Lessons from OLPC

by David Week on 02 July 2013


The news, finally, is that OLPC (One Laptop Per Child) is falling apart. The news come via this post on OLPC News, which cites:

  • multiple failures on project delivery
  • poor evaluation results
  • top staff leaving in droves.

I thought OLPC would fall apart about two years ago. There were two flaws in its business model:

  1. It is hardware oriented—as reflected in its name. Development is a social process, not a hardware process. Like the difference between sustainable village water supplies, which start with community engagement and ownership, and “one pump per village.” The former is development, the latter is not going to work—as we learned in the 1960s.
  2. OLPC can never keep up with the mainstream manufacturers. It was clear then that smartphones and netbooks were getting cheaper and better at a prodigious rate. A single firm, with limited capital, using proprietary hardware and proprietary software can never, ever sustain the pace of innovation required to keep up.

There are three lessons here for architects working in development:

Development is not a hardware problem

Architects love to design stuff, and I see endless efforts to design one-off “model” buildings, new forms of disaster shelter, new forms of housing, new forms of school. None of these, for the most part, ever take off. Most of them don’t even have any influence, no matter how well thought through or smart they are.


  • Development is a social process
  • Development problems are social process problems, not built form problems
  • There are always local builders and designers, who don’t look kindly to foreigners air-dropping supposed solutions into their country
  • No matter how smart the air-dropped designs, they always miss many important factors which local builders and designers  know about, but are not readily visible. (In other words: buildings are the way they are for good local reasons, not because local people are bad designers.)

Solution: Work on incremental improvements to the process of building, as understood by local experts, with local exports. Use all your skills as an architect, not just your drawing board skills.

Development is not an island

Remote, rural, or poor areas may look like special worlds, separated from the big smoke and the middle class, but in fact rich and poor both live in the same economy. You can’t design for this supposed “special world”, without realising that you are competing with the whole economy, and—increasingly—the global economy.

If you design your beautiful, sophisticated, passive thermal building, and next month the Chinese start selling $35 solar air conditioners (a scenario that’s not unlikely), you will be left in the dust.

Solution: Watch what people are using today, from the global economy. Work with the developments in that economy. Do not attempt to build an “alternative” solution predicated on an alternative economy and alternative social world.

Business is business

Change takes time. Learning takes time. It’s difficult to undertake efforts in sustainable development if you yourself, or your team, are not a sustainable enterprise.

I love this passage from Dickens’ character Mr Micawber, which expresses the key to business sustainability pithily and poetically:

Annual income twenty pounds, annual expenditure nineteen pounds nineteen and six, result happiness. Annual income twenty pounds, annual expenditure twenty pounds nought and six, result misery.

Solution: Have a sustainable business model, which ensures at least a chance that money in is greater than money out over the long term. And then execute well.

Digiprove sealCopyright secured by Digiprove © 2013 David Week
  • Christoph Derndorfer

    Hi David,

    thanks a lot for sharing your perspective and your link back to my piece over on OLPC News.

    While I agree with most of what you’re saying I don’t agree with you when it comes to the “OLPC can never keep up with the mainstream manufacturers.” flaw and partially also with the “Development is not an island” notion.

    The OLPC XO-1.75 and now XO-4 models are very appropriate devices for their intended use (a children’s educational tool) and environment (remote and poor regions of the world) and second to none, particularly nothing that mainstream manufacturers offer.

    The XOs general robustness, easy repairability, low power consumption (at least since the ARM-based XO-1.75), versatility (microphone doubles as sensor input, great wifi reception, dual-mode display), etc. combined with its price make for a a very good value proposition. In that sense I do believe that OLPC found an underserved niche. It’s not quite an island, I’ll give you that. But I think it’s at least an overlooked peninsula.

    Though as you point out having a solid device only goes so far when you (a) don’t manage to build a sustainable business model around it and (b) don’t actually show convincing evidence (or strong indications) that you solve the problem you claim to address…

    Just my 2 eurocents. 🙂

    • Hi Christoph. I’m not intending to disparage the quality of the design or fitness for purpose of the machine itself: only that these by themselves are not sufficient for success. When I saw in the original article that OLPC had missed two ship deadlines for new products, I see that as a sign that it is undercapitalised. But it is in the long run competing with companies that invest billions in product development.

      I had an Osborne 1 when I was working in remote rural Papua New Guinea, in the early 1980s. It was great. That product died when the company founder announced the Osborne 2 before having it ready for market, thus killing sales. Then I bought a Mac 128k, another excellent machine for its time. Apple Corporation almost died after a series of missteps, facing bankruptcy just before the return of its founder.

      Business is a complex thing, and the best product does not always win. Products themselves to do not stand alone, but are part of complex social-technical systems. We need to look beyond the product, to the system.

Previous post:

Next post: