For the teams that walk Indian villages and wards with a phone in hand.
Bharat Survey is made in India survey software that grew out of a field-operations team’s frustration: volunteers filled forms from home, reports showed everything ‘complete’, and decisions rested on bad data. The tools on the market were either voter-list utilities or foreign research software that knew nothing of India’s panchayats, wards and booths. So we built the tool we needed ourselves.
The gap in Indian election and survey tools that Bharat Survey was built to close
We used to run voter surveys and field operations for different leaders. Every campaign raised the same questions: Did this survey really happen in that village? Was this volunteer actually in the field today? Is the percentage in the report calculated on the right denominator? Answering them meant paper sheets, WhatsApp photos, Excel copies and phone calls — and at the end, a report nobody fully trusted.
Looking for an answer, we examined five platforms and a dozen Play Store apps. The conclusion was clear: none combined three capabilities — business control (plans / seats / billing), field operations (assignments / live tracking / GPS checks) and survey quality (validation / review / back-checks).
Most election management software in India is built around the voter list — search by EPIC number, caste and age reports, voter slips, booth-wise turnout counters. Useful, but there is no real survey logic: no skip rules, no validation, no version control, no back-checks. Where GPS exists, it is just a ‘tag’ — nothing tells you whether a submission came from inside the assigned panchayat or from 14 km away. And the most common user complaint is “check your internet”, meaning offline mode doesn’t really work.
Tools such as SurveyCTO, KoboToolbox and ODK are strong on form logic and data quality — we learn from them and import their XLSForms directly. But they have no state → district → block → gram panchayat → village hierarchy, no urban body → ward, no Lok Sabha → assembly → booth; every project means building areas by hand. The interface is English-first, plans are in USD, and data may sit outside India. For an assembly candidate or a district NGO, that distance is too great.
We filled the space in between. A Hindi-first app that runs a full day without network on phones with 2–3 GB of RAM. All-India administrative geography — 36 states and UTs, 6.6 lakh+ villages, 4,000+ assembly constituencies — preloaded with LGD codes. GPS verification that sorts every survey into five location states and never counts ‘outside area’ as accepted. Seat-based plans in INR, with GST invoices. And data in India, with consent and retention aligned to the DPDP Act 2023.
The first detailed data load is Uttar Pradesh; the first pilot runs in one assembly constituency with about 10 volunteers. From there, the rest of India — and beyond elections, to corporate market research, NGO baseline and endline studies, and government door-to-door surveys. We are in the pilot phase; this website shows no client logos or results because we don’t have any yet.
The whole product exists to protect this one moment. The surveyor introduces themselves, explains that taking part is voluntary, and asks the questions. The form is saved on the phone, proof of place and time is attached — and the report counts only what actually happened on the spot.
Keep scrolling and the story moves forward. Where a date is firm, we show it; where it isn’t, only the order. The last card is still unfinished — and we are showing it that way.
Running voter surveys and field operations for different campaigns, the same three questions kept coming back: did this survey really happen in that village? Was this volunteer actually in the field today? Is the percentage calculated on the right denominator? The answers were scattered across paper sheets, WhatsApp photos and copies of Excel files.
Indian election software, global research platforms and Play Store survey apps — we logged features, pricing structures and user complaints side by side. The most common complaints: sync nobody trusts (“check your internet”), failing OTPs, no skip logic or validation, and opaque monthly pricing for a three-month campaign.
The finding fit in one line: no single tool combined business control (plans, seats, billing), field operations (assignments, live tracking, GPS checks) and survey quality (validation, review, back-checks). The brief was written for exactly that gap — with acceptance tests, so every claim can be checked.
Four kinds of client — political, corporate and research, NGO, government — on one platform; booth tools and the war room for political clients only. Hindi-first, with English alongside. Maps built on OpenStreetMap data. Pilot builds shipped as a direct APK so Play Store review never holds up fieldwork. Personal voter-roll import kept out of v1.
Fourteen pages at bharatsurvey.org: solutions, features, coverage, the pricing structure, a privacy policy and a demo form. One rule from the first day: no client logos, no made-up numbers, no list prices until plan tiers are final.
The home page and inner pages were rebuilt — real maps, animated sample dashboards and far deeper content: how-tos, checklists, a glossary and ‘when this is not right for you’. The aim: you should know what the product does, and what it doesn’t, before the demo.
The order from here: server and company panel (clients, plans, seats, geography data, form builder, QA), then the client panel (team, assignments, live map, reports), then the Android app, and then the first pilot in one assembly constituency with about 10 volunteers. Each stage passes its acceptance tests before the next begins.
