We’re in the thick of a state legislative session here, which means that action alerts are flying fast and furious. I think there’s a lot of poorly written and ultimately ineffective alerts out there, but I can’t *prove* it.
Which leads me to an idea…
What if we organized the infrastructure to write and test different versions of a bunch of action alerts, measure the results, rewarded the winners, and wrote up the lessons learned — with statistics? We could structure it as a contest to give aspiring alert writers an incentive to participate.
We’d need to find a couple of organizations with lists large enough to randomly sample/segment, and organize the system for the alert facts to get out to the writers and turned around quickly. This all seems totally doable.