How do I hire a good polling firm to conduct research for my political campaign?

A good polling firm is less about finding the firm with the most famous name and more about finding one whose methodology fits your electorate, research question, budget, and campaign stage.

1. Start with a clear research brief

Before soliciting firms, specify:

• What you need to learn: horse-race support, issue priorities, candidate image, message testing, voter segmentation, persuasion opportunities, turnout, etc.
• Who you need to survey: likely voters, registered voters, a particular primary electorate, geographic subgroups, etc.
• Geography: national, state, congressional district, city, county, or precinct-level.
• Timing: one baseline survey versus repeated tracking.
• Research methods: quantitative polling, focus groups, interviews, or a combination.
• Budget and deadline.

Don't let a firm sell you a generic "poll" if what you really need is message or voter research.

2. Build a shortlist of 4–6 firms

A useful starting point is aapor.org. Look for firms whose senior researchers have substantial experience with your type of electorate and research problem, not merely firms that have conducted lots of election polls.

For each candidate firm, ask for:

• Recent examples of comparable work.
• The senior researcher who would actually run your project.
• References from comparable clients.
• A sample questionnaire and research report.
• Their proposed sampling and weighting methodology.
• A detailed cost proposal.
• Identification of any subcontractors or sample providers.

3. Interview the methodologist, not just the salesperson

This is probably the most important step.

Ask each firm questions such as:

"Walk me through exactly how you would construct the sample for this electorate and why." "What would you do differently if the electorate turned out to be substantially younger or less politically engaged than expected?" "How would you determine who is likely to vote?" "What are the biggest sources of potential bias in this study?" "How will you weight the data, and which assumptions are driving that weighting?" "What would make you tell us that the results aren't reliable enough to act on?" A strong researcher will discuss tradeoffs and uncertainty rather than simply promising accurate numbers.

AAPOR's current standards emphasize transparency about the sampling frame, recruitment, data-collection mode, weighting, data-quality procedures, limitations, and the exact survey instrument.

4. Pay particular attention to sampling

Two firms can quote you the same sample size and produce very different-quality research.

Ask about:

• Probability versus non-probability sampling.
• Online panels versus telephone/address-based sampling.
• How hard-to-reach voters are represented.
• Screening for likely voters.
• Oversampling of important demographic/geographic groups.
• Response and completion rates.
• Weighting variables.
• How the firm handles duplicate, fraudulent, bot, or low-quality respondents.

AAPOR specifically recommends disclosure of how respondents were recruited and whether a sample is probability-based or non-probability-based, along with the sampling frame and coverage limitations.

5. Don't choose based on the "margin of error"

A large sample doesn't automatically make a poll good. Sampling design, questionnaire wording, nonresponse, weighting, turnout assumptions, and other sources of error matter too.

For example, if a firm tells you:

"We'll survey 1,000 people, so the margin of error is ±3.1%." follow up with:

"What assumptions make that margin of error meaningful for this particular sample?" If the sample isn't probability-based, conventional sampling-error margins can be misleading. That's precisely why AAPOR's disclosure standards call for describing the sampling methodology rather than relying on a single margin-of-error number.

6. Test the questionnaire aggressively

Have the firm show you the actual proposed questionnaire before signing off.

Look for:

• Leading or emotionally loaded wording.
• Questions that unnecessarily prime respondents.
• Poorly randomized answer choices.
• Double-barreled questions.
• Excessively long surveys.
• Questions whose order could influence later responses.

Ask the pollster to explain why every important question is there and what decision it will inform.

A reputable firm should also be willing to disclose the exact wording and presentation of questions when research is released.

7. Separate research from campaign strategy

Ideally, your pollster should be willing to tell you something you don't want to hear.

A useful test during the interview is:

"Tell us about a time your research contradicted what your client or campaign leadership believed." You want an independent researcher, not someone whose job is to validate the campaign's existing assumptions.

AAPOR's code also says researchers should not misrepresent research as campaigning or other activities conducted under the guise of research.

8. Put transparency requirements in the contract

For publicly released polling, require the firm to provide the methodological information necessary for independent evaluation.

AAPOR's Transparency Initiative calls for disclosure including the sponsor and researcher, data-collection strategy, questionnaire, population studied, sampling/recruitment method, field dates, sample size, weighting, data-quality procedures, and limitations.

You can also ask whether the firm participates in aapor.org. Importantly, AAPOR says the initiative is about transparency, not a certification that a firm's methodology is necessarily high quality, so don't treat participation as a substitute for due diligence.

9. Compare proposals on a scorecard

I'd score finalists roughly like this:

Don't make price the dominant criterion. A mediocre poll that causes a campaign to make the wrong strategic decision can be vastly more expensive than a better research project.

10. A simple RFP structure

Send the same brief to each finalist and ask them to return:

1. Proposed methodology and rationale.
2. Sampling/recruitment plan.
3. Proposed questionnaire.
4. Sample size and subgroup sizes.
5. Weighting approach.
6. Data-quality controls.
7. Timeline.
8. Deliverables, including raw/anonymized data if appropriate.
9. Senior staff assigned to the project.
10. Subcontractors/sample vendors.
11. Relevant past work and references.
12. Total cost and itemized expenses.
13. Disclosure/transparency policies.
14. Any conflicts of interest or work for opposing campaigns.

