UPSETRESEARCHContact

Areas of Expertise

Insights

Reach the people who will decide the race, ask them real questions, and read what they actually say back.

Start a scope

What Insights is

How Insights works

Insights runs in waves through the race instead of arriving once as a report. Respondents are invited where the voters of your district already are: text, email, the campaign’s own web page, social posts, a code printed on literature, and phone follow-up where a number is good. People choose to take part.

Every wave carries a short set of fixed questions and at least two open ones. A person reads every open answer and quotes it in the memo in the voter’s own words. The phrase a voter uses for the problem on their street is the phrase that belongs on your mail piece.

What a wave gives you:

  • Which issues people raise without being prompted, and in what order.
  • The words they use for those issues, verbatim.
  • How firmly they hold a position, and what moves them off it.
  • Which of your messages drew a response and which drew nothing at all.
  • What people already believe about you, your opponent and the record in office.
  • How all of that shifts between one wave and the next, because the same questions are asked the same way each time.

The words we use

Why this work is called Insights

The June 2026 industry code of ethics reserves the words poll and survey for probability samples, where every person in the population has a known, nonzero chance of being selected. Respondents to this firm’s research choose themselves. That is not a probability sample.

So this firm uses neither word for its own work, in a memo, in a pitch, in a press release or in conversation. The service is called Insights. When the work of other organizations is cited here, it is called a poll, because that is the correct word for it.

  • What self-selected responses can carry: direction, which issues matter and in what order, the language voters use, intensity of feeling, which messages land, and change over time when the same questions run wave after wave.
  • What they cannot carry: a margin of error, any sentence of the form “X percent of voters in this district believe Y,” weighting presented as a repair for self-selection, or a projection of the result on election night.

Every number that gets written down arrives with the count of respondents, how those respondents were recruited channel by channel, and the dates of fieldwork, written out as 03/14/2027 to 03/21/2027. The limit of the study sits on the front page of the deliverable in the same type size as the findings.

Where a question genuinely requires a probability sample, the right answer is a poll from a firm that runs them, and you will hear that.

Find out what your voters actually think

Send the office, the election date and what you have set aside for research. We write back with a scope, a price and a plain statement of what the research can and cannot tell you.

Contact Upset Research