standard, but at one, and Sony was set back, you know, fairly dramatically as a result of losing that standard fight. So Rustici is in this situation, if you've won an ADL contract, the Department of Defense is going to pay you to think and you want to come up with something that meets their need, but also is beneficial to your business. And did you do that and how did you go about it?

David: [00:25:00] That's a really interesting question.

Tom: [00:25:03] Or just tell the story.

David: [00:25:05] Of course, of course. Yeah. You know, our Rustici software's sort of sprang into existence and had so much success because they became experts in this, in this emergent standard. That was in some sense, like I said, kind of already there. And so becoming an expert in that standard, didn't, you know, the the tone within the environment was not we own scorn. And so I think there was a similar mentality when we approached X API, which at that time was was called Project Tin Can. I don't think we walked into it going, we're going to own this or we're, this is, this is, this is going to be ours. Our name will forever be associated with this. And, and we'll kind of have exclusive, I don't know, you know, benefit to, to this thing. We really did, you know, approach it. And this is all to the credit of, of Mike and Tim, who were running Rusty Software. They they sectioned off. Really? It was a fellow named Ben Clark and myself. And sort of put us off to the side. And there wasn't a lot of mixing between what was going on generally at Rustici Software commercially and this research project and and so from that perspective, I think we, I think we legitimately approached it and did the initial 0.9 launch from the perspective of this is what we think the community needs. Now, that said, of course, you know, you're working on it.

David: [00:26:49] You're you have you have made the business on on the basis of being an expert in a standard. We obviously want to be a leader going into the next one. And so we're already thinking about, okay, implementations of this thing, the premier implementation of of a learning record store so that we're the first to market with that. And that eventually was what, what what fired off the idea of a reporting and analytics store, which became a watershed. And so, yeah, absolutely. There was always a sort of what? Step two and what step three and what step four. And and you know, lest anyone walk away thinking that it was sort of pure insider baseball or unfair, we put in a ton of work to to get the thing adopted. We put a ton of investment from a marketing perspective to actually get the thing launched off the ground once, once that spec was was actually written. And so from that, you know, going back to that earlier point of a standard is nothing without adoption. Yeah, that part of it, if you if you're trying to make that happen you know, you've got to put in, you've got to put in the time and the money to, to make it work. And it's really a marketing and political thing as much as anything else.

Tom: [00:28:10] What was the reaction? So you've got Tin Can is now a, a sort of draft standard out for public comment kind of thing. What was that initial reaction in the learning tech community.

David: [00:28:23] We had a good kind of elevator pitch for it, which I think let people orient to it very quickly, which was. And another kind of sign of the times, if you rewind to that time, activity streams was a thing. Facebook was like really just moving into like major mature popularity at that point. And they were very much proponents of activity streams as a data format, which happened to be. Based off the same core structure. So really the kernel of the idea that became project X API. Was let's take activity streams and, and put it in the learning domain and give it some extra affordances that make it fit for that domain.

Tom: [00:29:09] Can you quickly define an activity stream?

David: [00:29:12] Yeah. Activity streams are events happening over time with that same idea of an actor, a verb, and an object. In the social world. It's just things like Tom posted a cute picture of his pug or whatever.

Tom: [00:29:28] I have done.

David: [00:29:28] That, you know? Tom commented on Suzy's post, right? You know, that kind of thing. And so but but Facebook being Facebook very technical, they were putting out, you know, information and documentation and so on about activity streams as a standard format. And so there was already this air of. Okay. Event based JSON based standard. Oh it's activity streams in the learning domain. And so I think I think that environment was set up so that people went oh yeah okay. Right. That makes sense. Tell me more. Right, right.

Tom: [00:30:05] So you wasn't that you looked at Facebook and said, let's do what they do for the standard. It was more that you created a standard and then spotted Facebook as being a potential tailwind.

David: [00:30:18] Oh, it was it was as much the former as the latter. To be perfectly honest, we had seen what was going on with with activity streams and Facebook, and they were already being loud about it. And when looking at the learning domain and thinking about the problem set that I described earlier about these new modalities and platforms it was a matter of sort of putting those pieces together. And I have to give credit to Tim Martin, who was really the first person to suggest that idea of, well, what if you took this kind of event based thing and you just applied it into these domains? Could we could we have a format and a communication standard that worked over there? And that was the that was the basis that the research was, was developed on.