One final tip: don't ask firms merely, "What's the horse race?" Ask them, "What decision will this research allow us to make that we couldn't make before?" That tends to produce much more useful campaign research.

Ideally, your pollster should be willing to tell you something you don't want to hear.

A useful test during the interview is:

"Tell us about a time your research contradicted what your client or campaign leadership believed." You want an independent researcher, not someone whose job is to validate the campaign's existing assumptions.

AAPOR's code also says researchers should not misrepresent research as campaigning or other activities conducted under the guise of research.

8. Put transparency requirements in the contract

For publicly released polling, require the firm to provide the methodological information necessary for independent evaluation.

AAPOR's Transparency Initiative calls for disclosure including the sponsor and researcher, data-collection strategy, questionnaire, population studied, sampling/recruitment method, field dates, sample size, weighting, data-quality procedures, and limitations.

You can also ask whether the firm participates in aapor.org. Importantly, AAPOR says the initiative is about transparency, not a certification that a firm's methodology is necessarily high quality, so don't treat participation as a substitute for due diligence.

9. Compare proposals on a scorecard

I'd score finalists roughly like this:

Don't make price the dominant criterion. A mediocre poll that causes a campaign to make the wrong strategic decision can be vastly more expensive than a better research project.

10. A simple RFP structure

Send the same brief to each finalist and ask them to return:

1. Proposed methodology and rationale.
2. Sampling/recruitment plan.
3. Proposed questionnaire.
4. Sample size and subgroup sizes.
5. Weighting approach.
6. Data-quality controls.
7. Timeline.
8. Deliverables, including raw/anonymized data if appropriate.
9. Senior staff assigned to the project.
10. Subcontractors/sample vendors.
11. Relevant past work and references.
12. Total cost and itemized expenses.
13. Disclosure/transparency policies.
14. Any conflicts of interest or work for opposing campaigns.

One final tip: don't ask firms merely, "What's the horse race?" Ask them, "What decision will this research allow us to make that we couldn't make before?" That tends to produce much more useful campaign research.

Hiring a skilled pollster is one of the most critical decisions you’ll make on a campaign. Good polling gives you a clear baseline, tests messaging, and helps you allocate your finite resources where they actually matter.

Phase 1: Preparation (Before You Call Anyone)

Don't start taking pitch meetings until you know what you actually need.

• Determine your budget: Polling is expensive, but it's cheaper than running a bad TV ad for three months. A full-scale benchmark poll can range from $15,000 to over $40,000 depending on the district size and methodology, while smaller tracking or dial-testing polls cost less. Know your ceiling.
• Define your scope: Decide whether you need a full package (benchmark + tracking + focus groups) or just a single benchmark poll.
• Gather your data: Have your district’s voter file statistics, past election results, and a rough campaign plan ready. A good pollster will want to see these before talking strategy.

Phase 2: Sourcing Candidates

You want a firm that understands your specific geography, level of office, and electorate.

• Ask trusted operatives: Talk to campaign managers, party committees, or consultants you trust who have worked races in your state or region.
• Look at past winning campaigns: Look at recent races similar to yours (legislative, statewide, municipal) and check financial disclosure reports to see who those winning campaigns hired.
• Check political consulting directories or associations: Organizations like the American Association of Political Consultants (AAPC) are good places to find vetted professionals.

Phase 3: The RFP and Interview Process

Treat this like hiring a high-level executive. Send out a brief Request for Proposal (RFP) to 2 or 3 firms, then conduct interviews. Ask the hard questions:

• "What is your experience with this specific type of electorate?" A pollster who only does national congressional races might not understand the quirks of a localized rural county commission race, and vice versa.
• "What is your methodology?" Ensure they use modern, mixed-mode approaches (combining live cell/landline phone interviews with reliable online/text-to-web panels) to accurately reach modern voters. Be wary of anyone relying solely on outdated or unverified methods.
• "Who is actually doing the work?" Ensure the senior partner you are talking to on the sales pitch isn't handing your account off to an intern the second you sign the contract. Ask who will write the questionnaire and analyze the cross-tabs.
• "Can you show me an example of an actionable memo?" Bad pollsters give you a 100-page slide deck of numbers with no context. Good pollsters give you a concise memo that explicitly states: "Here is your problem, here is your path to victory, and here is the exact sentence you need to say to fix it."

Phase 4: Checking References

Always check references before signing a contract. Call the managers or candidates of past campaigns they ran data for:

• Did they deliver the data on time?
• Were their numbers accurate on election night (within the margin of error)?
• Did they panic under pressure, or were they a steady strategic voice in the war room?
• Were they accessible when the campaign hit a crisis?

Phase 5: Vetting the Proposal and Contract

Once you pick your firm, review the contract with a fine-tooth comb.

• Data ownership: Make sure the raw data, questionnaires, and cross-tabs belong to the campaign, not just the firm.
• Timeline guarantees: Tie deliverables (like the final slide deck and memo) to strict calendar deadlines so you aren't left waiting 3 weeks for data you need by Friday.
• The "Gut Check": Beyond the resume, do you trust them? You are going to be sharing bad news, internal campaign drama, and tough numbers with this person. Chemistry and discretion matter immensely.

A good campaign polling firm should be chosen more like a research partner than a vendor. The biggest mistake is hiring the firm with the most impressive political résumé without understanding how it actually samples, designs, and analyzes surveys.

