One system from research question to defensible answer.
Specify a study, configure the audience, check feasibility before you commit, field it with live monitoring, and close with a result that carries its own methodology. No stitching together a panel vendor, a survey tool, and a spreadsheet.
The problem
Fragmented research is slow because it is fragmented
A typical study touches four vendors: someone to design the instrument, someone to supply the sample, someone to host the field, and someone to reconcile the output. Every handoff loses context, and none of them owns whether the result holds up.
The cost is not only time. When the number is challenged, nobody in the chain can produce the full record — who qualified, what was asked, what changed mid-field — because no single system ever held it.
What you can run
Ten study types, one workflow
The research question changes; the workflow does not. Each study type below uses the same specification, audience configuration, feasibility gate, and reporting artifact — which is what makes results comparable across studies instead of only within one.
Benchmark
Establish the baseline you will measure everything else against.
Message testing
Compare framings before spend commits you to one of them.
Rapid reaction
Read response to an event while the decision window is still open.
Issue salience
Separate what people say they care about from what moves them.
Candidate image
Track favorability and attribute perception, not just topline support.
Ad testing
Evaluate creative against the audience it is actually aimed at.
Persuasion
Measure movement, not just standing preference.
Tracking
Repeat the same instrument over time so change is real, not noise.
Policy tradeoffs
Force prioritization instead of collecting agreement with everything.
Custom studies
Designed with our research team when the question does not fit a template.
How a study runs, end to end
- 01
Create the study
Two doors into one system: build the study manually, or start from an AI-assisted design session.
Both paths produce the same structured specification, so a study built by hand and one drafted with assistance are reviewed, versioned, and reported identically.
- 02
Configure the audience
Define geography, eligibility, demographics, civic attributes, quotas, oversamples, and screeners.
Audience definition is part of the specification rather than a vendor's internal choice, so the population you asked for is the population on record.
- 03
Check feasibility before launch
See qualified audience, incidence, achievable sample, expected field time, incentive, and cost first.
Feasibility is reviewed against the verified audience before anything fields, so a study is repriced or rescoped instead of failing quietly in the field. Estimates are estimates — we do not guarantee completion windows.
- 04
Field with live monitoring
Watch completion, quota fill, and integrity flags as responses arrive.
Problems surface while they are still fixable: a quota that will not fill, a screener that is too tight, or activity held for integrity review.
- 05
Close and report
The study closes on schedule and reports with its specification and method version attached.
Results publish as an artifact with provenance rather than a slide, so the number can be re-examined later on its own terms.
Audience configuration
You define the population; the platform holds you to it
Sample structure is a decision, not a post-hoc correction. Audience inputs are declared up front and recorded with the study, so the population you specified is the population the result is reported against. Targeting is never narrowed to the point where an individual could be identified.
Geography
National, state, district, or media market.
Eligibility
Whether the study requires the actual voting population or general adults.
Demographics
Standard population attributes used for representativeness.
Civic attributes
Participation history and civic engagement characteristics.
Quotas & oversamples
Structure the sample deliberately instead of correcting for it later.
Screeners
Qualify participants into the study before any substantive question.
Before you commit
Feasibility is a gate, not a hope
Before a study fields, the design is checked against the verified audience. If it is not fieldable, you learn it while it is still cheap to change. These are estimates, not guarantees — what we commit to is that you see them before spending, and that the assumptions behind them stay attached to the study.
- Verified audienceParticipants eligible to be invited at all
- Population matchGeography, eligibility, and demographic fit
- Screener incidenceShare who qualify into this specific study
- Achievable sampleWhat can realistically complete in the field window
Reviewed before any spend commits
- Qualified audience
- Expected incidence
- Achievable sample
- Projected field time
- Incentive required
- Resulting cost
Structure of the feasibility readout. Bar widths illustrate narrowing, not measured values — every figure is computed per study against the verified audience before launch.
What holds a fielded study together
- Participation requires identity verification, so responses come from confirmed real people rather than anonymous traffic.
- Every study runs behind integrity controls that detect duplicate, coordinated, and low-quality participation before results are published.
- Published results carry provenance: the specification, method version, and computation run behind a number are recorded and inspectable.
- Methodology is stated in plain language — sampling frame, weighting approach, margin of error, and known limitations travel with every result.
- Verified audiences and automated fielding compress the time between a research question and a defensible answer.
- BetaThe platform supports advanced research designs including trackers, recontact studies, experiments, MaxDiff, conjoint, and pre/post measurement.
Feasibility figures are estimates and field windows are not guaranteed. Tracking, policy-tradeoff and other advanced designs are enabled per account as each method reaches production availability — the platform will tell you which are available to you rather than implying all of them are.
What backs this up
Verified audiences and automated fielding compress the time between a research question and a defensible answer.
See published resultsParticipation requires identity verification, so responses come from confirmed real people rather than anonymous traffic.
How verification worksThe platform supports advanced research designs including trackers, recontact studies, experiments, MaxDiff, conjoint, and pre/post measurement.
Published results carry provenance: the specification, method version, and computation run behind a number are recorded and inspectable.
Open a result with its methodologyEconomic value
What you stop paying for
Consolidating design, audience, fielding, and reporting removes the coordination tax: no vendor reconciliation, no re-fielding because a screener was wrong, no separate spend to prove a result was sound.
Studies get cheaper to defend and cheaper to repeat, which is what makes measuring change over time affordable instead of aspirational.