Tom: [00:31:04] And so you said you had you had a good elevator pitch initially, but it never my understanding is that it's never been super widely adopted. Is that correct?

David: [00:31:15] Yeah, it's been a it's been a mixed bag. It's been a mixed bag for sure. It has taken hold for sure. It has taken hold in a way that I think it's still going to be relevant within that, within that set of practitioners ten years from now. It may be bigger ten years from now. It may not be. And that's that part's really hard to tell. We thought that the adoption curve was going to move much faster than it actually did. In part, I think, because the browser context has proven a little more versatile than we thought. The idea that everything was going to be native on your phone proved not true. And you're able to execute things inside of, of of a JavaScript context in, in a lot of places that maybe we didn't anticipate. And and, and. Yeah, I mean, you know, there's sort of the 80 over 20 rule where you're still getting really that 80% value of or 80% of the use cases captured in, in a, in a relatively more simple method in terms of are people completing things and what are their scores? You know, that's still a key thing that people are looking out for. But that said having been through the journey of watershed, which I haven't talked about at all. Yeah. You know, people are using xapi in very interesting, very valuable and very fascinating ways. And I don't think that they're just going to stop tomorrow.

Tom: [00:32:57] So this probably may be a good transition into the watershed story. When I saw this concept of activity streams back when I was in learning tech and you all produced a little video that showed the, like, actor action object, and the light bulb went off for me. I was like, holy, you know, Holy shit, this is really cool. And I thought it was applicable to so many things other than learning and that to keep it, you know, with this anchor of of online compliance training, the dreaded compliance training felt like a real waste of what was a very interesting idea. And so have you seen the adoption of Xapi outside of learning? Are people using it to measure things other than, you know your resume?

David: [00:33:57] This is a fascinating topic. Yeah. So this in the reality of xAPI, this is sort of in some ways the, the crux of of what xAPI means. Could have meant might still mean, You know your comment about using it to track compliance training and how sort of wasteful that would be? You definitely were not alone in that thinking. And when, when we, when we sort of pushed it initially to, to get it off the ground, that was absolutely part of the narrative is this is not about the past, this is about the future. Ironically enough, when I really think about it, I think that that was a bit of a missed opportunity. I think that we could have should have sought to implement Scorm parity with Xapi out of the gate and said it's that and the future, as opposed to saying no, no, no, don't worry about that. Leave that for Scorm. This is for something else. I think looking back, I would have liked to see here's the bridge and here's the future. That said, yes. So without a doubt probably the most powerful thing about X API is its flexibility. And without a doubt, the absolute worst thing. The biggest liability for X API is its flexibility.

David: [00:35:32] You can use it to capture events about all sorts of things that are well outside the concept of learning, and sometimes you do find yourself at those boundaries going, this doesn't, this doesn't work, this doesn't feel right. Why is the actor, in this case, a system that's monitoring a daily number of site traffic? It does work for that. You can still use it that way. And certainly we have and, and in a perfect in a, in a perfectly honest sense, practically that that has become an important part of, of what it means to deploy X API. You know, in a live environment, it's being able to have a single format that integrates some learning data and some non-learning data. We've collected Salesforce data and sales numbers into Xapi data to put it alongside sales training. Right. And that's, that's that's net net, that's been a benefit. But it can get strange. Right? Because how do you have a it was designed for a learning domain. And therefore there are some really clear ways that you're supposed to apply learning events into that standard. You start to get much more variability and need for almost secondary standards or secondary rule sets on top in these other domains. Yeah.

Tom: [00:36:54] So that is fascinating. When I, when I first saw and we'll wrap up this thing about standards, because I want to hear your vision of the future for edtech. Where do you think it's going? When I saw xAPI initially, one of the things that also went off in my head was like, well, this is a pretty chatty standard. And so what is the criteria for what's worth recording and what isn't? Not all activities that I do are worth remembering whether I'm a processor on a server or whether I'm a human being who's studying Shakespeare. Not all of my activities add up to anything. And so it was you were going to be collecting potentially, like every transaction, I'll say. So, you know, Tom reads the to be or not to be sonnet or, you know, the web server starts a session. Any of those things could conceivably be captured by API. Somebody somewhere has to make a determination about what's meaningful. So is that what watershed does? And can you talk to me a little bit about you've collected all of this interesting data. How do you begin to make sense of that?