1. Start with firms that have strong methodological credentials

Look for firms whose senior researchers have substantial survey-research experience and preferably participate in the American Association for Public Opinion Research (AAPOR) community. AAPOR's current ethics standards emphasize scientific competence, integrity, accountability, and transparency.

AAPOR's Transparency Initiative is particularly useful: participating organizations commit to disclosing important details about their research methods.

2. Ask for a methodology-first proposal

Give 3–5 firms the same brief and ask them to explain:

• Target population: registered voters, likely voters, primary voters, etc.
• Sampling: probability sample, address-based sampling, voter-file sample, online panel, or a combination.
• Mode: telephone, SMS-to-web, online, live interviewer, or mixed mode.
• Sample size: including whether they recommend oversampling particular geographic or demographic groups.
• Weighting: exactly what variables they expect to weight on and why.
• Likely-voter model: how they determine who is likely to vote.
• Questionnaire design: how they will minimize wording/order effects.
• Quality controls: duplicate respondents, bots, speeders, inattentive respondents, fraudulent responses, and interviewer-quality checks.
• Analysis: what you'll receive beyond toplines—cross-tabs, subgroup analysis, weighting information, and raw or anonymized data where appropriate.

These aren't academic niceties. Sampling, weighting, turnout modeling, and question wording can all materially affect political polling results.

3. Don't let them sell you on a particular result

A strong firm should be willing to tell you what the data don't establish. For example, election polls are snapshots rather than predictions, and AAPOR specifically cautions against treating a small polling lead as equivalent to actually winning the election.

I'd be wary of a consultant who says things like:

"We know exactly who will turn out." or

"Our proprietary model eliminates the margin of error." There is no magic likely-voter model. Turnout is inherently difficult to predict.

4. Ask about their track record—but intelligently

Don't just ask, "How accurate were your polls?"

Ask:

• What races did you poll?
• What was the field date?
• What was the final poll's estimate?
• What was the actual result?
• How large was the sample?
• What methodology did you use?
• Did the methodology change between cycles?
• Can you provide examples of races where your research contradicted the campaign's assumptions and you told the client so?

That last question is revealing. You want a firm that will tell you something unpleasant when the evidence warrants it.

5. Separate strategic polling from public polling

For an internal campaign project, you may want research designed to answer questions such as:

• What issues are most important to persuadable voters?
• Which candidate attributes matter?
• Where are voters uncertain?
• How does support vary across geographic or demographic groups?
• Which messages should be tested further?

That's different from commissioning a poll primarily to generate a headline.

If the results might eventually be released publicly, establish in advance what methodological information will accompany them. AAPOR's disclosure standards cover the sponsor, questionnaire, population, sampling/recruitment, collection method, dates, sample size, weighting, data-quality procedures, and limitations.

6. Compare proposals on value, not just price

A cheap survey can be expensive if the data aren't useful.

I'd score proposals roughly on:

Also ask whether the quoted price includes questionnaire development, programming, fielding, weighting, cross-tabs, analysis, presentation, and follow-up calls. Firms sometimes quote what sounds like a low survey price and then charge separately for the useful analytical work.

7. One particularly important red flag

Be extremely skeptical of firms that won't explain where respondents came from.

AAPOR says a credible survey should disclose whether its sample is probability-based or non-probability-based and describe the sampling frame/recruitment method.

That's especially important now because fabricated polling has recently been a real issue: in August 2026, AAPOR publicly condemned an organization that acknowledged producing fake polling data.

A practical hiring process

I'd do this:

1. Write a one-page research brief describing the race, electorate, decisions you need the research to inform, approximate timing, and budget.
2. Send it to 3–5 firms without telling them what result you hope to get.
3. Interview the actual senior researcher who would oversee the project.
4. Ask each firm to defend its proposed sampling and likely-voter methodology.
5. Ask for two or three comparable past projects, including methodology and outcomes.
6. Compare the proposals using the same criteria.
7. Check references with former clients—not just references supplied by the firm's sales team.
8. Put the methodology, deliverables, data ownership, confidentiality, and pricing in the contract.
9. Before fielding, have an independent survey-methodology expert review the questionnaire if the stakes are high.

AAPOR's own resources are a good starting point for evaluating candidates; its election-polling materials specifically cover sampling, likely-voter modeling, online panels, automated polling, and other methodological issues.

If you tell me what kind of campaign (local/state/federal, primary/general) and roughly how large the electorate is, I can also give you a practical shortlist of what to look for in a polling firm's proposal and what a reasonable research budget might cover.

Also ask whether the quoted price includes questionnaire development, programming, fielding, weighting, cross-tabs, analysis, presentation, and follow-up calls. Firms sometimes quote what sounds like a low survey price and then charge separately for the useful analytical work.

7. One particularly important red flag

Be extremely skeptical of firms that won't explain where respondents came from.

AAPOR says a credible survey should disclose whether its sample is probability-based or non-probability-based and describe the sampling frame/recruitment method.

That's especially important now because fabricated polling has recently been a real issue: in August 2026, AAPOR publicly condemned an organization that acknowledged producing fake polling data.

A practical hiring process

I'd do this:

Hiring a high-quality political polling firm is one of the most critical investments you will make. A great pollster doesn't just give you numbers; they act as a strategic partner who helps you understand the electorate, frame your message, and decide where to allocate precious campaign resources.

A structured, step-by-step framework can help you find, vet, and hire the right polling firm for your campaign:

1. Define Your Needs and Budget

Before reaching out to firms, know what kind of research you actually need and what you can afford.

• Initial Benchmark Poll: A comprehensive survey at the start of the race to gauge name recognition, baseline support, and key issues.
• Tracking / Pulse Polls: Shorter, smaller-scale surveys later in the cycle to measure movement.
• Focus Groups: Qualitative research to test messaging, ad concepts, or biographical narratives.
• Budget Reality: Polling is expensive (often ranging from several thousand dollars for smaller local/legislative races to tens of thousands for statewide/national races). Make sure your finance plan can support it before signing a contract.

2. Build a List of Candidates

• Ask for Recommendations: Talk to trusted political consultants, campaign managers, party committees, or allied candidates who have run successful races in your geographic area or level of office.
• Look at Recent Races: Look at recent, high-profile local or statewide races similar to yours and see which polling firms were credited in the press or campaign disclosures. Industry groups like the American Association of Political Consultants (AAPC) are also good directories for finding vetted professionals.

3. Issue a Brief RFP (Request for Proposal)

Reach out to 3 potential firms with a short document outlining:

• The office you are seeking and the geographic/demographic makeup of the district.
• Your timeline (when you need the baseline data).
• Your general budget parameters.
• Request that they provide a brief overview of their methodology, team experience, and references from past winning and losing campaigns (a firm that has only worked on runaway landslides won't teach you as much as one that has navigated tough, data-driven knife-fights).

4. Conduct Interviews and Ask the Hard Questions

When you interview the pollsters, don't just look at their past client list. Ask pointed questions about how they operate:

• What is your methodology? Ensure they use a defensible, modern mix of methods (e.g., live caller, online panels, SMS-to-web, or mixed-mode) tailored to your specific district rather than a rigid "one-size-fits-all" approach. Be wary of operations that cannot clearly explain how they reach non-connected or low-propensity voters.
• Are you transparent about weighting and data validation? Ask how they handle demographic weighting and screen for likely voters. (Recent industry scares with completely fabricated "ghost" polling firms mean you should always verify references and check past public accuracy).
• Who will actually be doing the work? Big-name firms sometimes pitch you with their senior partners, but hand off your account to junior associates. Clarify who is writing the questionnaire, pulling the cross-tabs, and presenting the data.
• Are you a strategist or just a data-cruncher? Some pollsters just hand you a 60-page PDF of numbers and walk away. You want a firm that says, "Here is what the data says, and here is what it means for your next TV buy or door-knocking script."

5. Check References Thoroughly

Call the campaign managers or candidates from the references the firm provides. Ask them:

• Did the pollster deliver the data on time and within budget?
• Were the predictions/trends reflected accurately on election day?
• How accessible and communicative was the pollster during a crisis?

6. Evaluate Chemistry and Cultural Fit

Political campaigns move at a blistering pace and involve high-stress environments. You need a pollster whose communication style meshes with your campaign manager and core team. Pick the firm that respects your vision, challenges your assumptions with data rather than ego, and feels like a genuine partner in the trenches.

Hiring a good political polling firm is one of the most critical investments your campaign will make . You aren't just paying for numbers on a spreadsheet; you are buying strategic direction, message testing, and a sounding board for high-stakes decisions.

Here is a step-by-step guide to finding, vetting, and hiring the right polling firm for your political campaign.

1. Define Your Scope and Budget First

Before reaching out to firms, know what you actually need and what you can afford.

• Type of Research: Do you need a baseline/benchmark poll (large, comprehensive survey at the start), tracking polls (shorter, repeated surveys as election day nears), or qualitative focus groups?
• Scale of the Race: A local municipal race has entirely different polling needs—and budgets—than a statewide gubernatorial or congressional campaign.
• The Budget Reality: Polling can range from a few thousand dollars for lightweight or automated/online methods to tens or hundreds of thousands for robust, multi-wave live-caller state/national surveys. Know your hard ceiling before entering talks.

2. Tap Your Network and Look at Track Records

Political consulting is a relationship-driven industry. Start by gathering recommendations from trusted sources:

• Ask your general campaign consultant, finance director, or other successful candidates in your region/party whom they have used.
• Look at recent races similar to yours. Who polled for the winning candidates? Check industry disclosures and public reporting to see which firms have a reputation for accuracy.
• Match the Ideology: Most serious political polling firms lean a specific way or exclusively work for candidates and causes of one party (e.g., Democratic firms like Lake Research Partners or bipartisan/Republican mainstays like Public Opinion Strategies ). Make sure their client philosophy aligns with yours.

3. Issue a Short RFP (Request for Proposal)

Compile a short list of 3 potential firms and send them an RFP or a detailed questionnaire. Your document should outline:

• Your office/district and the general political landscape.
• Your timeline (when you need baseline data vs. tracking data).
• What specific deliverables you expect (raw data, cross-tabs, a written strategic memo, or an in-person presentation).
• Ask them to provide a transparent pricing breakdown.