Product principles we do not trade away
These aren’t slogans; they are design rules. When a feature is debated, these five questions settle it. Keep scrolling — each rule opens, shows its example and makes room for the next.
Every record shows the state of its evidence — inside area, near boundary, outside, weak signal, suspected mock GPS. Where boundaries are approximate, the badge says ‘approximate’. Never a false ‘Verified’ badge.
‘Saved on phone’ and ‘reached server’ are shown as two separate counts. If the network drops, the dashboard says ‘last updated 18 minutes ago’. Old data is never dressed up as live.
Big buttons, fewer screens, one question per screen, Hindi, offline. The app is built for the field worker, not the admin — in harsh sun, on a cheap phone, on a weak network.
Tenant isolation, the introduction-and-consent screen, the contact vault, the retention policy — these aren’t fine print but parts of the product that clients can see and check in their panel.
Real screens, real flows, testable acceptance tests. No fake client logos, no made-up numbers. Sample data is always labelled ‘Sample’.
Commitments to every client, every respondent and every volunteer
The working day we design for — illustrative photos, real routine
Every screen starts with one question: how will this feel in the hands of a supervisor in the village square and a volunteer walking the lanes? Hover over a card or use the arrows. The photos are illustrative — not of any real client or respondent.
Bharat Survey works on equal terms for every party, candidate, company, NGO and government department. Survey responses belong to the client; we act as the processor. We do not profile individual voters — not in v1, and not in this product’s principles.
Read the privacy policyThe boundaries and working methods behind every Bharat Survey project
A product is defined by its ‘no’ as much as its ‘yes’. These four things will often be asked of us, and our answer is already settled:
Survey data holds private things: a household’s circumstances, opinions, sometimes phone numbers. So the rules are few and strict. Isolation: each client’s data is kept apart — at the API, database, file, export and live-connection level. Minimal collection: contact details only when the respondent gives separate consent; they sit in a separate vault, hidden from people who view reports. Accountable access: even our own team opens sensitive data only with a recorded reason, and it shows in the audit log.
Retention: data is kept for as long as the account or contract lasts and as needed for legal, audit and export purposes. When a plan ends, data stays read-only; it is deleted or anonymised when the client instructs or on a valid DPDP request. Backups: daily backups and weekly snapshots, with restores tested — because a backup that has never been restored isn’t a backup, it’s a hope. The full detail is in our privacy policy, which will be legally reviewed before launch.
Software only works if a 45-year-old party worker who has never done a survey on an app understands it by the second house. That’s why training takes 45 minutes, in Hindi, screen by screen — start shift, arrival check, consent, form, visit outcome, sync. Every volunteer gets a one-page Hindi guide and a short video; supervisors get a separate session on review and back-checks.
On the first day of fieldwork we are live with you: the first sync, the first ‘outside area’ alert, the first evening report, reviewed together. A record sent back for correction always comes back with a reason — so the volunteer knows what went wrong and doesn’t repeat it the next day. And at the end of the pilot, a written report: what worked, what didn’t, and what needs to change next — in our product too.
We tell volunteers one thing plainly: the app isn’t there to doubt you; it’s there to keep proof of your work. If you covered 38 houses in the sun, all 38 should count — and forms filled from home shouldn’t. For an honest volunteer, that is the biggest benefit of all.
How we work with every client: field first, honest numbers, Hindi throughout
Field first
“Did this survey really happen in that village?” — that question produced the five location states. “Was it saved on the phone or did it reach the server?” — that one produced two separate counts. We write the field question first, then its acceptance test, and only then the screen.
Honest numbers
Of 240 complete records, 210 accepted; 180 gave a valid answer to the water question; 90 reported a problem → 50%. Every report states its time period, form version and response count. A survey of whoever was convenient to reach is never called a ‘representative sample’ of the whole constituency.
Hindi throughout
A 20-minute demo using your own assembly constituency, district or ward. Setup in 1–2 working days. Volunteer training in 45 minutes, in Hindi. We are live with you on the first day of fieldwork. After that, email and call-backs in Hindi or English — whichever suits you.
A small team from the field, building for the field — and the roles we are hiring
We come from both field operations and software — we have run volunteer teams ourselves, counted paper sheets and watched Excel reports built on the wrong denominator. Names and photos will go up when the team is complete; until then, here are the roles we are looking for.
When Bharat Survey is not the right partner — yet, or at all
No one can be right for every organisation all the time. In the four situations below, we will tell you ourselves to wait, or to look elsewhere.