David: [00:38:08] Yeah. It's a very sound observation, too. And, you know, this ended up being a major story in data warehousing in general. We went into this, this, this this period of time that was all about big data and the idea of a data lake. And so you bring all the data in and you worry about, you know, sifting through it later. But for now, let's just let's get it in there. Let's get it in. And, you know, you ended up with data swamps instead of data lakes, right? This was the term that that started to float around and all this dark data, right where it's like, we've collected all this stuff and nobody's ever gonna have the time to even look at it or think about it, because you've got two data points sitting next to each other, where one is really significant and one is completely insignificant. Yeah. And so, absolutely, you know, we eventually ended up at a place where we realized, like, wait, one of the things that we need to do to add value here in this reporting tool is to make that delineation and provide ways to kind of filter out some noise and stick to the things because you if you've got, you know, I watched a cat video next to I completed a university course. Right.

David: [00:39:23] They can still just be two data points in the stream. And obviously they carry far, far different weight. And so developing a views into those, into those, those streams of data, you know, separated and be having some rules about how you sort of clean out some, some of the noise over time and retire that data because, you know, the other thing is that you may have fine data points that you do care about, but you probably only care about them for a certain amount of time. So if you're looking at specific interaction points within a piece of content, you're going to do a study about how people are engaging with that content. But you don't need five years of that data by any stretch, even three months of that data, because it's so fine, it's probably going to give you way more than you need, statistically speaking, for you to develop some views about it. And so, so developing some technology that would kind of take in those rule sets and go, okay, actually for this, for this stream of data, we're only going to keep about a month of it at a time. But for completions, we're going to keep those forever. You know, that kind of idea definitely ended up being a practical part of what Zapier was.

Tom: [00:40:32] And so did you all have some watershed? Did it have a services arm? I remember when I was working with you guys, you were working with giant companies like Verizon. Yes. Did you all have a services arm? Because I remember they had 50 people around a table trying to figure out how to get something done. It was kind of comical in a way. So they need someone to tell them what to do. And so if they've got their living inside of a data swamp, they need someone to help them make meaning of that swamp. Yeah. Did that become a big part of your business?

David: [00:41:01] Absolutely. Without a doubt. Watershed was and is in effect a sort of a combo product and services business as a whole. And a lot of that just becomes sort of what what to expect when you get into the reporting space. As soon as you start to engage with data, with different stakeholders, you get five people looking at some high level data points. You're probably going to get 30 different questions that could take you to another place. And so despite the fact that you might have a stable standard way to ask questions, let's say, and just to use more off the shelf stuff like power BI or Tableau or whatever you've got doesn't mean that those 30 questions are going to be known and answered ahead of time either. And so that process of engaging with the tool to continue to dig and develop views that you come back to showed up every single time that we engage with the customer. And so really what we were selling was a platform and then a services engagement to help use that platform to its best value. So tell me.

Tom: [00:42:14] Talk to me about Open LMS, where you are now. How that how you got there and what your activities are today.

David: [00:42:20] Yeah. So so so watershed was acquired by Learning Technologies Group back in the end of 2018. Beginning of 2019. And it was at that time that I moved to a managing director role there and sort of lead and grew. That startup was still felt like a startup for sure over the next five years. Having had a positive experience there and built some relationships with the people at LTG. There was as, as sometimes happens, just a shuffling of the pieces, some movement and reconfiguration of of kind of the broader organization. And the way to think about LTG conceptually is sort of as a portfolio company where they've got lots of business units inside of this, this bigger umbrella with a few shared services. And, and yeah, in that, in that grand shuffle of pieces, it looked like it made a lot of sense to take watershed in the state that it was and actually bring it back under a rusty software. So what's interesting is now it's sort of back with the family, quote unquote. And as part of that process to to try and start to optimize in some of the other business units, which which included bringing me over as the managing director for open LMS there in around November of last year.