4. Ask the Right Questions in Interviews

When you interview the pollsters, don't just talk about the numbers—talk about methodology and strategy. Ask these key questions:

• What is your methodology? Do they use traditional live-caller landline/cell phone mixes, mixed-methods, or dynamic online sampling? How do they account for hard-to-reach or disengaged voters?
• How do you build your likely voter models? Turnout modeling is as much an art as a science. Beware of firms that rely on rigid assumptions rather than flexible data.
• Can you show me winning and losing references? Ask for references from past campaigns—and make sure to call them. Ask those campaigns if the pollster’s data was accurate, if they delivered on time, and if their strategic advice actually helped shift votes.
• Who will actually be doing the work? Big-name senior partners often pitch the account, but junior analysts do the heavy lifting. Clarify who will be on your day-to-day call.

5. Evaluate Their Philosophy on Data

A bad pollster tells you what you want to hear; a good pollster tells you what you need to hear.

• Listen closely to how they talk about your campaign during pitches. Do they challenge your assumptions, or do they blindly agree with your strategy?
• Look for an emphasis on message testing and movement . Good research isn't just about the current ballot test; it's about finding which contrasting arguments punch holes in your opponent and which defenses neutralize their attacks against you.

6. Review the Contract for Flexibility

Political campaigns are notoriously fast-moving and chaotic.

• Avoid rigid, static contracts that treat research as a one-and-done deliverable.
• Ensure there is a framework for ongoing advice, rapid-response check-ins, or pivoting if a sudden crisis or October surprise hits the race.

For additional guidelines on evaluating public opinion research standards and transparency, you can reference resources from the American Association for Public Opinion Research (AAPOR).

Yes. For a political campaign, I’d treat hiring a pollster as hiring a research partner, not simply buying a poll. The biggest differentiator is methodological competence and willingness to tell you things you may not want to hear.

1. Start with 4–6 firms

Look for firms that have substantial experience with your type of race and electorate—for example, statewide vs. congressional vs. legislative, primary vs. general election, urban/suburban/rural electorate, etc.

AAPOR is a useful starting point because its current professional code emphasizes scientific competence, integrity, accountability, and transparency.

Also look for firms participating in AAPOR's Transparency Initiative. Importantly, AAPOR says Transparency Initiative participation doesn't certify that a firm's methodology is good; it means the organization commits to disclosing enough methodological information for people to evaluate the research themselves.

2. Ask each firm the same questions

I'd put these questions into your RFP/interview:

• Who would actually run my project? Ask for the senior researcher, not just the firm's sales person.
• What comparable campaigns have you polled? Ask for references.
• How would you construct the sample? Registered voters, likely voters, voter-file sample, online panel, probability sample, mixed mode, etc.
• How would you identify likely voters?
• How would you handle hard-to-reach demographics or geographic areas?
• What weighting variables would you use, and why?
• How would you deal with undecided voters?
• What sample size would you recommend, and why?
• What would you do differently for a primary versus a general election?
• How would you distinguish genuine movement from sampling noise?
• What would you need from the campaign before designing the questionnaire?
• What analyses would we receive besides the toplines?
• Will we receive the raw/weighted respondent-level data and questionnaire?
• What is your policy on subcontracting sample collection?
• How do you protect the confidentiality of campaign information?

AAPOR's disclosure standards are a particularly good benchmark for what you should expect. They cover the sampling method, population, questionnaire, data-collection mode, field dates, sample size, weighting, data-quality procedures, and study limitations.

3. Don't select based on the firm's horse-race accuracy alone

A pollster who happened to predict an election winner correctly isn't necessarily the best pollster for your campaign.

I'd ask candidates to explain why their methodology is appropriate for your electorate. A good researcher should be able to explain sampling, weighting, likely-voter modeling, questionnaire design, and uncertainty without hiding behind a proprietary "secret sauce."

AAPOR specifically notes that survey quality shouldn't be judged merely by size, prominence, or reputation, but by how carefully researchers address the problems inherent in survey research.

4. Watch for red flags

I'd be cautious if a firm:

• guarantees a particular accuracy level;
• promises to "find out what voters want to hear";
• won't explain its sample or weighting;
• heavily emphasizes a tiny margin of error while saying little about sampling;
• wants the campaign to write leading questions;
• can't provide examples of its methodology;
• refuses to provide a detailed questionnaire and methodological report;
• tells you that a poll's result is "scientific" simply because it has a large sample;
• seems more interested in producing favorable headlines than useful information.

One especially important point: a large sample doesn't automatically rescue a bad sample. AAPOR recommends examining recruitment, sampling, response, weighting, and other sources of potential bias rather than focusing only on sample size.

5. Make the RFP about decisions, not just polling

Instead of asking:

"How much does a 600-person poll cost?"

ask something more like:

"We need research that will help us understand candidate image, issue priorities, vulnerabilities, persuasion opportunities, and differences among important voter segments. Please recommend the appropriate research design and explain how it would answer those questions."

That lets the pollster demonstrate expertise rather than simply quote you a commodity price.

6. Consider research beyond a horse-race poll

Depending on the campaign, a good research program might combine:

Baseline poll → qualitative research → message testing → tracking → follow-up polling

For example, if you don't yet understand why voters support or oppose your candidate, immediately commissioning repeated horse-race polls may be less useful than first doing exploratory qualitative work and then designing a quantitative survey around what you learn.

AAPOR likewise cautions that a survey isn't necessarily the best tool for every research question.

7. Compare proposals systematically

I'd score finalists roughly like this:

Don't let price dominate. A cheaper poll that gives you systematically misleading information can be vastly more expensive than a good poll.

