GPS-verified, offline-first field app for election surveys.
Volunteers fill surveys without internet, every submission arrives with proof of place, time and signal, and results show up on your dashboard the moment they sync. Election surveys, voter surveys, booth management, market research, NGO baselines and government household surveys, on one platform, backed by field teams when you need them.
The surveyor is standing in a village lane with no network. The form is saved on the phone, location proof is attached, and the moment a signal returns the record clears its checks and lands on the dashboard. Nobody has to make a single phone call.
Most field surveys today are scattered across five places, so nobody knows which data is right. Bharat Survey keeps forms, team, location proof, QA and reports in one place.
Four rules that stop a candidate's money going to waste, keep a research and survey agency's data dependable, and keep the audit of a government survey clean.
Every submission records location, time, accuracy, the age of the fix and mock-GPS signals. The server matches it against the boundary of the assigned panchayat or ward and sorts it into one of five states. Only “inside area” records join the approved count; “outside area” goes to review and the leader or manager gets an alert.
Forms, areas, household lists and instructions download in the morning on the home network. In the field every section is stored in the phone's encrypted database. It syncs on its own, with retries, as soon as a network appears. The app always shows two counts: “saved on phone” and “reached server”.
During a shift, each volunteer's last location, time, accuracy and assigned boundary appear on a live map. The activity timeline shows shift start, arrival check, survey saved or submitted, pause and end. Alerts cover outside area, inactive for too long, too-fast interviews and pending sync.
Too-fast interviews, contradictory answers, impossible time/distance sequences and duplicates go to a review queue automatically, each with a risk score. The supervisor can send a correction back or run an independent back-check. Approved data becomes panchayat, village or ward-wise reports in CSV, Excel or PDF, plus a Google Sheets copy.
This is the sequence that runs in the field. At every stage you know how far the work has got, and every record carries its proof.
Build questions in the form builder or import an XLSForm, with skip logic, validation and Hindi/English labels. District → block → panchayat → village, urban body → ward, or assembly → booth: all already there with LGD codes.
Client panel · web · day 0–1Link each volunteer by name and mobile number, activate with an OTP and register the device. Assignment = volunteer + area + form + date range + target (for example 300 homes). If every seat is taken, ‘Request extra seat’ is one button away.
Client panel · web · day 1Shift start, arrival check, introduction and consent, section-by-section form autosave, and the visit outcome (complete / partial / refused / house locked / revisit). A full day without network; sync as soon as a signal returns.
Android app · Hindi · dailyThe server checks every record's location state, speed, duplicates and contradictions; suspect records go to review with a risk score. Approved data flows to the dashboard, live map, alerts, panchayat-wise reports and Sheets automatically.
Server · QA · dashboard · liveKeep scrolling and the day moves forward. The same sequence runs on every project, whether it is an assembly voter survey, an FMCG retail audit or a district-wide household survey.
Open the app on home WiFi or mobile data and today's assignment, the latest form version, the approximate area boundary and the household list land on the phone. The screen says it plainly: ‘Download complete, you can now work without a network’.
On their own phone the supervisor sees which volunteer has which panchayat or ward, how many surveys came back yesterday for correction, and today's target. If a volunteer does not turn up, a replacement takes the seat in two minutes.
Pressing ‘Start shift’ begins location sharing, with clear notice and only for the shift. On reaching the area the app records an arrival check: place, time, accuracy. That is the first proof the team is really on the spot.
The surveyor explains who is running the survey and why, and that taking part is voluntary. The time and version of consent are recorded. Then one question per screen, big buttons, skip logic, and every section autosaves to the phone's encrypted database.
Without a single phone call: where each volunteer is, whose shift is running, how many surveys reached the server and how many are still on phones. If the network has dropped, the dashboard says so honestly: ‘last update 18 minutes ago’.
Records sync as soon as there is a network, each with one ID, so nothing is counted twice. The server checks every record's location state, speed, duplicates and contradictions. Clean records are approved automatically; suspect ones go to the supervisor's queue with a risk score.
Pressing ‘End shift’ stops location sharing. The app shows: 38 surveys today, 38 on the server, 0 left on the phone. Records returned for correction appear on tomorrow's list with the reason.
