Saturday, May 3, 2014

Mind the Quicksand: A Word of Warning to EdTech Investors

I read about a new EdTech startup this morning in TechCrunch. It's called Galxyz. I started writing a comment on the article page itself, but then it got really long and kind of "meta." So I thought it might make an interesting blog post.

Here's the description of the company from TechCrunch:

Galxyz (it’s pronounced “galaxies”, and it’s a nightmare to spell) is building a science-focused game for tablets and smartphones. And the company really is just focused on one game — Rashid [the founder] described it as “an intergalactic science adventure,” one that kids could potentially play for years, battling a villain called King Dullard across the galaxies. As they do so, they’re also learning about science at their own pace.

You can check out their epic promo video here (I would embed it, but that might violate copyright).

Sounds interesting enough.  Although an antagonist called "King Dullard" is a little on-the-nose, if you know what I mean.

What's the new angle on education, I wondered?

Apparently the idea came to Rashid as he saw his own children playing educational games, which he found lacking in several ways. He said they weren’t engaging enough, the content wasn’t deep enough, or they required the parents to get involved in order for the kids to advance. That second point is why he’s focusing on a single title — so that kids can just keep playing rather than running out of material after a few weeks.

Wow - it seems so obvious once he points it out.  I wonder why this hasn't occurred to anyone before? 

Anyone who follows this blog knows that I believe it is entirely possible to make education orders of magnitude better.  And that I applaud any authentic effort to try.

But from where I stand, this has "Titanic" written all over it.  

I am confident that producing truly effective education at scale is harder than any business or engineering challenge the good people at Galxyz have ever faced.  

This isn't about them, though.  I wish them well.  For purposes of this blog post, they represent just the latest in an endless parade of investor-backed EdTech entrepreneurs who seem to have almost exactly the same creation story.  I'm not upset about it.  I'm more intrigued.

One aspect of education that fascinates me is that to outsiders it always looks so *simple*.

As simple as running across that wet, sandy patch of ground in the jungle to get to a fabulous treasure twenty feet away.  But's hard to believe no one has run over there and taken that treasure before...hmmm...don't you wonder why? (Hint: That's not sand.)

What follows is my prediction for this venture.  

To be clear, this is not what I wish for them.  I want truly better education more than just about anything, and I'm indifferent as to whether it comes from a for-profit or non-profit venture, as long as we get it.  And it's so close I can almost taste it.  

The prediction is based on a pattern that I see play out over and over and over and over and over again (that's right - five 'overs').  

It looks like this...

Day 1: Outlook bright, mattress so overflowing with cash I can't climb onto it so am sleeping on the couch, burn rate 7, feeling great, gonna make history and be a hero by fixing education because I'm...
(Choose all that apply)
  • A product of an education system, and therefore I know how to educate...more better...than anyone
  • More funner than anyone in education today
  • In possession of more revolutionary technology than those other people
  • More smarter
  • Mo' richer
  • A bit confused about the difference between money, intelligence, and expertise

Day 30: Epic teaser promo video produced, first prototype created and friends (who are also employees, or hoping to be) are saying it's definitely the next big thing - time to pick up the pace of hiring animators, writers, and engineers!  Mattress so overstuffed with clams that I rolled off it last night.  (Ouch.)  Outlook blazing, increasing burn rate to 9.0.

Day 180: A bunch more concepts storyboarded and prototyped, things seem to be moving right along.  Still, things are getting a bit confusing.  We discovered that narratives are inherently linear and learning appears to be inherently nonlinear, so that's creating some "design challenges" (so-called).  Can the kids go through the same narrative ten times if that's what they need in order to understand the concept?  But wait - even if they will tolerate that, isn't one definition of insanity "doing the same thing over and over while expecting a different result"?  Should we have multiple narratives for each concept? Wow - that would be expensive.  Let's stick with the one narrative.  Should we at least present each concept in multiple different ways to help kids understand it? But will all those choices confuse them?  And what other ways are there to present the concepts within a single narrative? This is taking longer than we thought - better hire more animators and engineers!  Mattress slightly overstuffed with cabbage. Burn rate goes to 11. Outlook bright.

