Areas of Expertise

Strategy

Strategy is the step where what voters said becomes what the campaign does on Monday.

From finding to decision

Every finding arrives attached to a decision

Research that stops at a chart hands the campaign a second job: work out what the chart means, at night, between a finance call and a candidate forum. Most of the value in research is lost right there.

So findings are written as pairs. Here is what we learned, and here is what we recommend you do about it. The recommendation names the action, who on the campaign does it, what it costs, the date it has to happen by, and how much weight the evidence behind it can actually bear.

That last part is graded openly, in three levels:

Act on this. The signal is consistent across waves and strong enough to carry a spending decision.
Act if it matches what you are hearing. The research points one way and your field program should be able to confirm or contradict it this week.
Signal only. Interesting, thin, and queued for a harder test in the next wave. Nobody should move money on it yet.

Grading evidence openly is a discipline. It makes it possible to say later which recommendations were strong calls and which were guesses, and that record is how a campaign learns whether to keep listening.

What the decisions are

The decisions that move a race

Campaigns in these races do not have a hundred levers. They have about six, and they get pulled with limited money and no time to reconsider.

Message order. Which of your issues leads, which supports it, and which one you stop saying out loud.
Language. The exact wording on the mail piece, the door script and the digital ad, taken from how voters said it rather than how the campaign would phrase it.
Field. Which precincts the volunteers walk this weekend, and which audiences to stop spending on entirely.
Money. Where the next $2,000 goes when there is no more after that.
The opponent. What voters already believe about them, and whether raising it helps you or helps them.
Timing. What gets held until the last three weeks, when most voters in a down ballot race start paying attention at all.

Targeting from Modeling decides who hears you. Strategy decides what they hear and when.

How it runs

How the month runs

A wave of Insights goes into the field with the questions that follow from last month's decisions.
Open answers are read and coded by hand, and the exact wording is kept.
A memo is written: findings, the limit of the study on the front page, recommended actions, confidence grade, cost and deadline for each.
One call to argue with it. Disagreement from the candidate and the manager is the most useful hour in the month, and it changes the recommendations.
The decisions that come out of that call get written down, and the next wave tests whether they worked.

Nothing is held back for a larger reveal later. A finding that matters in week two is sent in week two, by email, in a paragraph.

Three things strategy will not do:

It will not tell you what to believe. Positions belong to the candidate, and research describes the district rather than choosing your convictions for you.
It will not claim the evidence says more than it does to make a recommendation sound stronger.
It will not quietly drop an inconvenient finding because advice already given points the other way. When the research contradicts us, the memo says so.

Turn findings into a plan for the week

Send the office, the election date and what you have set aside for research. We write back with a scope, a price and the first set of questions worth asking your voters.

Contact Upset Research