David: [00:43:42] Open LMS is fascinating. It's it's certainly larger scale than, than than watershed was, and rightfully so, because it's in more of a base kind of long lived domain. It is a learning management system. It is that LMS concept that we talked about earlier. And in particular it is a Moodle based learning management system. So it's it's open source, which is a fascinating component of, of its story. And, and, and the offerings that we have. And it's very much related with higher ed, which was a space that I was traditionally not not in, you know, Scorm, in fact, and much of the time it rusty software was fundamentally in the, in the corporate space. So now so now being kind of more in higher ed and corporate sort of a hybrid market, if you will has has been a really fascinating view into the concerns of higher ed, the relationship of higher ed to corporate the changing face of higher ed, the admissions cliffs. I mean, there's, you know, there's some interesting pieces that I'm seeing that I hadn't seen before.

Tom: [00:44:52] So what are some of the differences in that market like? So tell me when when a corporate organization is looking for a learning management system, how do their needs differ from a university looking at one.

David: [00:45:03] Oh, so you know, this is changing over time. And and I think I think one of the most interesting narratives to watch within the broad learning domain that captures both is how they're, how they how the relationship between those two sides work and how they are moving towards each other in the way that they think. So I won't try to unpack, like the procurement process and the kinds of stakeholders.

David: [00:45:37] Picture, you know, usage patterns and so on. But conceptually, when you're talking about deploying this piece of technology for a purpose universities have long had had a concept of, of, you know, a specialization, a degree path, a major that is, is finding its way towards a skills based view of the world, which I think is taking more hold in the corporate space. And I think those two are trying to find each other right now, which is, you know, you've got a computer science degree, which is going to give you all this grounding about how circuits starting from how circuits work to assembly language to how you can create your own programing language, and you do all this very generalized stuff that gives you this, this really solid context to go do something that's more specific, which is language based or distributed databases or you know, cloud computing or whatever it might be. But, you know, the more you can cross that gap from the general to the, the some of the key specifics of the day you know, the more valuable as people are going to be on day one. Et cetera. Et cetera. And at the same time you know, the the writing is on the wall that they kind of have to do something there. You can't just, you know, charge whatever a degree costs these days, which seems to be just more and more and more every day. Right. You know, just a ton of money.

David: [00:47:18] And have somebody come out and feel like they now need to catch a lucky break to catch two years of experience, to actually be able to get the job that starts the career that they want. And so, so a you know the industry at large is going wait what about this thing in the middle. What about these MOOCs that are, that are skills based that are, that are that are doing credential programs. Sometimes those credential programs are backed by content from, you know, major universities. You know, these people seem to be coming out with skills that are ready on day one, even versus people who have gone and done this big investment of a of a major degree. And this is just of course, this is setting aside the whole idea of like, hey, learning how to learn is important. And the, you know, whatever the value of a generalized education, of course, personally, you know, and to your life. But but but but that said, you know, I think corporates looking, you know, in that direction going, okay, skills are a way for us to articulate how to have learning have more of an impact on, you know, business outcomes because we expect X, y, z, higher productivity, better communication, whatever it is that we need to train on. And universities are going, we really want to be known for students that are coming out and finding success, because that's how we sell to to more students.

David: [00:48:50] And in some cases, universities are investing in programs that may disrupt themselves. You know, almost on purpose and so are involved. You know, if you go to Coursera right now, right. You'll find content, great content from major universities, Harvard, University of Washington. You know, you name it. They they sort of know what they're doing because they have to sort out a sort of bridge to the future as well. And so, I don't know to try and circle back to the original question about how they're deploying those learning platforms. They're starting to look more like each other. They really still don't today. You still are going to have semesters and, and major student bodies that are flowing through and, and in particular here here's one analog that definitely rings true on both sides, which is monitoring for at risk. So who's at risk of dropping out of the university because they're not attending a class? Who's at risk of of dropping out of this organization because they're not engaged and they're not showing up and they're not learning and staying on pace. So so from that perspective, it's it's it's similar, but otherwise. Yeah, I mean, it's, it's you've got one mentality which is let's get them. Let's get them through these programs that we have. And the other sense you have, let's pull them to the place that we need them to be for the organization. Right.

