Field-verified election surveys with a method your newsroom can publish
Bharat Survey works as a poll partner for news channel and digital newsroom teams that need election numbers they can put on screen and still defend a week later. We design the sample, run interviews with our own offline survey app and field teams, verify every record by GPS and timestamp, and hand over vote share estimates with error ranges, a methodology note and the raw data behind each chart.
- Methodology note written for the broadcast disclosure
- Every interview GPS-verified and time-stamped
- Vote share with ranges, never a lone seat number
- Release timing checked against ECI rules
- In areaInside approx. boundary
- Accuracy±12 m
- Time09:42 · fresh fix
- Mock GPSNot detected
- Interview length11 min
- Comes regularly
- Comes sometimes
- Rarely / never
- Don't know
What a newsroom is really buying
As an election poll data provider we send the newsroom three things together: the estimates, the method and the evidence. Graphics teams get clean tables; the editor gets a disclosure note; the legal desk gets the release timeline.
Estimates in newsroom-ready tables
State and region-wise vote share, issue rankings and demographic splits arrive as CSV and chart-ready sheets, with the unweighted base printed beside every percentage so no one reports a split built on 60 interviews.
- Weighted and unweighted bases
- Confidence ranges per estimate
- Round-on-round comparison
Evidence behind each chart
Each interview carries location, time, duration and interviewer ID. If a rival channel or a party questions a figure, your team can show where and when the underlying interviews happened, without exposing any respondent's identity.
- GPS point and accuracy
- Interview duration
- Supervisor back-check status
- In areaInside approx. boundary
- Accuracy±12 m
- Time09:42 · fresh fix
- Mock GPSNot detected
- Interview length11 min
A seat projection survey agency has to get two things right: the vote share estimate and the conversion of that share into seats. We keep them as separate steps so each can be checked on its own.
- 01
Agree what will be published
State vote share, regional splits, seat ranges or issue tracking — each needs a different sample spread. We fix this with the editor before sampling.Day 1–2
- 02
Draw a probability sample
Assembly segments and polling-station areas are selected at random within regions, and households within them by a fixed interval. No voter-list personal data is used.Day 2–4Ram Kumar · shift on26.8467 N · 80.9462 E · ±12 m14Received12Accepted1Review1Outside - 03
Field interviews offline
Enumerators use a Hindi-first Android app that works without network, records GPS and duration, and keeps the vote question on a private screen.FieldworkSurvey 3/6 · AmenitiesOfflineHow is drinking water in your area?- Comes regularly
- Comes sometimes
- Rarely / never
- Don't know
✓ Section saved on phone · autosaveBackNext - 04
Verify and weight
Out-of-area, too-short and duplicate records are held back. Remaining data is weighted to Census and past turnout patterns, and the weights are disclosed.Daily3,000Target outlets2,184Visited1,962Audited1,871AcceptedDaily accepted auditsLast 12 days - 05
Model seats with stated assumptions
Vote share goes into a regional swing model. We report a seat range and the assumptions that would move it, not a single confident figure.Before release3,000Target homes2,184Visits1,962Completed1,871AcceptedDaily accepted surveysLast 12 days
- Comes regularly
- Comes sometimes
- Rarely / never
- Don't know
A vote share projection survey measures intention across a sample of adults and weights it to the electorate. With about 1,000 random interviews, a single share near 50% carries roughly ±3 points at 95% confidence before any clustering. Household surveys are clustered, so the effective sample is smaller, and state-level splits need their own adequate bases.
That is why we recommend reporting ranges and naming the sample size on screen. It protects the channel when the result lands at the edge of the range, and it is simply the correct description of what a survey knows.
Polling agency accuracy India: judge the method, not the claim
Every agency advertises a good record. Polling agency accuracy in India is better judged by what is disclosed: sample size per state, how stations and households were chosen, dates of fieldwork, weighting variables and how refusals were handled. A partner that will not share these before signing will not share them after a miss.
Most accurate exit poll agency India: why we don't use that phrase
Search for the most accurate exit poll agency India has and you will find many claimants and few published error tables. We make no accuracy promise. Exit polls face refusals, clustered samples and seat conversion error, and close multi-cornered contests magnify each. What we promise is a documented method and verifiable fieldwork.
Disclosure note includes
- Sponsor and field agency
- Fieldwork dates
- Achieved sample per state
- Sampling and weighting method
- Margin of error statement
- Question wording for the vote item
| Format | When fieldwork happens | Good for | Legal timing to check |
|---|---|---|---|
| Pre-poll tracker | Weeks before polling, in rounds | Mood, issues, leader ratings | Section 126 silence period in the last 48 hours |
| Exit poll | Polling day, outside stations | Early vote share estimate | Section 126A notified window |
| Post-poll survey | Within days of voting, at home | Why people voted as they did | Release after the 126A window for that election |
| Issue poll | Any time outside silence periods | Explainers and debates | General media norms |
Timing rules summarised from the Representation of the People Act, 1951. Always read the ECI notification for the specific election.
Sample
- How many interviews per state and region?
- Are households chosen at random or by quota?
- What is the design effect assumption?
Field
- Is every interview GPS and time-stamped?
- What share is back-checked, and by whom?
- Are women interviewed by women where needed?
Release
- Who holds data until the legal window closes?
- Will raw data be available for audit?
- Is a written methodology note included?
Not the right fit if
- You need a guaranteed winner call for a headline
- Results are wanted before the ECI exit poll window closes
- The questionnaire must favour a party or candidate
- You need voter-list based targeting of respondents
A good fit if
- You want a disclosed, repeatable method
- Coverage spans several states and languages
- Your legal desk wants records it can audit
- Editors prefer ranges over false precision
Political survey solutions
Constituency research, booth-level surveys and war-room dashboards.
Learn moreOpinion poll survey
Voter mood, issue and leader ratings between elections.
Learn moreExit poll survey
Polling-day fieldwork with station sampling and legal release timing.
Learn morePre poll survey agency
Pre-election rounds that track movement before polling.
Learn moreHindi: राजनीतिक सर्वे
Political survey services explained in Hindi.
Learn moreStraight answers to the questions people ask most about this topic.
Can a news channel commission an opinion poll during elections?
Yes, outside the restricted periods. Section 126 of the Representation of the People Act bars displaying election matter, including opinion poll results, on electronic media in the 48 hours before polls close in a constituency, and Section 126A bars exit poll results during the ECI-notified window. We plan release dates around both.
Do you publish results under your own name or the channel's?
That is agreed per contract. Most channels credit both the broadcaster and the field agency. Either way, the disclosure note names who conducted fieldwork, when, how many people were interviewed and how the sample was weighted, so viewers can judge the numbers.
How fast can a state-level round be completed?
It depends on the number of regions reported separately, languages and interview length. A short questionnaire in one state can usually be fielded within days once sampling is agreed; multi-state rounds take longer. We share a day-by-day plan with the proposal rather than a generic promise.
Do you guarantee accurate seat projections?
No. Seat projections depend on vote share estimates and on how that share translates into seats in close contests, both of which carry error. We give ranges, state the assumptions and provide the underlying data so your team can explain any gap between projection and result.
Tell us the states, dates and what you intend to publish. We will reply with a sample plan, field timeline and disclosure note template.