Day 240: Somehow a couple of teachers got wind of what we were doing and we invited them in for a demo.  They were really harshing on our mellow.  They think the whimsical five minute cut scenes between short sets of learning activities are too long - complaining about how that's going to "steal learning time" or something.  But don't they understand that's what makes learning fun?!  (And how many simoleons we have invested in those videos?!)  They also wanted to know what kinds of student performance data we are going to provide.  We showed them the awesome score meter and the leaderboards but they didn't seem to get it.  That's why we didn't want to bring any teachers up in here - we just knew they wouldn't understand our Vision.  Mattress still stuffed with ample Benjamins. Burn rate holding steady at 11.

Day 330: Brought some kids in to play test for the first time.  They thought the narrative was kind of hokey and felt tacked on to the learning activities.  Lenny's kid called it "chocolate covered broccoli."  That's why I didn't want to bring students up in here - we just knew they wouldn't understand our Vision. The writers and animators are getting a little up in arms because people keep changing requirements and trying to mess with their narrative.  I sympathize with them - the narrative is the hard and expensive part, so shouldn't we revise the science material to fit it instead of the other way around?  Running a bit behind schedule.  Mattress feeling a bit lumpy.  Better pull back on contractors to conserve cash.  Burn rate reduced to 8.

You can imagine where it goes from here - the way of 38 Studios or a quick exit that amounts to a soft crash landing.  Like I said - quicksand.  The pattern is that they don't discover where the real complexities in education lie until they are in it up to their necks.

Here's some unsolicited advice for engineers and investors who are eyeing EdTech:
  • If you think you can crack the code on better education with money, you are wrong
  • If you think you can crack the code on better education with raw intelligence, you are wrong
  • If you think the core challenges in EdTech are technical, you are wrong
  • If you think the silver bullet is in "making learning fun" or "engaging" students, you are wrong
  • If you think the solution is in making clever lessons for each concept, you are wrong
  • If you think you can solve this problem with better graphics, animation, and narrative, you are wrong
  • If you think bringing even all of the above elements to the table *must* be sufficient to crack the code on better education, you are still wrong - though you might be able to generate a positive ROI that way.  Or not.

Relay Ventures, Andreessen Horowitz, and Emerson Collective (the investors backing Galxyz) - you are welcome to use the above as a litmus test when evaluating EdTech pitches in the future.  (Acknowledgement as the source is always appreciated.)

Don't get me wrong: I strongly believe there is a path to better education.  This just isn't it.

I can hear the incredulous guffaws now.  We are on Day 30 and the outlook is blazing - we have an epic promo video and the prize is practically within reach!

I don't really see any point in debating the matter right now - this is a prediction, not a challenge.  Let's check back in a year and see where things stand.  I would be pleasantly surprised to be wrong.  But I wouldn't bet on it.


  1. Well said son, I agree with you of course... Not because I'm your dad, because I think you're right.

    Have a great day...

  2. Michael:

    Who, in your opinion, is doing it right? Do you have a positive litmus test (something like the Joel Test to measure yourself against? Any starting pointers or literature for a software developer interested in EdTech but overwhelmed by where to start?

    1. Hi, mj.

      Thanks for the comment. I know the Joel Test well.

      I haven't put together a positive litmus test but you've given me a great idea for a future blog post!

      Let me reflect on this a bit and I'll post some initial thoughts.

    2. mj - I've been thinking hard about the Litmus Test idea since I read your comment. Thanks for the good suggestion.

      I need more time and space to provide an adequate response, so I'll save that for a follow-on blog post. But I wanted to give you some thoughts now.

      EdTech involves two dimensions - Education and Technology

      One of the biggest litmus test criteria would be:
      * Which dimension gets priority within an organization (Ed or Tech)?
      If it's Ed, that's a point in their favor. If it's Tech (as it is with Galxyz), that's a yellow or red flag. The way people often talk about EdTech, they really highlight the Tech - I guess they assume the Ed dimension is a given. But it's not - it's by far the more challenging dimension. And if you stop and think about it for a second - how can you even begin to evaluate the quality of a proposed educational technology without knowing the expected value of the educational core first? In fact, just because people call it "educational technology" or *intend* for it to be educational doesn't guarantee that it *is*, in fact, the least bit educational. See, for example, this post by Balefire Labs that evaluates the "Top 10 Educational iOS apps" based on instructional design principles - and finds that 5 of them are *non-instructional* and three more get a failing rating: ("Only Two of Top 10 EdApps in iTunes are Worth Buying)

      Think about that for a second - fully half of the Top 10 Apps that are labelled "Educational" in iTunes do not meet the minimum criteria to provide any kind of instruction. And another three are among the very worst in terms of instructional design. But many people would consider all 10 of these apps as examples of EdTech and, in fact, because of their star ratings might even hold them up as exemplars of the cream of the EdTech crop. Even though they aren't educational. (Yikes.)

