Huge amounts of time and money are spent trying to persuade people who have already made up their minds, to turn out voters who don’t support our candidates, and to market a product or service towards those who would never purchase it. No matter the size of your organization, microtargeting can help you focus scarce resources where they will make the biggest difference.

In politics, an organization would ideally ask every voter who they were supporting, and then focus persuasion resources only on those who said that they were undecided, and GOTV (Get Out The Vote) resources only on supporters. In the real world, however, it is not possible to ID every voter. More and more voters have unlisted numbers, have caller ID, or are just unwilling to answer ID questions. There are also cost considerations. Even attempting to call every voter in a district can be prohibitively expensive. Microtargeting allows us to predict how voters who we were not able to reach directly would have responded.

  • Know how likely it is that a voter is undecided

Focus dollars and time spent on persuasion phones, mail, door-knocking, radio and television on the voters most likely to be undecided.

  • Know what issues undecided voters care about

Target the most powerful message to each undecided voter.

  • Know how likely it is that a voter supports your candidate

Focus GOTV (Get Out The Vote) efforts on the voters most likely to actually support your candidate.

  • Know how likely it is for a voter to contribute

Boost the returns from fundraising prospecting by mailing or calling those most likely to contribute if asked. At the same time, expand your contributor list by finding brand-new donors, rather than just resoliciting the same pool of known political donors.

How it works:

The idea behind microtargeting is to look at what we know about voters we were able to reach, and to use that information to build a statistical model to predict how voters we didn't reach would have IDed if we had reached them. Strategic Telemetry combines voter ID information from phone or door-to-door canvasses with detailed demographic and marketing information in order to build the statistical models. This technique can be applied to predict support for a specific candidate or issue, likelihood of being undecided, likelihood of voting, or to find likely contributors.

Strategic Telemetry uses a variety of statistical techniques to build the best possible model for each project. When building a model, it is vitally important to set aside some of the original IDs in a test set. The model is then built using the rest of the IDs. Once the model is completed, it is applied to the test set to verify its accuracy. The percent ID'ed as exhibiting a certain behavior (i.e., supporting our candidate, undecided, etc.) in each segment of the test set should be within the margin of error of the predicted value. It’s easy to build a model that looks impressive, but doesn’t hold up in the real world. That’s why Strategic Telemetry always uses test sets to validate models so that organizations can use them with confidence.