Tom: [00:50:14] So this gets to a topic I did want to talk about, which is the the conflict between learners and administrators.

David: [00:50:26] It's interesting to frame it as a conflict.

Tom: [00:50:28] It's absolutely, in my opinion, it's 100% a conflict. There's an old joke about colleges. There's two administrators walking down a path, and one looks at the other one, and he says, you know, college would be a great place if it wasn't for students and teachers. And I think there's a lot of businesses that think of themselves as this would be a wonderful business if we didn't have employees or customers.

And so I, I'm, I guess a Marxist in that sense. There's to me an irreconcilable conflict between the institutions and the people that they serve or allegedly serve. So a lot of the technology is designed to make the life of the administrator easier. Not necessarily the experience of the learner richer. You know, and I think about in my own life, what are the situations where I feel like I've had the richest learning experience, while some of them have been with books, which are still a wonderful technology. And a lot of them have been with individuals, or they've been in things like projects where there isn't an expert who has the answer and is waiting for us to get to that answer.

So many learning situations feel like bullshit. They feel that there's, you know, guess what's in the teacher's head or sort of there's there's a predefined answer. And so when I think about, like, the match you were talking about and I'm not trying to crap on your business, I mean, but I've been learning a long time, and I'm like I have a lot of mixed feelings about it.

Tom: [00:52:20] But anyway, when you're describing, it's like, okay, higher ed to justify $250,000 investment doesn't want people coming out and going, “I can't find a job. I just dropped 250 grand.” And then, you know, sort of standard answer to that in the press is, well, you studied the wrong thing. You should have studied computer science. You studied history. Stupid you. Nobody wants a history degree. Oh, crap. In my opinion, that stuff is, like, totally wrong. Because you brought up the point about, like I need to be able to think. I need to be able to learn the problems I'm going to attack, don't have predetermined answers. And in my opinion, delivery of a product is always a cross-functional process. It's not, “These six computer scientists are going to do computer science stuff, and these two English majors are going to do marketing stuff or English stuff. And then the problem is going to be solved.” It doesn't work that way. It requires the collaboration of those mixed skill sets in order to solve a complex problem. There's a anyway, I could go on. But I given so that that critique that I, that I've given of the innate conflict between learners and administrators first. Do you disagree? And if you do, what is the right way to think about it?

David: [00:53:55] There's definitely a lot to unpick there, I think. I think you just brought together a few different threads into into one sort of tangly mess.

Tom: [00:54:09] It was pretty tangly mess. Yes.

David: [00:54:11] Which you know, the sentiment of which I don't disagree with. So, so a you know, let's let's just sort of first try and extract the technology piece out of it. Platform is a platform. So a platform is going to show up in a bunch of different contexts. And ideally the platform is going to fit whatever your intent is as it changes over time. Right? You know, I, I have experienced what is surely, you know, under the hood, a learning management system in a bunch of of Coursera stuff that I've done after school, and it's been great. I love it. That's why I keep going back to it. And it's been valuable to me. It's been a place of joy. It's been a way to expand my thinking and all that kind of stuff. And that has nothing to do with sort of the issues that exist with the institution of academia. Certainly in some of the ways that I think you've characterized it, that again, I think a lot of those things are true. There's that ivory tower thing that people have pointed fingers at for a long time for, for good reason. Right. So, so, yeah. So to set the technology side apart, I think, you know, having a platform in which people can, can bring people digitally and develop them over time in different contexts, fine. You know, it's kind of neutral. It's sort of the idea of, like, the internet's good and the internet's bad. And the conflict between administrators and learners. In some ways, I relate to the idea of the conflict between the IT department and the rest of the organization, which is under under kind of, you know, your, your, your classic mentality is I don't want to be bothered.

