Monday, March 16, 2015

10 surprising things that education can learn from Netflix?

Netflix came from nowhere. Nobody saw it coming. Now a global channel that makes traditional broadcasters seem like history. Could something similar happen in the learning world? I think so.
Like movies and TV, learning experiences are delivered in roughly one hour, scheduled chunks; in schools as periods, in HE as one hour lectures. It’s a tired model – the teachers are tired, the learners are tired. We’re all tired. It's ripe for a player that produces the perfect storm of technology, cloud-based delivery, timewhift, AI, data-driven, content focussed, low cost delivery that Netflix created.
Young people are watching way less TV these days, TV is dying, and when they do watch stuff, it’s streamed, at a time that suits them. Education has to learn from this. I’m not saying that we need to replace all of our existing structures but moving towards understanding what the technology can deliver and what learners want (they shape each other) is worth investigation. Hence some reflections on the wonder that is Netflix.
1. Technology matters
According to the Head of Content, decision making at Netflix is 70% data, 30% human. That seems, in the long term, about right. Education and training operates at almost 100% human decision making, which is exactly why is not improving fast enough. Everyone’s getting impatient at the lack of progress – parents, teachers, policy makers, researchers, politicians. We seem to be forever stuck, ploughing the same expensive furrows. The answer to bad schooling is always more schooling.
The first lesson, is that we need to learn that technology matters in terms of delivery, data, recommendation engines, access, convenience and cost.
2. Coud-based streaming
Of course, Netflix as we now know it, was only possible when cloud-based streaming was possible on scale, and consumers had enough bandwidth to cope with streamed content. Cloud-based delivery is now becoming the norm for learning services. Content and management systems are moving to the cloud as this is now a necessary condition for success. Large adaptive systems need to be cloud based as they are only scalable on that model, especially if they are tracking and delivering services in real time, to large numbers of learners in a personalised fashion. The second lesson is scalable, cloud-based delivery.
3. Timeshift trick
Netflix, along with a couple of others, is reshaping the entire TV and movie industry by doing something that billions of customers want but the broadcast industry was slow to learn – like most things in life, people want to watch when they choose. Netflix outflanked both the TV and the movie business by being more intimate, with less of a presenter-led, patronising or blockbuster attitude. It’s addictive because you are in control and you have enough of a choice to fit your mood – movie, one episode, drama binge, comedy, documentary….. Timeshift, was the trick. Watch when you want. Free yourself from the tyranny of time.
Current educational institutions and delivery seem so like these traditional broadcasters, stuck in a scheduled, presenter-led world that is just so damn old-fashioned. Far too much learning is delivered in realtime, synchronous learning. It's time we recognised that learning can largely be done asynchronously. This is what online learning offers. This is unlikely to happen in schools, where the physical care of pupils is essential. Even there, however, there is room for improvement. But there are ample opportunities in post-secondary and adult education. We need smart, scalable online delivery.
4. Data drives delivery
Netflix famously turbo-charged their adaptive engine with a $1 million prize. They didn't actually use the winning algorithm but learnt from it and used something simpler. It paid off in spades. It is this subtle recommendation engine that makes it more than a library of stuff. It turns it into a living, breathing, personalised service.
That’s exactly what learning delivery needs – a more personalised service, one that always knows what you as a learner need and delivers the right stuff at the right time. This has been a key feature of Google (essentially an algorithm service), Facebook (algorithm driven ads), Amazon (algorithm driven recommendation engine) and now Netflix. AI is the new UI. There is every reason to believe that this has efficacy in learning, where knowing what a learner knows, doesn’t know and needs to know next, is the key to delivering efficient learning. This is the Age of Algorithms. AI is the underpinning technology that will shape things for the next decade and more.
5. Data drives content
