The impact funnel

Fermi estimates for political startups and funders

Political entrepreneurs need to describe their their potential impact in order to attract funding, teammates, and other support. Similarly, funders of political startups need to figure out what projects have a chance of big success. For both groups, I’ve found Fermi estimates¹ to be a useful tool.

Say you want to get people to call their mayor, and your strategy has four steps: (1) Email people, to get them to (2) join a training, on how to (3) ask their neighbors, to (4) call the mayor. You can sketch out your possible impact like this:

Multiplying all those numbers gives 400, which is the number of neighbor-to-neighbor conversations you’d expect to produce. If 1% of those 400 conversations produce a call to the mayor, then you expect to produce four calls.

This type of analysis can help in several ways.

Use 1: Validation

A Fermi estimate makes projections explicit, which is the first step towards testing them. Lightweight testing to validate each assumption can save a lot of time and money by catching aggressive estimates early. Tests can help attract partners by de-risking your project; “scaling up something that appears to work” is more appealing than “trying something totally new.”

Use 2: Strategy

Fermi estimates can help you figure out where to spend time. In the example above, is it better to improve the conversion rate of your initial email, or of your trainings? Or maybe you should skip the training step altogether? The Fermi estimate lets you play with different possibilities.

Use 3: Comparison

Funders and entrepreneurs benefit from knowing the cost-effectiveness of various projects. If your program above costs $500 and produces 4 calls, that’s $125 per call. Might there be some other approach that costs half as much? Or perhaps existing methods cost $1,000 per call, so your approach is way better? This method of comparison is reductive (see “Searching beyond cost-per-vote”) but can still be useful.

Use 4: Scalability

By sketching out a Fermi estimate where everything goes well, you can get a sense of how large a project’s impact could be. In How to Start a Political Startup, I quote Ted Suzman:

[T]he first question I ask is: if it truly succeeds, how much does it matter? Try to imagine the largest plausible success case. In that case, how glad would you be to have worked on the project?

Concluding questions

• Where have you seen Fermi estimates be helpful or unhelpful?

• What other back-of-the-envelope analyses have you found useful in explaining or projecting impact?

1. If you want more about Fermi estimates in general, or don’t know what they are, click here for a one-page overview.