Consumer Panels in India: How They Work and How to Judge One

AAPOR's task force on online panels, reporting in 2010, describes three sources of online sample (AAPOR Report on Online Panels, Public Opinion Quarterly, 2010):

The report's first conclusion is blunt: "Researchers should avoid nonprobability online panels when one of the research objectives is to accurately estimate population values." It adds that an opt-in panel can still be the right choice, because not all research aims at precise population figures.

So ask which kind you are buying, and match it to the claim you will make. In our view, matched cells from an opt-in panel can tell you whether pack A beats pack B. They are a weak basis for a sentence such as "X% of Indian households use Y". Our sampling methods guide covers the probability side in more depth.

ESOMAR publishes 37 Questions to Help Buyers of Online Samples (ESOMAR). Question 5 asks which sources are proprietary or exclusive, and defines both. Proprietary means "the sample provider owns the asset". Exclusive means the provider has "an exclusive agreement to manage/provide access to sample originally collected by another entity" (wording as reproduced in EMI Research Solutions' published answers). Use those definitions in your contract.

ESOMAR's notes to question 4 say a provider may deliver sample from one source, such as its own panel, or blend participants from several, and that the answer "might differ from country to country and from project to project". Some firms are aggregators by design: EMI says it has "never owned a panel" and works through a network of over 150 vetted online panels.

Neither model is wrong. They fail in different ways. Our summary:

For a tracker, the last two rows matter most. ESOMAR's question 26 asks how a provider manages "source consistency and blend at the project level". Question 16 asks whether a router or other technique allocates participants to surveys, and warns that biases "may arise from prioritization in the order in which surveys are presented".

Panel members answer many surveys, and repetition can change their answers. AAPOR's report cites a comparison of 20 US online panels in which, on average, 33% of respondents said they had taken 10 or more online surveys in the previous 30 days.

It also cites studies finding that experienced panellists were generally less positive than new members on measures such as likelihood to purchase. One of them found purchase intention and brand awareness less affected when the gap between surveys was long enough. These studies predate 2010, and other studies found few differences, so ask your provider for its own Indian figures rather than assuming either way.

ESOMAR's question 23 asks how often one person can take part, and how long they have been taking surveys. Question 37 lists metrics a buyer can request, including the share of paid completes from members with 0-3 months' tenure and the average number of paid completes per member, both trended by month.

ESOMAR's question 7 asks what validation a provider uses "to ensure that participants are real, unique, and are who they say they are". It asks for both the practical steps and the technologies, quantified where possible.

Push for specifics. What is checked, and against what? Is it done once at sign-up, or again for each survey? What happens to an account that fails? An answer that names no check is itself an answer.

Sign-up checks and in-survey checks are separate layers, and you need both. Speeding, straight-lining and fraud screens are covered in our survey data quality guide.

Send these eight questions before you buy, and ask for written answers. The good answers and red flags are our own judgement; the last column points to the matching ESOMAR question.

A provider that answers all eight in writing is easier to hold to account than one that sends a slide deck.

Size claims. Panel sizes are self-reported and measured in different ways. Research Live reported in April 2023, for instance, that Kantar aimed to "quadruple current panel capacity in India" within two years, to 40 million people (Research Live). That was a target, and no headline figure tells you how many people will answer your study. ESOMAR's question 37 shows the better form: active unique members fitting a target in the last 30 days. Ask for that count for your own target.

Towns and tiers. "Tier 2" means different town lists to different providers. Ask for the list your quotas will use; our Tier 2 and 3 research guide explains why it matters.

Language. Ask which Indian languages a questionnaire can run in, and whether members see it in a language they chose.

Offline households. An app or web panel cannot reach people who are not online. For village or low-connectivity audiences, plan face-to-face fieldwork; our rural research guide covers the options.

Verification is where our panel methodology page says most, so here is how it answers row 3 of the scorecard. Members join through the SuperJ app and complete Aadhaar-based KYC once, which the page says "prevents one person from holding several accounts". Each study then receives a zero-knowledge proof of verified status, so clients see answers and never identity documents. The 25M+ consumer footprint is the reach of the network; fewer people are eligible for surveys, and how many fit your study depends on its targeting.

The page also answers row 8: the panel is app-based, so offline households are under-represented. It does not cover rows 2, 4, 5 or 7, so ask for recruitment channels, routing, participation limits and monthly metrics in writing.

The cheapest test of any panel is a calibration survey with a question whose answer you already know from your own sales or CRM data. A 100-response check on Plus costs 100 × ₹150 = ₹15,000 (pricing). If you first want to see how the questionnaire looks on a phone, the Free plan includes 30 responses.