Panchayat and ward-wise progress, data quality, top issues and help requests, with the time period, form version and response count. Download a PDF or Excel file, or open the copy that has already arrived in Google Sheets.
From area assignment to the booth board: five modules, one dashboard. Switch tabs below to explore.
Last location, time, accuracy, shift status and the approximate boundary of the assigned area, on one map. With SSE realtime the dashboard updates in ~5 seconds on a normal network. If the network drops, the message stays honest: “last update 18 minutes ago”.
Keep scrolling: each layer opens, shows its job and makes room for the next. The same sequence runs for every record, from the phone to the report.
A Hindi-first, offline-first app that runs on phones with 2–3 GB RAM. Forms, areas and instructions are downloaded in advance; every section autosaves to an encrypted local database; location is shared only during a shift.
The server matches every record against the approximate boundary of the assigned area and sorts it into five location states. It also checks interview speed, duplicates, contradictory answers and impossible time/distance sequences.
Team and seats, assignments, live map, activity timeline, alerts, QA queue and four kinds of report: progress, quality, local issues and help requests. Fully usable in a mobile browser too.
An extra layer for political clients: booth master, booth in-charges and page in-charges, A/B/C/D ratings with reasons, daily booth reports, and an hour-by-hour turnout board on polling day.
Single/multiple choice, number, text, date, rating, photo, signature, barcode. Skip logic, validation (errors in Hindi), calculations, piping, repeat groups, cascading selects, multilingual labels. XLSForms from Kobo, ODK or SurveyCTO import directly.
A published form is locked; a change creates a new version; old answers keep the meaning of the old question, so wave-by-wave comparisons stay safe. Sensitive questions go live only after company approval.
At every home the surveyor says who is running the survey, what it is for, and that it is voluntary. The time and version of consent are recorded. Every visit gets an outcome: complete / partial / refused / no contact / house locked / revisit.
Last location, time, accuracy and shift status. Timeline: Shift Started, Arrival Check, Survey Started/Saved/Submitted, Paused, Resumed, Shift Ended. Honest timestamps; an old location is never shown as live.
Surveys outside the area, inactive for too long, too-fast interviews, pending sync, an offline permit about to expire: shown as a list, in order of priority.
A. Work progress B. Data quality C. Local issues D. Help requests. Filters: area/date/volunteer/version. CSV/Excel/PDF; a one-way Google Sheets copy (Overview, Area, Quality, Team, Issues).
Part number, building, voter count, coordinates; mandal/shakti kendra; booth in-charge, BLA, members; page in-charges with page ranges. Import from CSV or the CEO's PDF list.
A/B/C/D ratings with reasons and history; a 2-minute daily booth report; a compliance board showing which booths have not reported, and for how many days in a row.
Hour-by-hour turnout entry, a count of ‘supporters yet to vote’, lagging booths and missing updates, live on the war-room board; offline entries stay safe on the phone.
The volunteer app, the supervisor's QA queue, the live map, the booth board and the reports all run on the same data. Scroll down and watch each screen settle into place.
Four kinds of clients, one data engine you can trust. Hover over a panel: each one gets its own modules, language and reports.
Screens you can read in harsh sun, buttons you can press with one thumb, weak networks and low-cost phones. Every decision came from how fieldwork really looks. Hover over a card or use the arrows to turn the deck.
Role and assigned area are both checked on every request. A supervisor cannot see another team's data, a viewer cannot change anything, and client A's data has no route to client B.
Everything in your project, even from a phone browser. Seat usage (“9 of 10 used, 1 free”), a one-tap request for extra seats, making supervisors, assigning teams and areas, and downloading reports.
They see only their own team and area. Suspect records sit in the QA queue ordered by risk score; approve them, return them for correction with a reason, or assign an independent back-check. Granting an exception never erases the original proof.
A Hindi-first Android app with five or six screens and big buttons. 1 seat = 1 named person = 1 authorised device. Today's work, start/end shift, arrival check, consent, form, visit outcome and sync. That is all.
People who only look at results need no seat. Read-only dashboard, map and reports; no editing, no contact details. An agency can show each client only that client's project.
Creating clients, setting plans and seat limits, approving extra-seat requests, importing area data, reviewing forms with sensitive questions, and live support on day one. We handle all of it so your team can focus on the field.
