Managing network based enterprises teaches that(echoing Aboyeji’s comment) the best designs naturally scale. Focusing on infrastructure; creating platforms for exchange of local resources and creativity, promotes self-sustaining solutions that not only scale, but become viral.

These solutions follow a principle expressed in a quote from Keith Hoffman, (paraphrasing) “the job of (NGO) social enterprise is to express the needs of a community in the market place,” to motivate creation of value in context to these needs. The resulting needs based exchange, lifts the community and builds ecosystems that grow along with it.

Just one point (I am not sure but you may have mentioned this somewhat in your post):

Scale up should pay for itself.

If the scale up is not going to make the operation more self sustainable in the long run, it probably isn’t worth scaling up in the first place. If this does not happen, it will be difficult to measure success…besides, the point of aid is an end to aid…a scale up should close shop..and in a good way

The point 5 implicitly means also (the other way around) that if a pilot does NOT work well, the bigger scale may still work. Because of the general equilibrium effect that are not present at the pilot level…

One approach that the WB is advocating for is to get the governments to design the assignment to their programs in a way that it can be evaluated easily. This gives the opportunity to starts with a big enough scale and then to improve along the way if necessary.

I like the approach but obviously it has it own flaws. One of them being that in some cases, instead of providing the program to the people it will presumably serve most, they just randomize…

What’s needed instead is an understanding of key stakeholders that would benefit/lose from reforms, leading to a plausible successful coalition of supporters that could form to support a specific set of reforms, and what their roles need to be as the reforms go ahead (perhaps leading to a different formulation of the nature of the reforms, or a restructuring if project already approved), and a process approach for making it happen (and making sure reforms aren’t reversed by opponents). This rarely happens in my experience, but when it does, reforms supported by aid donors can be scaled up and succeed.

This is true of clean water: why is it still a problem when supplies and labor for simple wells are abundant?

Superstitions and local customs are not to be discounted. One sub-Saharan Africa example http://bandwidthout.org/2009/07/pride-and-hydration/ There were endless variations on this.

That’s why COMMUNICATION is the most important thing to scale up, especially among peers. It scales virally, can be low or no-tech…

I welcome suggestions for getting funders and implementers to prioritize this.

Thanks!

The aid will be a waste anyway, as all previous scaling ups,

but the improved governance might actually help people.

You can scale up what has a transferable management: inputs and tools are abundant or can be easily created or transferred.

I’m thinking mainly about knowledge-intensive projects but many other things may apply.

FontlineSMS is the best example I can think of:

- input is mainly knowledge. What is not abundant (usage of mobile phones) is easily transferable (how to build the network and operate the software — apologies for the simplification).

- tools are, again, knowledge; the software, which is freely available; mobile phones, quite “cheap” around; and some desktop PCs, again quite “cheap” everywhere.

BTW, just discovered your blog: loved your Sex, Aid, and Rock & Roll analysis

i.

SS

Footnote to #2) If what you do best is not industry best practice, or even good practice, don’t scale up. Just stop doing it.

I’d re-state Mr. Barnes comment above approximately as follows: Don’t assume because something works well and can be scaled up in one context that it can necessarily be scaled up in another. Every context requires nuanced approaches for even those interventions that seem most basic.