Why education startups do not succeed

Instead of a four-year, pass-or-fail model for education, how about switching to a incremental model? You can have 50 levels of applied knowledge, rising all the way to PhD. People can purchase small amounts as they can afford it. Certification systems could tie into the levels. Early levels could require 2 or 3 courses in other areas. Later levels could be more narrowly targeted. A universal directory could manage it all and allow competition between providers.

As some random internet commenter, I agree with the author: this field is getting ready for major innovation. Just not quite yet.

The one thing I'd like to see the data for is the notion that the poor in America are more likely to try new things. This has not been my experience -at all-. I hadn't even heard of a Montessori school until I came to Palo Alto, CA. The idea that the poor are more likely to try things would make sense, but as an Asian-American growing up in this country, I noticed that the poor Caucasians were more likely to be set in the their ways than the rich Caucasians.

Sal Khan's vision of what education should be like is spot on and I strongly agree with him. He seems to be sincere about this thing and has received millions in donations.

This is not a company looking for success by means of an 'exit'. They are actually making a difference to millions of students. In doing that they're being successful.

Wish we could see more of this instead of useless, gimmicky stuff.

I think that's why VC and education do not go well together. If you plan to make an exit, then you're not really there to support the next generation of people who want to learn. You're not a teacher or educator if you're there only to make the jump up to the next thing.

In that sense, the idea of passionate angels, and small startups that can find a niche and build a sustainable business out of it makes a lot of sense. It might not make you a millionaire, or won't disrupt the entire education market, but it should hopefully make you proud and help even a few people improve their learning experience.

disclaimer: my startup kenhub.com tries to build a (small) platform to make learning anatomy more easy and fun.

It damages the credibility of the author to make a statement like this. University of Phoenix degrees are less marketable than those of public universities. If someone wants specific evidence there is a Frontline episode on education that features UoP heavily. In addition, UoP is basically a parasitic entity exploiting a loophole. They take federal money intended to provide access to useful education, and people who think a degree will improve their marketability, fail to deliver to either party and pocket a percentage.

Otherwise, I pretty much agree with everything in this article based on what I've learned about the education business recently. It's a long, long, long game and my guess is that most investors (right now) will not have the stamina to play it. It will happen, but my guess is that tipping point will not be reached as fast as it does in other verticals.

So to make money on an education business, you either need very broad use (at least six figures of users), or something of very high value to those who can afford it - the thousand people for a thousand dollars model (much more than that, and you're actually selling to institutions).

When we say "startup" around here, we're specifically talking about a pretty narrow economic model... angel/venture funded software companies targeting $10M-$100M+ annual sales in less than a decade. It needs those hockey-stick graphs to justify the investment. How do you get that?

There are probably spaces that can be exploited by startups - mobile software that simplifies the lives of college students in the under-$100 range, for example. But really, I think the innovation mostly has to come as entrepreneurship within institutions.

I have a conviction that a quality education should be available to everyone. To that end, it really seems that most education resources should be free and open.

That said, developers deserve to be paid well for their time. I think the ed-tech world should focus on paying good developers a respectable rate for creating high quality, free and open resources.

> The average, middle class person thinks about education as an expenditure, not an investment

Is it a state-occupied (compulsory) education system that does so? IMO, it is. For example, many(most? vast majority of?) people go to college to get a diploma, not to get actual education. The system sucks and doesn't deliver any actual value(in terms of knowledge or personal growth, not in terms of signaling to the market), so no wonder that people don't view education as an investment.

http://www.digitalscienceacademy.com/

We are getting funding through Google Adsense and it has worked out so far.

I think this could be the future of education: delivering quality content with as little overhead as possible.

I guess we'll see how it works out.

For fine arts such as singing, piano, violin, other instruments in a symphony orchestra, guitar, dance, painting, sculpture, photography, etc. it appears that overwhelmingly the way children make progress is (A) their family has some expertise and encourages their children and (B) the education is available in the home, in the church (e.g., for choir), or from private tutors, e.g., a local piano or violin teacher. E.g., for violin, consider Ms. Caroline Goulding as at

     http://www.carolinegoulding.com/

     http://www.youtube.com/watch?v=Tezau3hlRxs

For the consumer market for grades 7-12, before a consumer will devote much time or energy to education, they need to know that there will be a 'payoff' from official education 'certification' which mostly means better grades in grades 9-12 or higher scores on tests of the College Entrance Examination Board (CEEB), e.g., as at

     http://en.wikipedia.org/wiki/College_Board

So, really such consumer education has to aim at existing graded courses in grades 9-12 and/or subjects tested by the CEEB.

Now we face a hard lesson: Doing well on these aims is "not a spectator sport", more work than just watching TV, and a lot of hard work.

For such learning, the first source is the student's usual school. For more, the student may want some tutoring. A few students want to do better than what is available in the school.

For tutoring, here's an approach: Find what are the more popular textbooks and develop some videos tied closely to those main books. Then a student who needs tutoring, or just to get ready for a test on Monday, could use the videos for just the place they are in their textbook.

For going beyond what is available in the school, a student could use some guidance and then the learning materials. For going beyond, we are really aiming just at some CEEB tests, in particular, the tests of the College Level Examination Program (CLEP) as at

     http://en.wikipedia.org/wiki/College_Level_Examination_Program

E.g., maybe a student wants to do well in calculus. So, get a book on calculus aimed at the CLEP test in calculus.

For still more, that is, for an ambitious student in grades 9-12, just aim at the subject matter tests of the GRE.

For vocational, the audience of Hacker News the unique, world class cream of such education. Why? Because in the US, nearly all education in practical computing is from self teaching from books, on-line fora, practicing at a computer, on-line documentation, e.g., the Microsoft MSDN Web pages, etc.

Improving such vocational education? Not easy because it would involve at least writing better books than the best ones already available from O'Reilly, Microsoft Press, and many other publishers and better Web sites, fora, etc. For one company in education to improve on all those sources is hopelessly difficult.

So, where do I see a significant startup opportunity in education for grades 9-12 and vocational education? I don't. Really, that's good news: The CEEB seems to have the needed means of 'certification', their tests. Otherwise, get some good learning materials, commonly just textbooks although maybe PDFs downloaded from the Internet.

The bottleneck is the work -- no spectator sport and no royal road. Instead hard work. E.g., look at the left hand finger tips of Ms. Ibragimova; they are bent backwards. So were those of Jascha Heifetz. Why? Thousands of hours of intense practice. Hard work.

I did some such things: (1) I'm no good at violin, have too little talent and never practiced nearly enough, but I did make it through the D-major section of 'Chaconne' and parts of the rest mostly via being self-taught. (2) I never took freshman calculus. The college where I did my freshman year was not so good and pushed everyone into a 'college algebra' class a bit beneath what I'd already done in high school (in the same city as the college but the best high school in the city -- Dad carefully selected where he bought his house). So, I showed up for the tests only and otherwise got a good calculus book and dug in. For my sophomore year, I went to a much better college and one that happened to have a quite good math department and just started on their sophomore calculus from Johnson and Kiokemeister, then also used at Harvard. It's a good calculus book. Did fine: Made As in that calculus, got Honors in Math, and 800 on the math subject matter test of the GRE. (3) I continued with a lot of largely independent study much like I'd used for freshman calculus, and that background was crucial for my Ph.D. in Engineering -- really some applied math. (4) My career has been in applied math and computing. I've taught computing in two well-known universities, but I never really took a course in it and, instead, was essentially self-taught, as is nearly standard on HN.