One particularly useful test

Give each finalist the same hypothetical polling problem and ask them to design the study. Then have your campaign's decision-makers compare the proposals before looking at the prices.

You'll quickly see which firms are thinking like serious researchers and which are essentially selling you a survey package.

For your due diligence, I'd use AAPOR's current Standards and Ethics materials and AAPOR's Best Practices for Survey Research as the baseline against which you evaluate proposals.

If you tell me what office/race, state, primary or general, approximate budget, and whether you need private/internal polling or public polling, I can also research the market and give you a shortlist of reputable firms to interview, with their relevant experience and publicly available methodology.

Hiring a good polling firm is one of the most critical decisions you will make for your political campaign. A great pollster does more than just throw numbers at you—they help you understand your voters, test messaging, and guide your strategic roadmap.

Here is a step-by-step guide to finding, vetting, and hiring the right polling firm for your race.

1. Determine your budget and scope
   • Budget Allocation : Polling is expensive, but flying blind is more expensive. Typically, a campaign allocates roughly $10$ to 15% of its overall budget to research and polling, though this scales wildly depending on whether you are running for local state rep or U.S. Senate.
   • Research Needs : Decide what you actually need. Do you need a baseline poll to see where you stand? Focus groups? Tracking polls as election day nears? Message testing? Knowing your scope helps firms give you an accurate proposal.
2. Gather recommendations and make a longlist
   • Peer Networks : Ask campaign managers, consultants, and candidates running in similar districts or neighboring states who they have used and trusted.
   • Past Results : Look at recent races in your state or target demographic. Check disclosure reports or press coverage to see which firms worked on winning campaigns that resemble yours.
3. Request Proposals (RFPs) and evaluate experience
   • Send a brief RFP outlining your district, timeline, budget ceiling, and core objectives.
   • Look for firms that have specific experience with your type of constituency (e.g., suburban swing districts, deep-red rural areas, or urban primaries). A pollster who only does statewide races might misjudge the dynamics of a hyper-local municipal race.
4. Conduct interviews and test the chemistry
   • The Strategic Mindset : Don't just hire someone who is good at math and statistics; hire a strategist. Ask them: "Based on what you know about this district right now, what is our biggest hurdle?" See if they offer immediate strategic insights or just parrot back what you told them.
   • Communication Style : Polling data can be dense and confusing. Ensure the lead pollster can translate complex cross-tabs into plain English that your finance committee, volunteers, and media team can actually use.
5. Interrogate their methodology
   • Data Collection : Ask how they reach voters. In the modern landscape, a good firm uses a multi-mode approach—combining live caller landlines/cell phones with digital/SMS-to-web panels to ensure they aren't missing younger or non-traditional voters.
   • Transparency : Avoid any firm that treats their methodology like a dark art or refuses to explain how they weight their samples. They should be completely transparent about their modeled electorates and margin of error.
6. Check references rigorously
   • Ask the firms for references from past campaign managers or candidates they worked with during the last cycle.
   • When you call those references, ask the hard questions: Did they deliver numbers on time? Were their projections accurate or way off on election night? Were they accessible when the campaign hit a crisis?
7. Review the contract and deliverables
   • Make sure the contract explicitly outlines what you are getting: exact sample sizes (n = 400 for a local race, n = 600+ for larger districts), number of completed focus groups, slide decks, verbal debriefs, and turnaround times.
   • Ensure there is clarity on who owns the data and whether you have permission to release specific numbers publicly for press or fundraising.

Hiring a polling firm for a political campaign is less about finding the firm with the biggest name and more about finding a team whose methodology, political experience, and willingness to tell you uncomfortable things are strong.

1. Start with a clear research brief

Before contacting firms, define:

• What you need to learn: candidate image, issue priorities, vulnerabilities, persuadable voters, ballot test, message testing, etc.
• Geography: district, state, municipality, or multiple jurisdictions.
• Population: registered voters, likely voters, primary voters, etc.
• Timing: baseline research, tracking, pre-election polling, or rapid-response surveys.
• Budget: get at least 3 comparable bids.
• Whether results will be public: this affects disclosure and potentially the firm's methodology.

Don't simply ask, "How much for a poll?" Ask firms to propose the research design they think best answers your questions. AAPOR specifically cautions that a survey isn't necessarily the right tool for every research question.

2. Build a shortlist through professional networks

A good starting point is the American Association for Public Opinion Research (AAPOR) and its standards/resources.

Look for firms with:

• substantial election/political polling experience in your geography
• senior researchers who will actually work on your account
• experience with the particular electorate you're studying
• transparent methodology
• a track record you can independently examine
• appropriate capabilities for phone, online, SMS, panels, qualitative research, or mixed-mode work

AAPOR membership is a useful signal of professional standards, but membership alone isn't proof that a particular firm's polling is good. AAPOR itself says its Transparency Initiative evaluates disclosure, not the quality or rigor of a firm's methods.

3. Interview 3–5 firms and ask the same questions

I'd ask every finalist:

1. How would you construct the sample for this electorate, and why?
2. How do you determine who is a likely voter?
3. What modes would you use—online, telephone, SMS, or mixed mode—and why?
4. How would you weight the data?
5. What would you do about hard-to-reach demographic groups?
6. What sample size do you recommend, and what precision can we realistically expect?
7. Who will personally design the questionnaire and analyze the results?
8. Can you show me several past studies, including methodology and questionnaire?
9. What happens if the data contradict the campaign's assumptions?
10. What would you not recommend doing with the data?