It comes as a shock to people when they learn that Netflix gathers data, not only on what’s popular, cross-referencing actors with genres and sub-genres, as well as scene analysis. Netflix knows what you don’t like, when people tend to lose interest and when they drop out. This data informs subsequent programme choices, even content commissioning and scene construction. Sarandos, Netflix Head of Content, describes House of Cards as “generated by algorithm”. They calculated that the demographic that loved political thrillers also loved Kevin Spacey. This insightful content production is the result of a careful understanding of their audience, not the locked-down, play it again Sam, costume drama behaviour of the BBC. In a fascinating statement, Sarandos describes Netflix as providing a new kind of viewing “more like reading a novel – some nights you read two pages and fall asleep and other nights you stay up all night to watch”. That’s exactly it.
Learning has a lot to learn from this. We need to use data to build and improve content and delivery. This needs to happen in the creation and curation of content, as well as at the course and degree level. Everything needs to be seen as a system that improves with use. Education is a slow learner, let's make it learn fast.
6. Multi-device
Most watch Netflix on a big TV. The adoption of large-screen TVs, made the movie and TV watching experience that much more enjoyable. To hell with all that tiny TV stuff – bigger is better for movies, drama, sport –you name it. At the other end of the scale Netflix has benefited from the tablet, laptop, mobile and 'phablet' revolution, other outlets for watching content. They want it all. Predicting and accepting technical change worked to their advantage.
Listen up learning folks, learners want services to work across the board - on all devices.
7. User experience
The Netflix user experience is slick, easy to use, easy to navigate, searchable and personalised.
Compare this with most educational experiences. Most  are curiously disjointed, complex and inconvenient. It takes a huge amount of effort to understand and navigate the system or even a course. We must make it easier to find things, study things, get help, get feedback and get assessed. This needs constant attention, lots of A/B testing and a relentless attitude towards the learning experience.
8. Breadth of content
The scale of the content matters. Lifting movie rental out of the tape and CD business from which they came, was Netflix's masterstroke. Offering a large number of movies, TV series, documentaries and kids programmes paid dividends. Netflix pay others for most of their content. But they also make content themselves. Increasingly, you find yourself watching Netflix originals. This is no accident. Once you have recouped your costs the margins are enormous. Remember that their content creation is driven by data, huge amounts of data. That gives them the edge over other programme makers.
This breadth in both volume and type is important in a learning service. Too many institutions stick to the same old curriculum and qualifications. They need to be more agile and deliver a variety of courses, some short (Movies), some longer (box-sets).
9. Low cost
Netflix is a volume business. They keep the offer large and price small by winning customers. The more they customers they get, the more the price can remain low.
This approach is essential if educational supply is to meet demand. It is becoming unacceptable to pile debt on learners and governments by refusing to look at low cost options and now that expensive degrees do not necessarily lead to jobs that justify the original investment, it's game on. Don't look for nickel and dime cost reductions, look for massive cost reduction. 
10. Go global
Netflix is in 90 countries with 70 million plus subscribers. We badly need some big, global education content delivery. Brilliant, scalable content that teachers and learners love. MOOCs are getting there, showing what can be done but still far too long, drip-fed (like old TV) and over-scheduled (semester-long courses were never the real demand, just a feature of the old, supply model). We need subscriptions in the tens then hundreds of millions. Education needs a Google, Apple, Facebook, Twitter or Netflix. I’m tired of the corner-shop mentality, the attitude that teaching and learning has nothing ‘essential’ that can’t be scaled. We’ve come further than we think with Google (amazing search), Wikipedia (crowd-sourced knowledge-base with humble hyperlinks), YouTube (more analysis here) but we need to keep forging ahead.
Conclusion

Have I seen signs of this happening? Sure. The nearest service I’ve seen is Duolingo, which I've looked at in detail here, as it works on any device, is truly adaptive, personalised and sensitive to my needs as a language learner. More importantly, the very recent emergence of AI, namely adaptive learning in the EdTech area, shows that this is coming. There’s some very smart people backed by smart money that will make this happen. You won't see it coming.

1 comment:

Allison Rossett said...

Really smart, Donald. But still light years away from how most universities operate and dream.

You know I am a believer. http://www.allisonrossett.com/2014/05/03/whither-higher-education-to-the-opera/