Duplicate and bot screening
Device and response-pattern checks catch the same person answering twice or scripted completes.
- Duplicate device flags
- Open-end gibberish checks

Partner panel sourcing, quality checks and field top-ups for people online panels miss
If you are searching for an online panel provider in India, here is the honest position first: Bharat Survey does not own a consumer panel. We source online sample from established third-party panels, script and quality-check the survey, and add face-to-face interviews with our own field teams for audiences that panels under-represent — rural households, older adults, low-income and Hindi-only respondents. You get one dataset, one set of quality rules and a clear note on which interviews came from where.
An online panel survey in India reaches people who use the internet regularly, are willing to join panels and complete surveys on a phone or laptop. That skews urban, younger, more educated and more English-comfortable than the population. Weighting can correct some imbalance on age or city tier, but it cannot create respondents a panel doesn't have — such as older rural women or first-time smartphone users.
For many studies that is fine. Concept tests for a D2C brand aimed at metro buyers can run entirely online. For a mass FMCG launch, a financial inclusion study or anything described as all-India, the online sample needs a field component, and the report should say so.
We set quotas for the full target population, fill what online sources can reach well, then send field teams with the same questionnaire in the app to fill the remaining cells. Every record carries a mode flag so analysts can check whether answers differ by mode before combining.
Bad online data is common and cheap. These checks run on every source, including our partners'.
Device and response-pattern checks catch the same person answering twice or scripted completes.
Interviews far faster than the median, or failing attention items built into the questionnaire, are removed before counting toward quota.
Contradictory answers — a non-car owner rating their car's service — are flagged and reviewed.
Removed interviews are replaced, and the removal count by reason is reported in the methodology note.
Consumer, brand and retail research.
Learn moreUsage and attitude, concept and product tests.
Learn moreFace-to-face research where panels don't reach.
Learn moreRepeated waves with consistent samples.
Learn morePhone-based short surveys.
Learn moreStraight answers to the questions people ask most about this topic.
No. We don't own or operate a consumer panel. For online sample we work with third-party panel partners, named in the proposal, and apply our own quality checks. Our own capacity is field interviewing with trained teams and our survey app.
Not on its own. Internet users who join panels differ from the overall population in location, age, education and language. A mixed design that adds field interviews, with weighting and transparent disclosure, gets closer, but the report should still describe the sample accurately.
We script in English and Hindi as standard and can add other Indian languages with translation and back-translation. Availability of respondents in each language depends on the partner panel, which we check before confirming quotas.
Pricing is on quotation. It depends on incidence of the target group, interview length, number of completes, languages, and how many field interviews are needed to cover offline groups.

Tell us the target group, sample size and deadline. We will tell you honestly what online sources can cover and what needs fieldwork.