Built for the volunteer, not the admin. Big buttons, more tapping than typing, a font-size setting and one question per screen. Native (Kotlin) performance on phones with 2–3 GB RAM, with no heavy webview. Photos shrink on the phone itself (5 MB → 200 KB) so data costs stay low.
Every section goes to an encrypted local DB; data comes back after the app closes or restarts; 24-hour Offline Work Permit.
Foreground and user-initiated, in line with Play Store policy. No background tracking.
Read-only inside the same app for leaders and managers: Overview → Map → Activity → Alerts.
No leader, programme manager or officer has to add areas by hand. State, district, block, gram panchayat, village, urban local body, ward, Lok Sabha, assembly constituency and polling station, with official LGD codes and names in Hindi and English. Lok Sabha, assembly, municipal or panchayat: any election, in any state.
Project, area data, forms, seats, assignments and dashboard, because the areas are already loaded.
In Hindi, with live support on day one and separate review and back-check training for supervisors.
From submission to dashboard on a normal network (SSE realtime), measured during the pilot.
Winners, margins, turnout and maps for every assembly and Lok Sabha constituency, built from Election Commission of India reports. Search a seat, or start from a state.
Open reference pages built from ECI, Census 2011, LGD and MCA records, with the source credited on every page.
The field team is the biggest cost in any campaign or study. If 10–20% of surveys are filled from home, the daily allowance is only part of the loss: every decision built on that data (where to campaign, which issue, which village) points the wrong way. Enter your own numbers below.
This estimate uses only your inputs; it is not an industry claim. Bharat Survey keeps “outside area” records and too-fast interviews out of the approved count.
Survey data holds people's private information. That is why security is part of the product for us, not the fine print.
Data on our servers, reports in your format. XLSForms from Kobo, ODK or SurveyCTO import directly, so there is no need to rebuild your forms.
The Google Sheets copy is one-way: if Sheets is down or hits a quota limit, the app and dashboard keep running. An API comes in a later version.
Behind every survey sit signals of place, time, speed, device and consent. Bharat Survey adds them up to show the one count you can base a decision on.
Every method has its place. Below are the things that matter most in large field surveys in India, whether that is an election survey, a voter survey, market research, an NGO baseline or a government survey.
| Capability | Paper + Excel | Google Forms / WhatsApp | Foreign research tools | Bharat Survey |
|---|---|---|---|---|
| Works a full day without internet | ||||
| Location proof and 5 states on every submission | Geo-tag only | |||
| Surveys outside the assigned area left out of the count | ||||
| Automatic checks for too-fast interviews, duplicates, contradictions | Partial | |||
| India's geography (LGD, assembly, ward, booth) pre-loaded | ||||
| Hindi-first app on low-cost Android phones | — | Partial | ||
| Skip logic, validation, form versions, XLSForm import | Basic | |||
| Live map and alerts (during shifts only) | Partial | |||
| Booth management and war room | ||||
| Seat-based team, replacements at no charge | — | — | Per user / per submission | |
| Data in India, DPDP-aligned consent and audit log | Not by default |
This compares general categories, not any single product; capabilities vary across individual “foreign research tools”. Details: all features.
A field survey app is the mobile application a surveyor uses to go door to door, shop to shop or office to office, ask questions and record answers straight on the phone, with no paper forms and no data entry afterwards. It is also called CAPI (computer-assisted personal interviewing). In India the work is different for three reasons: networks in villages and small towns are unreliable, the surveyor is often doing formal work on a smartphone for the first time, and the units of geography (gram panchayat, ward, assembly constituency, polling station) do not exist in foreign tools out of the box.
‘Offline’ means more than a form that opens without a network. The real test is an eight-hour field session: after a hundred surveys, a phone restart and the app being closed, not one confirmed record is lost, and when the network returns the same record can be sent ten times and still be stored on the server once. In Bharat Survey every submission carries a permanent unique ID, and the app always shows two counts: saved on the phone, and reached the server.
Most election survey apps and data collection apps only geo-tag, which means attaching a latitude and longitude to the record. That does not tell you whether the spot was inside the surveyor's assigned area, how old the fix was, how accurate it was, or whether a mock-location app was running on the phone. Bharat Survey combines all of these to place each record in one of five states, and only ‘inside area’ records join the approved count. Where boundaries are approximate, the badge says so: ‘approximately inside area’.
