The Platforming of Education
An Introductory Sidenote
In late 2007, Web browser pioneer, entrepreneur, and now venture capitalist Marc Andreessen wrote a really great blog post about Internet platforms. When I decided I’d write about education platforms as one of my Top Ed-Tech Trends, I immediately searched for it. I had only a vague recollection of what Andreessen had written. But I figured it’d give a good definition of Internet platforms and be an interesting frame for what I wanted to say about the “platforming of education." Because, ya know, it’s Marc Andreessen. (Link to Wikipedia entry, in case you don’t know.)
Slight problem: some time in 2009, Andreessen deleted most of his blog posts, including that one.
So a shout-out of thanks here to the folks who archived much of the site. And also, WTF Marc Andreessen deleting his blog?!
Defining the Education Platform
We throw the term “platform” around a lot in tech-speak, using it to refer to everything from software to hardware, from applications to operating systems, from websites to the Web and the Internet itself. In tech-marketing-speak, “platform” is often meant to invoke greatness or aspirations thereof: to become a platform is a goal — “the next Facebook,” if you will.
Marc Andreesen offered a good definition of platforms in a 2007 post titled “The Three Kinds of Platforms You Meet on the Internet”:
A “platform” is a system that can be programmed and therefore customized by outside developers — users — and in that way, adapted to countless needs and niches that the platform’s original developers could not have possibly contemplated, much less had time to accommodate.
Andreessen goes on to argue that there are 3 levels of Internet platforms — 1. Access API, 2. Plug-in API, and 3. Runtime environment — the differences depending on where and how developers run the code, as well as the amount of technical expertise and financial resources necessary to do so. Level 3 platforms are the best for developers, Andreessen contends, adding that “I believe that in the long run, all credible large-scale Internet companies will provide Level 3 platforms. Those that don’t won’t be competitive with those that do, because those that do will give their users the ability to so easily customize and program as to unleash supernovas of creativity.”
There are some flaws with Andreessen’s taxonomy here (best articulated by Fluidinfo’s Terry Jones) — least of which being his argument that Ning (the company he was then chairman of) sits at the pinnacle of Level 3. But as we think about education platforms — who is building them, who is developing third-party apps for them, why, and how — I still think it’s a useful technical framework.
But the programmatic aspect of platforms is just one part of why they’re important in education. There, they offer functionality that includes content, course administration, assessment, analytics, communication/collaboration, and external apps. Education platforms (and I should clarify here that I’m focusing specifically on Internet education platforms) are a fairly new development, I’d argue, and while the Web remains my favorite education platform, what we witnessed in 2012 is probably less about the open Web and more about the development of closed commercial platforms.
To say that education platforms are a fairly new development is not to say there aren’t major education / technology companies. Of course there are. Publishers. Software makers. Student information systems. Interactive whiteboard makers. But until recently there weren’t really platforms, certainly not Internet platforms in education — not the way in which Andreessen describes them at least.
Of course, until recently, there’s been a dearth of Web APIs in education — and that’s the technical pre-requisite for even Level 1 of Andreessen’s platform definition. In fact, as I wandered the rows of exhibitors at the ISTE conference this summer, asking them if their product had an API, it became pretty clear that they’re still few and far between.
As I wrote about APIs back in April, that means
Educational data is stuck in silos, something fostered by educational software – administrative and instructional – that makes it cumbersome at best and impossible at worst to move data in and out of systems. As a result, there’s lots of extra clerical work that educators and administrators have to do – recreating rosters, copying grades, downloading CSVs, copying-and-pasting, and so on. All because educational apps and software do not, as a rule, talk to one another.
The cost of upgrading schools’ various software packages and databases is high, But there were several major advancements this year in developing APIs to help retrofit legacy systems, including the launch of the non-profit Shared Learning Collaborative (SLC) and two for-profit startups Learnsprout and Clever. (See also: Top Ed-Tech Trends of 2012: Education Data and Learning Analytics for more details on these.)
