Market research software with its own respondent panel keeps recruiting, respondent verification and the audit record with one company, so you review one screening standard. Buying survey sample separately widens reach for B2B and low-incidence audiences, with a standard per provider to review. Your study's audience decides which fits. aytm reviewed 1.1 million survey attempts in Q1 2026.
The figures in this article come from the Q1 2026 edition of the aytm Data Quality Benchmark Report. We'll update them when the next edition publishes in November 2026.
Contents
- What is a respondent panel?
- What should respondent verification cover?
- What should you ask a sample provider about respondent verification?
- What does owning the panel let a vendor measure?
- How does a built-in panel compare with buying sample separately?
- When is buying sample separately the right call?
- How does aytm verify its respondent panel?
- FAQ
- Before your next study
Every research platform you evaluate says its data quality is strong, so you've learned to ask where the survey sample comes from before you take that at face value. Some market research software comes with a respondent panel the platform recruits itself. Other software sources survey sample from outside providers, study by study.
Many insights teams already work with more than one sample provider, and lining their screening standards up against your own takes real time. The question usually lands after fieldwork, when a stakeholder asks how the sample was screened and you're the one who answers. We're going to go through what a respondent panel is, what verification should cover, what owning the panel lets a vendor measure, and when each path fits your study's audience.
What is a respondent panel?
A respondent panel is a standing group of people who've agreed to take surveys, and the company that recruits them decides how they're checked. A proprietary panel is one a company owns and recruits directly. When that panel comes inside the market research software you field with, it's called a built-in panel. The other path is buying survey sample from a sample provider for each study, sometimes through a sample marketplace that pools several panel providers. On either path, panel quality depends on who checks each respondent and what they can show you afterward. That makes ownership of respondent verification the question to put to every platform on your list.
What should respondent verification cover?
Good respondent verification covers four jobs, in order: reviewing the survey, screening who arrives, checking the answers and recording what passed. A design review looks at the questionnaire before launch, because a survey that wears people down gets careless answers from honest respondents. Entry screening looks at the traffic arriving at a survey and checks each person's identity before their answers count. Answer checks look at the responses themselves, for patterns like speeding or failed attention checks. The audit record then lists which checks ran, so you can review them before analysis and a colleague can follow your work.
Each job also has a limit. An identity check tells you who answered and says little about how carefully they answered. An answer check catches the patterns it was built to look for, so ask what those are.
One design choice runs across all four. A check can block a respondent outright, or it can warn and leave the call to a person. Blocking on a single signal can drop honest respondents and change who's left in your sample, so ask which checks warn, which block, and who acts on a warning.
What should you ask a sample provider about respondent verification?
Ask every vendor the same five questions, so you can line their answers up side by side:
- Who recruits the respondents in my study, and who verifies their identity?
- What checks run before a respondent enters my survey?
- What checks run on the answers, and what triggers a removal?
- Do you publish your removal rates?
- What record comes with each study, and which responses does it cover?
These five focus on verification. For the broader set, our guide to comparing survey panels on data quality and fraud protection asks seven questions that cover every quality rate a panel should publish. ESOMAR, the global association for market research, publishes 37 questions to help buyers of online samples, and we've published our answers to all 37.
What does a strong answer include?
What does owning the panel let a vendor measure?
Every quality rate needs two things beside it before you compare it: its base, meaning what the rate is out of, and its period, meaning when it was measured. Our guide to comparing survey panels on data quality and fraud protection walks through the core rates any panel should publish, including pre-survey removal and post-survey cleanout. Two more rates depend on who owns the panel, and both come from the aytm Data Quality Benchmark Report (Q1 2026).
Incorrect demographics: how often does a verified respondent's profile fail to match?
Incorrect demographics is the share of verified respondents whose answers inside a survey don't match the demographics on their verified profile. Only a company that holds verified profiles can measure it, which makes it the clearest test of whether a panel's verification holds up once people start answering. A rise can mean the check got stricter, so ask what changed before you read the trend. aytm's incorrect demographics rate was 1.6% of verified respondents in Q1 2026, up from 0.6% a year earlier, because verification coverage widened and a stricter test found more.
Duplicate attempts: what does a panel owner catch in outside sample?
A duplicate attempt is the same person trying to enter a survey more than once. A company that owns a verified panel can check arriving outside sample against its own identity records before fieldwork begins, which is how it catches duplicates in traffic it didn't recruit. A peak month can sit far from a typical one, so ask which traffic a duplicate rate covers and over what period. In Q1 2026, third-party duplicate attempts peaked at 46% in a single month, a figure that covers outside sample only and excludes PaidViewpoint, aytm's own respondent panel.
How does a built-in panel compare with buying sample separately?
The two paths differ on who does the work and how many standards you end up reviewing.
A built-in panel puts recruiting and respondent verification with the company that runs your software, so the respondents and the checks on them come from one place. Buying sample separately widens the pool a study can draw on, and each sample provider applies its own screening standard for you to review. Two limits apply on either path. A record shows only the checks that ran where it was produced, so a blended sample needs one from each source. A removal rate describes one period, so ask for more than one quarter before you compare two vendors.
When is buying sample separately the right call?
