[cpks]

> The repeated “success, scale, fail” experience of the last 20 years of development practice suggests something super boring: Development projects thrive or tank according to the specific dynamics of the place in which they’re applied. It’s not that you test something in one place, then scale it up to 50. It’s that you test it in one place, then test it in another, then another. No one will ever be invited to explain that in a TED talk.

Perhaps. That model works very well in Silicon Valley. It's just that the expectation is that most projects will fail. Development projects are similar, only with an unrealistic expectation of success early on, and without appropriate processes to fail fast, recover money from failures, etc.

What I'd like to see is fundraising with:

* Big figures * Long-term windows * Milestones

In other words, I raise $10 million as $2 million per year, contingent on meeting specific milestones. Or $200k/$1 million/$4 million. Or whatever. As a fundraiser, I have a strong interest to write realistic plans -- not just grandiose but impractical visions -- since if I fail to execute, I lose funding. I also have a confirmed, steady, long-term source of funding, so I'm not spending 90% of my time selling.

If I do fail, the money is recovered.

[wozniacki]

  In 2010, “Frontline” returned to the schools where they had filmed children 
  laughing on the merry-go-rounds, splashing each other with water. They 
  discovered pumps rusting, billboards unsold, women stooping to turn the wheel
  in pairs. Many of the villages hadn’t even been asked if they wanted a 
  PlayPump, they just got one, sometimes replacing the handpumps they already
  had. In one community, adults were paying children to operate the pump. 

  Let’s not pretend to be surprised by any of this. The PlayPump story is a sort
  of Mad Libs version of a narrative we’re all familiar with by now: Exciting new
  development idea, huge impact in one location, influx of donor dollars, quick
  expansion, failure.

Frontline Video from 2005 "The Play Pump" :

http://www.pbs.org/frontlineworld/watch/player.html?pkg=entr...

Frontline Video from 2010 "Troubled Water" :

http://video.pbs.org/video/1533347949/

[cpks]

And it cost $15 million.

The annual cost of product failures in the US is $100 billion.

Big whoop. PlayPump is 0.1% of that. Your share of the PlayPump (assuming equal cost to everyone in the US)? A nickel. You don't just fund a PlayPump. You fund a thousand PlayPumps. Nine hundred fail, and one hundred have impact.

Someone tried to do good. Perhaps they didn't succeed. What happened as a result? They got ridiculed. That's the failure of PlayPump. Not the folks who did it, but the folks in 2010 mocking them. It brings about a culture which doesn't take risks, and that's a culture which cannot bring about change in the world.

[notahacker]

The Silicon Valley "success, scale, fail" process doesn't work in Silicon Valley either if you look at average long term returns to average VC investors, rather than all the positive side effects.

The difference is that the returns from successfully scaling up in SV make those successes glaringly obvious and desirable to emulate

Scaling up across 105 more villages, on the other hand, simply produces more anecdotes for the pitch deck for the next wave of funding, which will need a bigger amount (and face a greater chance of it being hampered by evidence of something, somewhere having gone wrong). Appropriate milestones and accurate measurement are as much a part of the problem as the solution. Imagine gauging the relative suitability of an early Pets.com and Amazon for follow on funding not from their sales, SV style, but by interns flying out to foreign lands to quiz villagers on how their purchases had improved their quality of life.