While these three three help address some of the problems surrounding software integration, interoperability, and data portability, other API developments addressed sharing metadata (e.g. the Learning Registry), tracking learning experiences (e.g. the Tin Can API), and issuing digital credentials (e.g. Mozilla Open Badges). (Noodle Education published a directory of education APIs — along with an interesting metaphor to help explain them — over the summer. Mulesoft’s APIHub lists a lot more.)
Platforms can provide lots of benefits — for users (who get an integrated experience), for third-party developers (as Andreessen argues, who can easily build — and hopefully monetize - their apps on it), and for platform owners (who get to monetize it all). Rather than being stuck with the software as it was originally written, platforms allow functionality to be extended and customized. That's a big win for education, I'd argue, as it means teachers and students can have more personalized user experiences.
Of course, platforms also raise questions about data security and privacy (particularly in education, where the reg flag of FERPA is often thrown about).
There are questions too about the Terms of Services, something that the crowdsourced and crowdfunded TOS;DR project — yes, that’s a play on the Internet lingo tl;dr (too long; didn’t read) — sought to tackle this year by reviewing various apps’ terms based on things like data portability, anonymity, cookies, data ownership, copyright, censorship, and transparency about law enforcement requests (something I’d urge us to pay attention to for education sites as well).
But how the terms work for Internet platforms isn’t just a question for users; it’s one for developers too. And while this doesn’t neatly fall under the category of “education platforms,” the changes to the Twitter platform this year should serve as a cautionary note to users and developers alike. Like many ed-tech companies, Twitter has raised a lot of venture capital but has struggled to find a business model (A LOT of venture capital: $1.16 billion-ish).
Despite growing popularity — thanks to users and third-party developer alike — in 2012, Twitter seemed to close its previously open(ish) platform. It introduced numerous restrictions to its API for third party developers (e.g. rate-limiting, shutting down Twitter access for certain apps), arguing that these were in the service of “delivering a consistent experience.” In August, Twitter unveiled changes to its API that included updates to the “developer rules of the road” and outlining the types of tools that it wanted to see built on the Twitter platform.
Twitter said it would encourage development in the upper-left, lower-left, and lower-right quadrants and crack down on those tools in the upper-right — which is quite frankly where all the cool stuff happens.
Twitter can do what it wants, some folks argue. It’s Twitter’s platform. And that’s true. But its success is thanks to innovations from third-party developers and users. And it’s our data — all those tweets that we still cannot download.
I can't help but wonder: are there lessons for us here as we see the rush to “platform education”?
Who’s Platforming Education?
So who's racing to platform education? Below I've written about some of the major platform plays and players. But I have focused here on Internet platforms. That means that Microsoft, despite the reach of Windows, isn’t on my list of folks who are working on education platforms. I could make the same argument for Apple, I suppose, although its mobile platform in particular has helped spawn a sizable educational app ecosystem. Who else does that leave?
One of the most important Internet platforms is Amazon, and its Amazon Web Services — where you can run virtual servers and processors — make it one of the very few “Level 3” platforms. Amazon has made very few ventures into education (it launched a “Whispercast” service in October to make it easier for schools to sync Kindles and it did start renting textbooks this year too), but via AWS, the Amazon platform powers a lot of other companies. Coursera runs on AWS. Edmodo runs on AWS. Hack Education runs on AWS. (You know, all the movers and shakers.)
At one point I would have listed Chegg as a major education platform, particularly after it made a string of acquisitions in 2011 (buying CourseRank, Cramster, Notehall, Student of Fortune, Zinch, and 3D3R Software Studio) to extend the company’s products and services beyond simply “textbook rental.” But in 2012, I didn’t hear a peep from Chegg.
Other incumbant education companies have made it very clear that they see the importance of developing as an Internet platform. Learning management systems, for example, despite being on the Web haven’t traditionally been “of the Web.” But they now seem to recognize that they need to open APIs, be programmatically extensible, and be user- and developer-friendly. (This is one of the things that makes Instructure different from its competitors, I think. It’s been architected as an Internet platform from the get-go.) For its part, Blackboard seems to moving towards a platform (although traditionally it’s done this by acquiring the companies that had features it wanted to add rather than building a platform that others could offer those features on). The LMS giant had a very interesting 2012, with the acquisition of Moodlerooms and Netspot, two Moodle support companies (with an associated promise that it was committed to open source) and with the resignation of co-founder Michael Chasen.