Buying sample separately is often the practical choice, and the reason is usually reach. A niche B2B audience or a low-incidence category can need more people than one panel offers at volume, and blending in outside survey sample is often how a team hits quota. Some teams also have years of history with a sample provider, and that relationship has value worth weighing. The deciding question stays the same on either path: does each provider's respondent verification meet the bar you're willing to put your name on, study by study? The five questions earlier in this piece give you a consistent way to find out.
What does a blended B2B study look like?
Say your study needs IT decision-makers at mid-size companies, and the built-in panel can fill only part of the quota. The platform blends in partner sample to reach the rest. Before fieldwork, ask which partners will supply the remainder and whether their respondents pass the same entry screening as the panel's own. After fieldwork, your data comes from two sources, so check that the study record covers both and shows how many responses each source lost to cleanout. If one source lost far more than the other, you know where to look before you present the results.
How does aytm verify its respondent panel?
aytm verifies its panel in four layers, in order: Prevent, Protect, Purify and Prove. As our piece on why data quality works as a discipline explains, each layer backs up the others, and together they carry the load.
Prevent: how does the Respondent Experience Score work?
The Respondent Experience Score reviews the survey itself before launch. It gives the survey's author a score on a 0 to 500 scale with prioritized fixes, and it re-scores in real time as the survey is edited. The score warns and never blocks, so the author decides what to change. This layer comes first because a survey that tires honest respondents produces careless answers that later checks then have to catch.
Protect: how does aytm check traffic and identity?
Traffic and identity are separate questions, so this layer runs two checks. Traffic Sentinel filters suspicious traffic at the source, catching duplicate attempts, location mismatches, and IP and VPN violations before anyone reaches your study. PaidViewpoint verifies each panelist's identity when they register. Our piece on how we use AI to protect your research explains that Traffic Sentinel checks traffic in real time and blocks known bad actors before they reach a survey.
Purify: how does aytm screen the answers themselves?
A verified respondent can still answer carelessly, which is where Data Centrifuge comes in. Data Centrifuge screens each response for speeding, straight-lining and gibberish in open ends, and routes each signal to automatic rejection or to review by a person. It draws on the Qualcha Library's 11 attention-check types. An attention check is a question built to show whether a respondent is reading closely. One missed check is a single signal, so a pattern across several says more about a respondent than any one result.
Prove: what record does aytm give you for each study?
The first three layers help your stakeholders only once you can show them, and that's the job of the per-study audit report. It shows which checks ran on a study and which responses were removed or sent for review. Read it beside aytm's published rates: the report covers your study, and the rates cover the whole panel.
aytm holds ISO 20252 certification, the standard for market, opinion and social research, and ISO 42001 certification, the standard for AI management systems.
FAQ
Which market research platforms come with their own respondent panel built in?
Market research platforms with a built-in respondent panel come in two broad kinds: platforms that recruit and verify a panel of their own, and platforms that pair their own panel with partner sample. To spot either on your shortlist, ask each vendor who recruits the people in your study, who verifies them, and what they hand back after fieldwork. aytm is one example, fielding studies with PaidViewpoint, its own panel, alongside partner panels, and verifying respondents in four layers: Prevent, Protect, Purify and Prove.
What's the difference between a built-in panel and a proprietary panel?
A proprietary panel is a respondent panel a company owns and recruits directly. A built-in panel is a proprietary panel that comes inside the market research software you field with, so the platform and the panel sit with one company. In practice the two terms often name the same thing. On either label, the useful question is who owns respondent verification
Who runs screening when the panel is built in?
Screening on a built-in panel belongs to the company that owns the panel, from entry checks through the study record. Your part shifts to reviewing how that company screens and reading what it reports. Ask what triggers a removal, and read every rate against its base and period before you sign off on the data.
What does panel quality actually measure?
Panel quality measures how well a panel's respondents hold up under checks on identity, attention and the answers themselves. Removal, cleanout, incorrect demographics and duplicate rates each show part of that picture, and each needs its base and period beside it. A rate quoted alone tells you little about panel quality.
Can you blend a built-in panel with outside sample in one study?
Blending a built-in panel with outside sample is common when a study needs more reach than one panel offers. The respondents in your final data then come from more than one source, so ask for each source's record and compare their rates on the same terms. aytm's Sample Engine draws on PaidViewpoint and partner panels, and aytm quality-tests each partner panel before adding it.
Is a built-in panel faster to field?
Field speed depends mostly on your audience's size and incidence, on either path. A built-in panel can save the setup that comes with lining up separate providers and reconciling their screening standards, and a hard-to-reach audience can still take longer to fill on any panel. Ask each vendor for a feasibility and timing estimate on your actual audience before you compare.
Does a built-in panel work for B2B audiences?
A built-in panel works for B2B sample when it profiles members by role and industry or runs specialty panels for professional audiences. Ask how a B2B respondent's job title and role are verified, because a professional audience is harder to confirm than a consumer one. aytm offers specialty panels for audiences such as senior management, IT decision-makers and healthcare professionals.
Before your next study
Line up every platform and sample provider on your shortlist against the same five questions, and set their answers next to each other. To compare one online survey panel against another, put their published rates on the same base and period first, the method our guide to comparing survey panels on data quality and fraud protection walks through rate by rate. If a built-in panel fits your study's audience, you can see where aytm's four layers sit in the platform.
See the platform in action
.webp)
.jpeg)