One little victory: One Christmas at the farm of my wife's family, I was upstairs practicing violin. A bright daughter, about 9, of one of my wife's sisters came up and watched. So I put my violin bow in her right hand, showing her how to position her fingers to hold it (the more natural Russian way Heifetz used and not the less natural German way!), put the violin under her left chin, and let her make sounds. Her mother had been trying without success for years to get that girl interested in music, but the next day her father asked me "Now, how much is a violin going to cost me?". She was a bright girl: She was at a large, very competitive high school, the pride of the community, and in her senior year her parents were surprised to discover that she was by a wide margin the head of her class. She got Valedictorian. In college the retraced the steps of my wife and got PBK. She got her LLB at Harvard and started at Cravath, Swaine & Moore. Bright girl.

So, net, really, just do the darned work, and no great educational startup opportunity.

For a reader who got this far and heard of the Bach 'Chaconne' here for the first time, here's what it is: At one point, Bach wrote six pieces for unaccompanied violin (he did the same for cello). Three of the pieces he called called 'Sonatas' and the other three, 'Partitas'. These pieces are secular, unlike his other ecclesiastical music. While Bach was mostly a keyboard player, especially for organ, his violin and cello pieces show amazing knowledge of violin and cello, even if heavily he tried to have them play chords more appropriate for keyboards.

The crown jewel of the unaccompanied pieces is the last part of the Partita #2 called the 'Chaconne' which was a old dance form.

A lot of the music is trying to play on 3-4 strings at once. Those efforts can view as chords sometimes and just more than one melodic line at other times. Bach was good at doing such things.

The piece is in three sections, D-minor, D-major, and D-minor again and is a 'theme and variations'. The variations have great variety, slow to fast, calm to agitated, low pitches to high.

The D-major section starts off calm. Soon it is playing some nice arpeggios with some triplets. The triplets get to be more and more intense, say, 'insistent'. The end of the section is essentially the climax of the piece and is the most difficult section to play. Depending on how it is played and heard, the last bars can sound like chords or several melodic lines, each heard before. How to play the last few notes is up to the violinist, but Heifetz makes a big climax there and then quickly moves to the start of the last D-minor section which can view as some 'cathartic' relief from the intensity of the climax at the end of the D-major section.

Due to the use of the chords and the relatively non-melodic main theme, for a listener the piece takes some effort to follow well enough to 'like'.

What it all 'means' is for the listener to decide. But it does appear that something intense and determined is going on underneath there somewhere. Playing it is great fun -- parts of it are a great way to scream out to the heavens the spirit of humanity or some such.

There is an old joke: At times the piece is played on guitar. So, at a concert a guitarist in the audience sat waiting for the concert to begin. With nothing else to do, he mentioned to the person next to him "The Bach 'Chaconne' sure is difficult to play."

The next person was composer Castelnuovo-Tedesco, known to be a man of few words, who said nothing until the concert was over at which time he turned to the guitarist and said, "The Bach 'Chaconne' is the greatest piece of music ever written." It's my favorite.

Again, the bottleneck in education is the need for the students to do hard work. Given that, there's lots of great music, good violins and pianos, textbooks, CEEB tests, colleges, universities, guidance, YouTube videos, etc. With the hard work, we can soar to the heavens. Without the hard work, we're stuck in the mud.

1. Public financing. This is the most successful; even if there's a reputation of bureaucracy and mediocrity (largely because public schools have to contend with the full range of inputs) the system actually works extremely well. A lot of people (read: short-sighted conservatives) just don't want to pay for it, so they gripe about the public-school model itself, rather than (admittedly, serious) issues with execution.

2. After-the-fact donations. The problem here is that it converges on a Nash equilibrium where the rich get the resources anyway because they are most likely to be future donors. (Depressing news: in spite of the "economic miracle" that college is supposed to work, but hasn't for a long time, the #1 predictor of whether someone will be wealthy is having wealthy parents.)

3. Non-dischargeable debt. Student debt might have been a good idea (a) before it started having effects itself on pricing --making college again unaffordable and replacing minor debt loads with indentured servitude-- and (b) were the economy less volatile than it has been since 1990 or so. It's horrible now.

None of these solutions (of the three, public financing being the only one that works) admit themselves well to the startup business model.