That is a fair requirement, and right now we can’t meet it — the product is in its pilot phase. If your procurement process demands three references, come back in a few months; or join the pilot and see the results for yourself.
We don’t send field teams and we don’t take on questionnaire analysis contracts. The field team is yours; we give it the tools, the training and support on day one. If you don’t have a team, choose a survey agency first — it can work on our platform.
Search by EPIC number, voter slips, voter-by-voter tagging, bulk WhatsApp messaging — we don’t build these. Not in v1, and voter profiling is ruled out on principle. Other Indian tools on the market cover that need.
Bharat Survey keeps ‘outside area’ and too-fast interviews out of the accepted count, and always calculates percentages on valid answers. In the first week your count will often go down. If the goal is filling a report rather than learning the truth, we are the wrong partner.
What Bharat Survey is, how it differs, and what we don’t do — straight answers.
An Indian SaaS platform — a web panel plus an Android app — that lets organisations run door-to-door and field surveys through their own volunteers or enumerators. Every submission is checked against GPS, and results land on a live dashboard. It is built for election surveys, booth management, market research, NGO baseline and endline studies, and government surveys. Seat-based plans, Hindi-first, data kept in India.
Yes. The product is built in India, for Indian field teams, and the servers are hosted in India. Each client’s data is isolated, encrypted in transit and at rest, audit-logged and backed up daily, with consent and retention rules aligned to the DPDP Act 2023. The company’s legal name and registration details will be added to this page and the footer before launch.
No. We are a technology platform and work on equal terms for every party, candidate, company, NGO and government department. No client’s data reaches another client at any level — API, database, file, export or live connection. Questions are worded neutrally, and sensitive questions go live only after our review.
Those are mainly voter-list and booth tools: voter search, slips, turnout counters. We focus on survey quality and verification — real form logic (skip rules, validation, XLSForm), GPS checks that sort every record into five location states instead of just tagging it, an offline-first app, transparent seat-based plans, and uses beyond elections in research, NGO and government work. We do not import personal voter-roll data in v1.
They are strong research tools, but they don’t know India’s administrative map — panchayat, ward, assembly constituency, booth. Their interfaces are English-first, plans are priced in USD and data may sit outside India. Bharat Survey ships with LGD-coded geography for all 36 states and UTs, a Hindi-first app, seat-based INR plans and data in India. Your existing XLSForms import directly.
Not yet, and we will not invent them. The product is in its pilot phase; the first pilot will run in one assembly constituency with about 10 volunteers. Every dashboard and figure on this website carries a ‘Sample’ label. When a real pilot report exists, it will appear here only with the client’s permission.
Yes. The app is native Kotlin, designed for phones with 2–3 GB of RAM, with a Hindi-first interface (English is one tap away), large buttons, minimal typing and a full day of work without network. Photos are compressed on the phone to save data. More Indian languages will follow in later versions.
A small team that comes from both field operations and software. We will publish names and photos once the team is complete; the roles we are hiring for are listed on this page. Write to hello@bharatsurvey.org with the role in the subject line.
The website has been live since September 2026. The platform is being built in stages — server and company panel first, then the client panel, then the Android app, then a first pilot in one assembly constituency with about 10 volunteers. Each stage must pass its acceptance tests before the next begins, so we do not publish dates. Book a demo and we will tell you exactly where things stand and whether a pilot fits.
We are a software platform. The field team is yours — your volunteers, enumerators or agency staff. We handle setup, a 45-minute training in Hindi, live support on the first day of fieldwork and help after that. The questionnaire is yours too, though you can start from our form library.
You do. Survey responses belong to the client; in DPDP terms you are the data fiduciary and we are the processor. We never show your data to another client, never sell it to a third party and never build our own analytics product on it. After a plan ends you can still export in read-only mode; data is deleted or anonymised when you instruct us or on a valid DPDP request.
Yes. The state and agency setup includes a multi-client workspace: the agency keeps each client’s project separate, and each client sees only its own dashboard. For partnerships, write to hello@bharatsurvey.org with ‘Partnership’ in the subject line.
The photos are illustrative — none show a real client, volunteer or respondent. Dashboards, moving map points and figures use sample data and are labelled ‘Sample’. The geography figures (36 states and UTs, 6.6 lakh+ villages, 4,000+ assembly constituencies) are facts. A real pilot report will appear here only with the client’s permission.
Demo, partnership, press or careers — we reply within 1 working day. Book a 20-minute demo using your own assembly constituency, district or project.