An assembly voter survey, an FMCG retail audit, an NGO baseline and endline, and a district administration's household survey all need the same things: forms, a team, areas, proof, QA and reports. The difference sits in the top layer. Political clients need a booth master, booth in-charges, page in-charges, A/B/C/D ratings and a polling-day turnout board; a research agency needs quotas and multi-client workspaces; an NGO needs permanent household IDs and follow-ups; a government department needs LGD codes, enumerator–supervisor–block–district roles and an audit trail.
The size of a field team is known in advance: 10, 50 or 200 people. So the price rests on the same thing: 1 seat = 1 named person who can fill surveys. However many surveys they submit, the fee does not grow; swapping a volunteer is a replacement, not a new charge; and people who only view reports need no seat. An eleventh person becomes active only when a seat is added, and the server checks this on every activation.
These words come up in almost every demo. Knowing them beforehand makes the conversation easier.
Straight answers to the questions every demo starts with.
Two things. We run field surveys (election and opinion surveys, market research, NGO baselines and government household surveys) with trained field teams, and we provide the software behind them: an offline, GPS-verified Android app, a web dashboard, and booth management for campaigns. You can hire us for the fieldwork, use the platform with your own team, or combine both.
Ask for proof rather than promises. A good partner shows where and when each interview happened, explains how fake or rushed surveys are caught, states the denominator behind every percentage, keeps data in India with a clear consent policy, and tells you plainly what it cannot do. Bharat Survey is built around exactly these checks, so you can verify them in a demo.
The strongest method is evidence at the moment of the interview. With every submission we record location, time, accuracy, the age of the GPS fix and mock-location signals, then match it against the assigned panchayat or ward. Records outside the area stay out of the approved count, and too-fast interviews and duplicates go to a supervisor's review queue automatically.
Most survey companies in India either sell a tool or run fieldwork. We do both on one data engine that is built for Indian conditions: a Hindi-first app that works a full day offline on low-cost phones, India's geography pre-loaded with LGD codes down to village, ward and polling booth, and seat-based pricing instead of per-submission fees.
Yes. Forms, areas and instructions download in advance, every section is saved in the phone's encrypted local database, and data syncs to the server automatically when the network returns. The app always shows how many records are saved on the phone and how many have reached the server. Our acceptance test is an 8-hour session with 100 surveys and no confirmed save lost.
No. Political campaigns get extra modules such as booth management and a war room, but the same platform serves companies running market research and retail audits, NGOs running baselines and endlines, and government household or facility surveys, for Lok Sabha, assembly, municipal or panchayat work in any state.
1 seat = 1 named person who can fill surveys in the field. A 10-seat plan keeps 10 volunteers active; an 11th becomes active only after a seat is added, with company approval and payment. Replacing a volunteer is a swap, not a charge, and viewers who only read reports need no seat. Pricing is shared on quotation.
The app is Hindi-first with an English option, uses big buttons and little typing, and is built in native Kotlin to run on Android phones with 2–3 GB RAM. Photos are compressed on the phone so mobile data costs stay low. There is no iOS app at present.
All data is hosted on servers in India. Each client's data is fully isolated, encrypted in transit and at rest, and every important action is written to an audit log. Backups run daily, and consent and retention follow the DPDP Act 2023. Contact details are stored in a separate vault from survey answers.
Yes. XLSForm import supports question types, skip logic, constraints, choice lists and multilingual labels. Anything that is not supported appears line by line in the import report, so you never have to rebuild an old form from scratch or guess what changed.
Because area data is already loaded, a project for one assembly seat or district usually goes live in 1–2 working days, including forms, team seats, assignments and the dashboard. We run the 45-minute volunteer training in Hindi and provide live support on the first day.
Not in v1. Importing personal data from electoral rolls carries risk under the DPDP Act and ECI rules. We work with booth-level masters (part number, building, voter count) and your team, not with profiling individual voters, and voters' locations are never tracked.
Location is collected only during a shift, after the volunteer presses ‘Start shift’, with clear notice, and it stops at ‘End shift’. There is no background tracking, in line with Play Store policy. Leaders see a volunteer's location only for the time they were on shift.
In 20 minutes we set up your constituency, ward or project live — team, assignments, offline app and dashboard.