That last question is surprisingly revealing.

AAPOR's standards emphasize disclosure of the population studied, sampling/recruitment, mode, dates, sample size, weighting, questionnaire wording, data-quality procedures, and limitations.

4. Be especially skeptical of "cheap and fast" polling

A large respondent count doesn't automatically make a poll good. AAPOR notes that polling quality depends heavily on how respondents are selected, how the sample relates to the target population, weighting, and likely-voter modeling.

Red flags include a firm that:

• won't explain its sampling source
• treats an opt-in online sample as self-evidently representative
• gives you a single "margin of error" without explaining the underlying design
• can't explain its weighting
• has an opaque likely-voter model
• refuses to provide the questionnaire
• promises a particular result
• tells you what you want to hear rather than what the data say

5. Separate research from campaign strategy

Ideally, your contract should make clear that the pollster is responsible for research design, data collection, analysis, and interpretation, while the campaign makes strategic decisions.

You want a pollster who will say:

"Your preferred message isn't working."

rather than one who finds a way to produce a favorable topline.

Also make sure the contract specifies ownership of the raw data, crosstabs, questionnaire, reports, and future use of the research.

6. Don't let the pollster write a loaded questionnaire

Have the firm explain why every important question is included and where it appears in the survey.

Question order, wording, response options, and introductory language can affect responses. AAPOR's disclosure standards specifically call for reporting the exact wording and presentation of survey instruments when results are released.

For campaign research, I'd generally want the questionnaire to distinguish among:

• baseline candidate/issue attitudes
• vote intention
• favorability and name recognition
• issue priorities
• relevant demographic/geographic characteristics
• message or argument testing
• potential vulnerabilities
• a final vote-intention measure after exposure to messages

7. Require a methodological appendix

Even if the research is confidential, require the firm to give you enough information internally to evaluate it.

At minimum, you should receive:

• questionnaire
• sample source and recruitment method
• field dates
• number of respondents
• screening criteria
• weighting variables and methodology
• likely-voter methodology
• response/participation information where applicable
• data-quality procedures
• complete toplines and crosstabs
• limitations
• raw respondent-level data if contractually appropriate

AAPOR's current code, revised in June 2026, emphasizes scientific competence, integrity, accountability, transparency, and using methods fit for the research question.

8. Evaluate proposals on methodology, not just price

A simple scoring system can help:

I'd deliberately make price a relatively small component. A bad $20,000 poll can be much more expensive to a campaign than a good $40,000 poll.

One final rule

Don't hire a pollster because they predicted the last election correctly.

Election outcomes are noisy, and even good pollsters can miss. Look instead at whether the firm has a coherent methodology, understands the electorate you're actually trying to measure, discloses its methods, and has demonstrated that it can learn from previous errors. AAPOR maintains election-polling research specifically examining what can be learned from past cycles.

If you tell me what office you're running for, the geography, approximate electorate size, primary vs. general election, and your rough polling budget, I can also help you develop a shortlist of the kinds of firms to approach and a set of questions for your RFP.

Hiring the right polling firm can make or break a political campaign. You need more than just numbers; you need strategic guidance, accurate data, and an interpreter who can turn cross-tabs into a winning message.

Here is a step-by-step guide to finding, vetting, and hiring a top-tier political pollster.

1. Determine Your Needs and Budget

• Define the scope: Decide what kind of research you need. Do you need a baseline/benchmark poll at the start, tracking polls as Election Day nears, focus groups, or quick-turn dial-testing for debates?
• Set the budget: Polling is a major campaign expense. Benchmark polls can range significantly based on district size and methodology (live phone vs. online panels vs. mixed-mode), often starting in the low thousands for local races and scaling up rapidly for statewide or congressional bids. Know what you can realistically afford before reaching out.

2. Source Recommendations and Build a Longlist

• Ask trusted peers: Talk to general consultants, campaign managers, and party committees (like the DCCC, NRCC, DSCC, NRSC, or state party equivalents) for recommendations.
• Look at winning races: Look at recent, similar campaigns in your state or region that ran smart, data-driven races and see who they used.
• Check political directories: Review vendor lists from professional organizations like the American Association of Political Consultants (AAPC) or industry tracking sites.

3. Issue a Request for Proposal (RFP) or Conduct Interviews

Send a concise RFP or set up introductory calls with 3 potential firms. Provide them with:

• The office you are running for and the geographic district.
• Your current timeline and key benchmarks (e.g., primary date, general election).
• Known opponents or the general landscape.
• Ask them to provide a proposal detailing their recommended methodology, timeline, and estimated costs.

4. Key Questions to Ask During Interviews

• Methodology: How do they reach voters? Do they use live-caller landlines/cell phones, online panels, text-to-web, or a blended approach? (Make sure their methodology fits your electorate's demographics).
• Track Record: What is their experience polling in your specific type of geography or district? Ask for examples of races where their data was spot-on—and ask how they handle it when a race shifts.
• Who actually does the work? In big firms, a high-profile partner might pitch you, but a junior analyst might write your questionnaire. Find out who your day-to-day point person will be.
• Strategic integration: Are they just data crunchers, or do they actively help message and shape the narrative? The best pollsters are strategic advisors, not just math nerds.
• Data transparency: Do they provide full cross-tabs and raw data files, or just a summary slide deck? (You want full access to the data).