      Also, see my response to Peps McCrea's comment below and see his post on EdSurge on a related topic:

      Stay tuned for more on this!

    3. I understand why these things happen. I love writing software, and after doing it for several years, I've wanted to make more of an impact. Education is one of the obvious categories. Unfortunately, I tend to see solutions through the tools that I'm familiar with. I think there's a lot of folks like me that just jump in and start doing. Any info that gets me closer to making things that have real, lasting value is appreciated. I'm looking forward to future posts (and others showing up with great comments like this post).

    4. I wrote a chapter with Zachary Stein and Howard Gardner that you might find interesting. It's called "Bridging Between Brain Science and Educational Practice with Design Patterns." Chapter 16 in the book "Neuroscience in Education: the Good, the Bad, and the Ugly."

      Preview on Google Books. I can't paste the link right now for some reason. Or email me and I will give you more info. Right up your alley-design patterns for education.

  3. Helpful to see this explicated. We need more writing like this as it helps us all to challenge and understand what works and what doesn't. Without it we will just keep recycling the same old narrative. Would definitely like to see your effort at a positive litmus test. For me, consideration of pedagogical design at the conceptual is what makes the difference between edtech and tech. More:

    1. Thanks for the link. Your article is thoughtful, well-articulated, and highly relevant. I had a couple of the exact same criteria in mind. In particular, near the top of the list would be "Has someone with deep expertise in learning design in an executive role within the organization."

      Another criterion would echo your implicit point that the litmus test for EdTech is not primarily a litmus test on the technology. This may seem counterintuitive, given the way EdTech is typically framed and funded. The core needs to be about the education dimension of EdTech, and the technology needs to support that (via access, scalability, and affordances that represent new or more effective learning design options).

      Galxyz, for example, leads with the technology and other elements that should be secondary (like graphics, narrative, game play), and that's a red flag.

      I was also thinking of Dan Meyer as one positive example passing the litmus test. As you pointed out, what he is doing does not necessarily scan as "EdTech" for many people (despite his effective use of video) because the technology is not "in your face" enough. But that just demonstrates the point - the way "EdTech" has been popularly constructed as a concept is flawed.

      I've spent some time thinking about the "Litmus Test" idea since mj raised it and I'll post a more carefully thought out follow-up to this post soon.

      I look forward to continuing the dialogue.

  4. It's surprising to me that any company that is floating an idea like Galxyz (that is hard to spell!) is getting VC. It definitely suggests to me that these VCs don't understand that there currently is no viable revenue model for apps that makes money. (And we know that VCs are looking for considerable ROI.)

    I agree that having learning designers (or instructional designers) on board is important for early planning. But my experience is that not all learning designers are created equal and only a portion of them, who have been trained in a particular way, focus on the functional aspects of design. I care less about having a formally-trained "learning designer" on board than I do about someone who asks some basic questions, such as:
    - what is the goal learner performance?
    - how will I measure it?
    - what are the component parts of the composite performance?
    - how will I measure those?
    - how can I iterate my instruction based on what the student performance data tell me?

    I almost don't care what the background of the team is if they ask those kinds of questions and execute on them. I have actually found that experts in areas OTHER THAN education are often better at taking a functional approach to learning design...people from marketing and engineering, for example, who are accustomed to metrics and data guiding their decisions. I find that educators sometimes get too "mushy," trying to explain away why the data show what they show, instead of just listening to them and changing their approach. (I say this as an Educational Psychologist and Instructional Designer, by the way.)

    How about this idea....perhaps where we really need talented and functionally-based learning designers is on those VC teams. How would the funding patterns change if we had input at that level?

    1. > perhaps where we really need talented and functionally-based learning designers is on those VC teams. How would the funding patterns change if we had input at that level?

      I think this is a very interesting idea.