Like the LMSes that are seeing the need to become platforms so too are the textbook publishers. We saw signs of this last year with Pearson and its launch of OpenClass, which as The Chronicle of Higher Education described at the time would be “a free LMS that combines standard course-management tools with advanced social networking and community-building, and an open architecture that allows instructors to import whatever material they want, from e-books to YouTube videos.”
Pearson has a very close relationship too with Knewton, which in turn offers an analytics platform that publishers can utilize to add adaptivity to their content delivery.
Another adaptive learning company Grockit took a different approach this year to (potentially) building out its platform. It launched Learnist in May — a “Pinterest for education,” if you will. (To my knowledge, Learnist doesn’t offer an API, so maybe I’m jumping the gun a bit on seeing this as a platform-play by Grockit.)
The startup that made a much clearer move towards becoming a platform this year was Edmodo. Or at least, that was the headline I gave the story in March when Edmodo opened its API to third-party developers. As I wrote then, “Using the Edmodo API, these companies can now connect their own apps to the Edmodo social network, meaning that they can tap into Edmodo’s badge functionality, feeds, assignments, grade book and so on. These apps are integrated with Edmodo, in terms of single sign-on, analytics, and – and this is key for both Edmodo users and app developers – roles (that is, who’s a teacher, who’s a student, what grade is the student, who’s a parent, who’s a principal and so on). According to Edmodo co-founder and CEO Nic Borg, the ability to integrate other tools with Edmodo has been the biggest request from users – some 6 million of them at last count, in some 70,000 schools.” 9 month later, “at last count” is now 15 million users. The company also raised $25 million this year, bringing its total investment to $47.5 million and prompting me to raise an eyebrow) about the potential for revenue and acquisition.
Following the acquisition of Wireless Generation and the hiring of former NYC Schools Chancellor Joel Klein in 2010, News Corp finally unveiled its plans for its newly rebranded education division, Amplify. And this is certainly a platform play as the company will offer digital learning tools for the K through 12 market focusing on three areas: assessments and learning analytics, digital curriculum (Common Core-aligned content for math, language arts and science), and content delivery. News Corp has also partnered with AT&T to pilot a tablet program in several schools. Wireless Generation is also helping to build part of the Shared Learning Collaborative.
Both Khan Academy and TED offer access to their video content via APIs (the former also offers user data, badge info, playlists, and exercises via the API). TED launched its TED-ED initiative in April — “lessons worth sharing” — and although teachers can download and remix some of those exercises, there isn’t really a way to do this programmatically (i.e. editing the lessons is manual). Are these two non-profits education platforms? Yes… maybe… sorta.
I think that “yes… maybe… sorta” also applies to this year’s biggest ed-tech trend, MOOCs and their role in and plans for platforming education. edX, Coursera, and Udacity haven’t opened APIs (to my knowledge), and while there are clearly efforts on the part of the former two to expand the content offerings by partnering with multiple schools, that seems to involve offline business development, not the “BizDev 2.0” Flickr co-founder Caterina Fake famously argued that APIs enable. But as George Siemens argued this summer, “MOOCs are really a platform” — or at least their proposed business models sure look that way.
Unbundle Education, Then Platform It?
There seems to be an interesting tension here with education platforms and the “unbundling of education.” We have heard a lot this year about “unbundling”: that many the services once offered by schools — content delivery, assessment, credentialing, research, mentorship, affiliation and networking, job placement — will no longer be monopolized by them.
But much like I argued in my Top Ed-Tech Trends story on MOOCs, I have to wonder if what we are “unbundling” from one institution (“school”) we are bundling and consolidating into another (into “the technology sector” broadly, and into platforms specifically). After all, content delivery, assessment, credentialing, research, mentorship, affiliation and networking, job placement — those are all the things that a good education platform would offer its users, right? That, along with the APIs for others to add apps and exchange data.
As Marc Andreessen recently argued, "software is eating the world." Will it "platform" the world as well?