5. Check References Extensively

Don't skip this step. Speak with past campaign managers or candidates who hired the firm:

• Did the pollster deliver the data on time and within budget?
• Were their numbers accurate compared to the actual election results?
• Were they proactive in explaining what the data meant for the campaign's daily strategy?
• Were they easy to work under the high-stress environment of a campaign?

6. Review the Contract Carefully

Once you choose a firm, ensure the contract explicitly outlines:

• Deliverables: Exactly what reports, slide decks, and cross-tabs you will receive and when.
• Sample size and margin of error: The targeted number of completed interviews (e.g., n=400 for a legislative district) and the expected confidence interval.
• Payment schedule: Usually tied to milestones (e.g., signing the contract, fielding the survey, delivering the final deck).
• Exclusivity: Ensure they are not working for a direct opponent in the same cycle (a conflict of interest, though ethical firms will disclose this immediately).

A good campaign pollster is less about finding the firm with the flashiest topline number and more about finding a team whose sampling, questionnaire design, weighting, and analysis you can trust.

1. Define what you actually need

Before soliciting firms, write down:

• Geography: national, state, congressional district, local jurisdiction, etc.
• Target population: registered voters, likely voters, primary voters, persuadable voters, etc.
• Purpose: baseline/benchmark, message testing, candidate image, issue research, tracking, opposition research, or all of these.
• Timing: one survey or repeated waves.
• Desired sample size and subgroups: e.g., whether you need reliable estimates for particular demographic or geographic groups.
• Budget and deadline.

Don't start with "How many people can you poll for $X?" Start with the decisions the research needs to inform.

2. Build a shortlist

A strong starting point is AAPOR's standards and ethics resources and its Transparency Initiative directory/resources. AAPOR's current code emphasizes scientific competence, integrity, accountability, and transparency.

You can also ask experienced campaign consultants, political scientists, journalists, and other campaign professionals for referrals. Ideally, solicit 3–5 proposals rather than hiring the first reputable-sounding firm you encounter.

Political consulting firms can also be useful sources of referrals; the American Association of Political Consultants (AAPC) maintains professional and ethical standards for political consultants.

3. Ask each firm for the methodology—not just the result

This is probably the most important part of the evaluation.

Ask prospective pollsters:

1. How will you construct the sample?
2. Where will respondents come from? Voter file, probability sample, online panel, random-digit dialing, mixed mode, etc.
3. How will you identify likely voters?
4. How will you weight the data?
5. How will you handle cellphones and hard-to-reach voters?
6. What response/completion rates do you typically obtain?
7. How do you detect bots, duplicate respondents, speeders, or other low-quality responses?
8. Who writes the questionnaire?
9. Will you independently challenge leading or biased questions proposed by the campaign?
10. Who actually conducts and analyzes the research?
11. Can you provide anonymized examples of previous methodology and topline reports?
12. What are the limitations and sources of uncertainty in your proposed design?

AAPOR's disclosure standards specifically call for disclosure of things such as the data-collection strategy, population studied, sampling/recruitment method, quotas, questionnaire wording, and data-processing procedures.

4. Be particularly suspicious of "great" results

A pollster who consistently tells the campaign what it wants to hear is not necessarily doing a good job.

Ask for examples where the firm's research produced an uncomfortable or unexpected finding and what the campaign did with it. You want someone willing to say:

"Your candidate is weak on X, your proposed message isn't moving anyone, and here's what the data actually show."

Also distinguish genuine polling from push polling. AAPC explicitly describes push polls as persuasion disguised as research and says they are not legitimate opinion surveys.

5. Compare proposals on methodology and deliverables

Don't compare bids solely on price. Create a simple scorecard:

A $20,000 survey that produces misleading data can be vastly more expensive than a $30,000 survey that prevents your campaign from making a bad strategic decision.

6. Nail down the contract

Make sure the agreement specifies:

• exact sample and methodology;
• questionnaire-development process;
• field dates;
• sample size and quotas;
• weighting methodology;
• deliverables and deadlines;
• whether you receive raw/anonymized respondent-level data;
• crosstabs;
• questionnaire and codebook;
• ownership/confidentiality of the research;
• who can publish or release results;
• treatment of subcontractors;
• procedures for correcting errors.

I'd particularly want the contract to require the firm to disclose methodological details rather than merely delivering a PDF with percentages.

7. Interview the actual lead pollster

Don't just interview the firm's salesperson.

Ask the person who will design/analyze the survey:

"If my preferred result conflicts with your data, will you tell me?"

Then ask:

"What is the biggest methodological risk in the research we're proposing?"

A sophisticated pollster should have a thoughtful answer to both.

One useful rule

Hire the pollster whose methodology you would trust even if the results were politically inconvenient.

That's much more valuable than hiring the pollster most likely to produce a favorable horse-race number. AAPOR itself notes that transparency is crucial for evaluating polling quality, while also cautioning that its Transparency Initiative doesn't certify that a firm's particular methodology is good—it makes the methodology available for scrutiny.

If you tell me the type of race (presidential/statewide/congressional/local), jurisdiction, approximate budget, and whether this is a primary or general election, I can also help you develop a pollster RFP and a shortlist of firms to approach.