(image courtesy of the lovely Wordle)

The vast majority of our visitors (86%) at School of Everything come to us by searching through Google. The image above is a snapshot of the kind of words that people are typing in before landing on a School of Everything page (the data is for the month of November and X-Factor definitely has something to do with the popularity of singing).

In the early days of thinking about School of Everything we sometimes described it as “an eBay for teaching and learning” and the analogy kind of stuck. People understood what we meant especially when we clarified that it was about face-to-face learning rather than e-learning.

Then we started to think about a similar business model to eBay — we would charge a commission on successful purchases made through the site. Initially we set this as 5.25% — exactly the same as eBay. We didn’t think we would need the other parts of eBay’s business model such as listing fees.

We modeled the revenues we thought this would bring in based on how many teachers we thought would use the system and how many students would join up. Going back to the original spreadsheets we did, our assumptions were…

1. The number of teachers would grow by 40% per quarter until the end of 2009 to 7,500.
2. Teachers would take 10 payments through us per week.
3. We would have 10 times more members than teachers
4. We would have 30 times more visitors than members (although that’s not really relevant in this model)
5. The average transaction would be £20.00
6. After PayPal fees 2.5% of each transaction would come to School of Everything

and what happened…

1. Well, we got the that one right. We’re now coming to the end of 2009 and sure enough we have just over 7,500 teachers registered with School of Everything. The percentage rate has slowed a little but we’re still growing faster than ever before. The problem is how many of them use the payment system — a tiny number.
2. Way off. We don’t have any teachers who take more than 1 payment per week through us.
3. Out by 50% — we have about 5 times more members signing up compared to teachers.
4. In the last quarter we have had 300,000 unique visitors and 25,000 members.
5. The average transaction has been about £25.
6. We actually get about 3% of the transactions after all the PayPal fees.

So overall, we have roughly (order of magnitude) as many teachers, members and visitors as we thought but they’re just not paying or receiving payments through us. The model (remember this was before we’d built even the first version of School of Everything) said that we should have earned £225,886 from the commission model in the last quarter. That didn’t happen by a long way.

Back when we did our customer research, we found teachers who did indeed say that they would like to be able to take online payments. However, the basic problem seems to have been that people went around the system. They weren’t being sneaky, they just didn’t need our service enough. Teachers were perfectly used to taking payment for their services (of course) and although they complained about cheques being annoying and not get payments when people cancelled lessons at the last minute, it wasn’t a sharp enough pain for them to switch over to the service we built.

So to some extent the problem was the way we designed the service. By using PayPal we made the system easier to implement but harder for people to use. It meant that we needed teachers to have PayPal accounts in order to receive payments and also, to start with at least, students would have to leave the School of Everything site to go to PayPal to complete their payment. There were just too many steps to the process on both sides for it to be a really valuable service either for the teacher or the student.

Commission was just the first model we thought of and tried. I’ll give details of the other models in future posts. The question is whether this is the model we should return to and improve or whether we should come up with a completely new one.

It’s pretty difficult to talk about what you’ve got wrong. When you’ve been working on something like School of Everything very intensely for two years you can’t really blame the mistakes on anybody else. But the truth is that we need to rethink because we haven’t managed to make the idea financially sustainable yet.

Steve Blank talks about the myth of ‘first mover advantage’ and how actually many of the start-up successes of the internet age haven’t actually been first movers at all. Google didn’t invent search, Amazon didn’t invent online retail, eBay wasn’t even the first company to try to create a marketplace. They were all ‘fast followers’ who saw what other people had tried and improved on it dramatically by executing really well and finding business models that worked.

The challenge for us is that we were ‘first movers’. Nobody had tried to do what we were doing (actually much to our surprise in the early days). But we now need to be our own ‘fast followers’ as well. We need to turn on a sixpence and be able to learn from our own mistakes rather than those of other people and that’s going to be a bit painful because as we do so there are going to be quite a few “D’oh!” moments.

So I think the best policy is just honesty. We’ll be posting data about predictions that didn’t come true, metrics that never quite went in the right direction and evidence about the problems we’ve faced as well as the many successes we’ve had along the way. We have a hell of a lot of information though so if there’s something that would help you help us, just drop me an email (paul[at]schoolofeverything.com) and if I can, I’ll put it up.

We’ve been doing a lot of work over the last few days understanding the people who come to School of Everything. We currently have just under 100,000 unique visitors per month who mainly come from Google search (about 85%). Here’s the demographic information that Google give us about our visitors:

This tallies up with responses to a questionnaire we did in October that showed that our visitors are slightly more likely to be female, are mainly aged between 19–40 and tend to be in full time employment.

One of our current ‘guesses’ is that we’re not quite managing to match up our supply and demand effectively. I’ll post more soon about what we know about our teachers but would be interesting to know how our visitors compare to other sites.

The first step in Steve Blank’s method is what he calls customer discovery. He talks about developing a hypothesis, both about the product and the customer you’re trying to serve. In his talk he actually says ‘hypothesis’ is really just a more technical word for ‘guess’.

This is actually very like something we do with Social Innovation Camp, especially when we run shorter events at our meetups to help people generate ideas that could be entered into the camps. We have five things that teams have to think about. Somewhere between the two and making it a bit more relevant to our situation we’ve come up with this list of things to think about when coming up with ‘guesses’ as part of customer discovery.

1. What problem are you solving?
2. What’s your solution?
3. Who is it for?
4. How are they going to find out about it?
5. How much will they be willing to pay for it?

A few months ago I read an interesting blog post by Nigel Eccles about a book called ‘Four Steps to the Epiphany’ by Steve Blank. It was so interesting that I forked out and got it shipped from the US by Amazon.com. A few weeks later it arrived and I sat down to have a read and pretty promptly got stuck. It’s not exactly a rip-roaring yarn. In fact even Steve Blank admits it’s a bit ‘turgid’.

Now we’ve been forced to rethink what it is we’re offering people with School of Everything, I’ve gone back to the book and we’ve all been reading it at Everything HQ. The basic premise is that you should go through a process of ‘customer development’ rather than the standard way that start-ups tend to think of things which is ‘product development’.

Although the book is a good reference, I’ve found myself looking on the web for other resources. This post is really just a starting point if you haven’t had a look at the ideas around customer development before. Basically I wouldn’t bother buying the book until you really need to, I’d read these posts instead:

• Eric Ries on ‘customer development’
• Venture Hacks summary of the book
• Jon Bischke’s links to other customer development resources

And perhaps best of all, listen to this podcast of a talk by Steve Blank given at Stanford. It’s the clearest introduction to the ideas I’ve come across so far.