  5. enjoy all responses. few further thoughts:

    1. edtech has greatest impact when it uses technology to make learning happen in a way that was previously very difficult or impossible. using technology to offer the same learning experience, but just on an iPad adds limited value. at it's best edtech disrupts dominant pedagogies, rather than sustains them.

    2. agree that educators aren't always the best learning designers. teacher training doesn't teach the design process, and that's the bit that matters. teachers with an engineering/design head is what edtech needs.

    3. the best edtech tech comes from a back and forth iterative evolution of the affordances the ed and the tech provide each other. considering them separately is a limiting way of thinking.

    eg: we want to help people learn in this way > okay, let's develop some tech allows us to do this > done, but what we've made means we could teach in an even better way > etc.

    1. 1. Agreed
      2. Agreed
      3. Agreed
      (My point above was that people seem to focus too much on the tech and forget the ed, when in fact the ed bit is actually what edtech is about at the end of the day.)

    2. This comment has been removed by the author.

    3. Whoa...I have an entire posting's worth of responses. In fact, since the original post I've been mulling over my responses and logged on just to see what other's have said. I will post more later, but for now I want to reply to these following comments:

      Peps eq above mirrors exactly what I've been shouting out to the IT folks at schools.

      Karen's "basic questions" - AMEN. Though I am going to say that my initial knee jerk response was...No, teams HAVE to have an educator on board. That is true only if the educator will ask the types of questions Karen proposes. It is sad, even maddening, to me, a teacher, what passes as educational tech that gets vetted in our schools. That is about 15 blog's worth of response.

      Tagging on to Karen's response about how on earth (my spin) did a company like Galxyz get VC and why would they get the write up in the first place. Why did TechCrunch give them media exposure. It isn't just Galxyz that gets this type of exposure from edtech media. Who's on the editorial teams vetting this stuff. Teachers, are likely to say, "Oh, hey, I read about this in ..., those guys are experts, so it must be worthwhile."

      Look, I only have a few short hours in a day to facilitate learning. I don't need bells and whistles. I can provide the engagement on my own. I need ways to get at the deeper understanding that I can't do on my own, or that mirrors what I do, but at an exponentially faster rate. See Peps's #1 - make learning happen in a way that was previously impossible...YES...that is what I want! far as "being so close I can taste it" - we are there, I HAVE tasted it; it's an app the author of this blog co-developed. AND....I have about a 87 more ideas for apps and/or software that I need in my classroom. I don't need extra tools that are just fancy and fun...sure those are great and I use them...but don't market those as educational!!!

      Arghhh....I can't make this short, but my classroom of students awaits! More later.

  6. TSteachers Andhra Pradesh TET 2017 Apply Online Exam Date Schedule Syllabue for Paper I and Paper II Exam Pattern Scheme of Examination at School Education Department Official Website Andhra Pradesh TET 2018 complete details, Improtant dates, Eligibility details, Steps to Apply, Application fee, Admit card and other details. The Commissioner of School Education Andhra Pradesh Teacher Recruitment Board will release the AP TET Application Form 2018 ApplyOnline for Andhra Pradesh Notification at|Andhra Pradesh school education board is going to releasing the AP TET Notification for the recruitment teachers in government schools in Andhra pradesh state.The applications are required from eligible candidates from who are intended to get the teacher jobs in government schools ap.


  7. TSteachers Telangana Postal Circle Released recruitment Notification for 33 MTS Multi Tasking Posts in Telangana with 10th or ITI Educational Qualifications | Online Application Form will be available in Telangana Postal Circle official website from 09.12.2017 to 08.01.2018 Telangana Postal Circle MTS Recruitment 2017-18 Notification. In accordance with the notification, there are total 33 Multi Tasking Staff Jobs are available. To get this MTS posts, first, fill the application form which is active from 9th Dec to 8th Jan 2018 @

  8. That is the great one which I found on your website. Thank you so much for sharing this of luck it was a good job.
    digital marketing services in delhi

  9. nice blog.Very great article.thank you.
    we are providing app development training
    iphone job Oriented course

    iphone job training center

  10. Thanks for the useful information. Right on. It's more informative and easy to understand. Please help me suggest
    The Best School in Moolakadai
    The Best School in Vyasarpadi

  11. So much inspired with the documents you have shared. Thanks a lot, this was much useful and handy. Please do help me suggest
    The Best School in Perambur
    The Best School in Vyasarpadi