David: [00:56:11] We've got the system set up. You know, the users don't know what they're doing. Life would be a lot easier if we didn't have them, that kind of concept. And so I think that some of that is based off of inertia. I really don't want to be bothered. I really don't want to have to develop or implement or manage changes to a system because whatever, and because the learners need it, or because the teachers have expressed a need for it or whatever it might be. And so I think that's a natural source of, of maybe where that conflict has come from. I think one and not answer. I think one thing that's moving that in the right direction is SaaS. I think that having a cloud based system in which you don't have the burden of actually hosting the thing yourself and in which it is being developed by and hosted and provided by a third party helps ease that a little bit in the sense of, okay, no, that that can change over time without as much of a burden on administrators at all. The need for administrators kind of changes and shrinks a bit. And, and and you know, otherwise I would say. It's a big world. We can talk in generals and we can talk about, you know, sort of the 80% sense of things. But I, I will tell you this, that, that more so than than ever in my life or career, I have been exposed to really smart people who are forward thinking within the space of academics and how academic programs are put together and how the result of those academic programs.

Tom: [00:58:07] Where do you see it going then? So like, what do you what is the the positive vision for technology enabled learning in the future?

David: [00:58:16] That's a that's a, that's a big wandering path. And it I don't we haven't said the word. We haven't said the letters I yet. Yeah. You know.

Tom: [00:58:27] It's possible to get a conversation without saying this.

David: [00:58:29] If we may be, we may be late in the conversation to to to crack that one open. But, but, but I think that It goes back a lot to to the idea of can we articulate can we articulate development at a slightly more granular level than, here's the degree program that was designed ten years ago that we're still running? Can we talk about things as a mixture of skills in which you do have these ancillary, cross-functional elements that are kind of comprising around what we think of as maybe it's guided by the concept of job roles that are that are relevant in industry. Maybe it's guided by the concept of sort of more individualized specializations that are able to be pieced together because you've got more granular ways to bring together cross departmental material into more coherent packages that that you result in. I think that that breaking up of the degree is an interesting idea that could that could take hold and be part of this, this bridge between academic and corporate.

Tom: [00:59:51] Do you see collaborations between name brand institutions and name brand companies? There's some, I guess, sort of informal pipelines today about like, you know, everybody from Stanford goes into private equity or whatever. And do you see those formalizing a little bit more in the future?

David: [01:00:10] It's hard to say. I don't, I don't think I can, I don't, I don't think I can credibly speculate on it. I think that you're right. I think that, you know, by default there's sort of this concept of feeders, you know and demands there for, for professionals still very much right now. And so I think that that pull from industry is probably creating more of those relationships, you know, that we may or may not see.

Tom: [01:00:37] What do you want to do next.

David: [01:00:40] I mean certainly I want to, to see Open LMS succeed in, in the best way possible.

Tom: [01:00:49] What would that look like.

David: [01:00:50] Yeah I want to I want to see it grow. I want to see it grow. I want to see it be part of all of these kind of narratives and threads that we've, that we've covered over the conversation. I want to see it at the center of some of those really sort of innovative, you know, University 2.0 programs that connect you know, skills based learning you know, sort of breaking up the degree and the we haven't really touched on it, but sort of the, the reporting analytics pieces of that. So, so you know, taking that thing which, which you know, we're in a great position to do and that's, you know, that's probably part of the reason why it's, it's so forefront on my mind is we have an open source, pretty lean and mean adaptable platform. It really is. It's it's the whole mentality of the platform and Moodle platforms in general is this plugin idea. And to be able to create different footprints of the product experience off of a, off of a common platform. So we're well positioned to deal with changes of the future. I would love to see us kind of be a be a leader in the things that are emerging now that are that I think are going to become especially relevant.

Tom: [01:02:17] That's interesting. Well, David, it's been a lot of fun. I appreciate you coming and telling your story today.

David: [01:02:26] Hey, man, I hugely appreciate it. I'm happy to be here. I've really enjoy your reading, by the way. Or your writing, rather, I enjoy reading your writing.

Tom: [01:02:34] Thank you so much.

David: [01:02:35] Yeah. Yeah, it's it's a fantastic little newsletter. It's always full of interesting ideas and insights and everything else. So I was pleased as punch to come come talk to you today.

Tom: [01:02:54] The Fortunes Path podcast is a production of Fortune's Path. We help service and technology businesses address the root causes that prevent rapid growth. Find your genius with Fortune's Path. Special thanks to David Ells for being our guest. Music and editing of the Fortune's Path podcast are by my son, Ted Noser. Look for the Fortune's Path book from Advantage Books on fortunespath.com. I'm Tom Noser. Thanks for listening and I hope we meet along Fortune's Path.