On Point Polling · Party committees
Polling for party committees and caucuses
Parties and caucus committees decide where a whole slate's money goes. That needs more than one poll: a read on the statewide environment, estimates for every targeted district, recruiting research and an honest review after the election. We build programs that answer all of it from one consistent method.
Nonpartisan · Both parties · Client work kept confidential · Fielded in 48 to 72 hours
- D +13.1Our final 2026 national poll, against a D +14.0 count
- 2 of 42Final national polls, ranked by distance from the count
- 7 of 8Senate winners called in our final state polls
- 8 statesAnd the nation polled in 2026, every poll scored in public
The questions you need answered
Each decision, and the research that answers it
- Which seats are really in play?Likely voter modeling and MRPThe Meridian likely voter model on your sample or your universe, and MRP estimates for counties, legislative districts and cities.
- What is the environment statewide?Tracking programsThe same core questionnaire on a schedule you set, weighted to locked targets so the movement you see is real.
- Is this recruit viable?Should I Run viability studyBefore you announce: name recognition, image, the field, the electorate and a path to win, tested on the voters who will decide the race.
- Which voters should the coordinated campaign target?Modeled voter universesYour poll scored onto every record on the voter file: a support and turnout score for each voter, and universes for mail, doors and digital.
- What happened across the slate?Post election autopsyA survey of the voters who just decided the race, matched to the voter file, that shows who turned out, who switched and why.
- How do we follow every race on election night?Election night deskA live county results desk built on your polling, with expected vote and race calls for your audience or your war room.
Across a cycle
How a program usually runs
- 01
Off year
Post election review, recruiting viability studies and a statewide environment benchmark.
- 02
Targeting
MRP estimates for every legislative or congressional district from one statewide sample, and the target list.
- 03
The campaign
Quarterly, then monthly statewide tracking, district polls in the closest seats and modeled universes for the coordinated program.
- 04
Election night and after
A results desk across the slate, then an autopsy that starts the next cycle.
Services
Services for Party committees
Tracking programs
The same core questionnaire on a schedule you set, weighted to locked targets so the movement you see is real.
- Cadence
- Weekly, monthly or quarterly
- Typical sample
- 500 to 1,200 per wave
Likely voter modeling and MRP
The Meridian likely voter model on your sample or your universe, and MRP estimates for counties, legislative districts and cities.
- Inputs
- Survey responses and vote history
- Geography
- State, county, district, city
Should I Run viability study
Before you announce: name recognition, image, the field, the electorate and a path to win, tested on the voters who will decide the race.
- Electorates
- Primary and general
- Typical sample
- 400 to 800 interviews
Modeled voter universes
Your poll scored onto every record on the voter file: a support and turnout score for each voter, and universes for mail, doors and digital.
- Input
- Your poll matched to the voter file
- Output
- Scores for every record
Benchmark and horse race polls
A statewide, district, county, city or national survey on your race and your questions, with leaned likely voter results and full crosstabs.
- Turnaround
- 48 to 72 hours in field
- Typical sample
- 400 to 1,500 interviews
Election night desk
A live county results desk built on your polling, with expected vote and race calls for your audience or your war room.
- Coverage
- Any statewide or national race
- Inputs
- Your polling and the county model
Post election autopsy
A survey of the voters who just decided the race, matched to the voter file, that shows who turned out, who switched and why.
- Timing
- Days to weeks after the election
- Typical sample
- 800 to 1,500 interviews
Voter data and text programs
Voter file datasets for your universe and text programs over the file, for surveys, persuasion and turnout.
- Data
- L2 voter file
- Channel
- SMS and MMS
Good to know
Working with us
One method across the slate
Every district is weighted to the same voter file targets and the same likely voter model, so seat to seat comparisons are fair.
District reads from one sample
MRP carries a statewide poll down to districts too small or too numerous to poll one by one.
Sharing results with candidates
Results a committee gives to a candidate can count as a contribution under FEC rules for federal races and state law for state races. We date every field period and delivery for your compliance staff.
Questions
Before you hire a pollster
Do you work with county parties as well as state parties?
Yes. County and local party committees commission district and countywide polls, autopsies and voter data.
Can one poll cover many legislative districts?
Yes. A statewide sample with MRP estimates every district, and we add oversamples in the seats that matter most.
Is the work nonpartisan?
The firm is nonpartisan and works with both parties. Each committee's work is confidential to that committee.
Tell us the race and the deadline.
Send the office or issue, the geography, the timeline and a budget or sample size. You will have a written proposal and quote the same day.