Geo-tag
Latitude & longitude on every survey
Our method is built to make a finding defensible before it is published. We listen to the same population in more than one way, verify every record at the point it is collected, and document how the work was done—so a conclusion can be examined, not merely trusted.
Face-to-face fieldwork, telephone interviewing and qualitative depth are designed to corroborate one another, rather than leaning on a single instrument to carry the weight.
Every interview is supervised, geo-tagged and open to back-check from the moment it is recorded—verification is part of collection, not an afterthought.
Methods are written down and travel with the work, and findings are reported as the field returns them—independent of who commissioned the study.
Samples are drawn to mirror the population they speak for—stratified across region, locality type and demographic profile, and carried out by trained investigators working in the geographies they know best.
Each study begins with a sampling frame designed to be representative, then stratified so that every relevant segment is reflected in proportion: region, locality type—urban, rural and tribal—and the demographic profile of the people being studied. Where a brief calls for sharper resolution in particular areas, such as aspirational or tribal districts, those areas are given fuller coverage and brought back to their true population share at the analysis stage.
Fieldwork is delivered through a pan-India network of trained investigators, recruited and briefed locally so that interviews happen in the respondent's own language and setting. Respondent selection and field protocols follow standard, documented procedures, which lets the same design run consistently across many states at once without drifting from study to study.
A purpose-built mobile application is the backbone of collection—surveyors record answers directly on the device—with computer-assisted telephone interviewing extending reach to respondents better reached by phone.
Field surveyors work through a custom application designed for the job. Responses are entered straight onto the device, which removes the transcription step where errors usually creep in. An intuitive interface keeps each interview moving, built-in skip logic routes respondents past questions that do not apply to them, and the questionnaire is presented in the respondent's own language. Once an interview is submitted, its record is locked.
Where households or areas are better reached by telephone, the same study is extended through computer-assisted telephone interviewing, run to the same script and the same checks. Together, in-person and telephone modes widen coverage without lowering the bar on how each response is captured and recorded.
Completed interviews travel to a central database as they are recorded, so data is secured and ready for analysis the moment it leaves the field—with no second pass of manual entry.
As each interview is finished, the application transmits it to a central database over any reachable network. Where connectivity is intermittent, records are held on the device and synced as soon as a signal returns, so nothing collected in the field is lost. Because the response moves straight from the device to the database, there is no re-keying and no window in which a record can be quietly changed.
Removing the lag between collection and the central office does two things at once: it protects the integrity of each record, and it lets the team see the study fill in as it happens—ready to monitor, validate and analyse without waiting for paper to come back from the field.
Every interview is stamped with the exact location it was taken, and surveyor movement is tracked—the foundation for a layered set of checks against fake or misleading responses.
Each survey is tagged with the latitude and longitude of where it was recorded, and the movement of field researchers is followed through their device. Together these establish, for every response, where it came from and whether the surveyor was actually there—the first line of defence against an interview that never really happened.
On top of the location record sits a layered verification routine. Short voice snippets and follow-up calls let a separate team confirm that an interview took place as logged; interviews are time-bounded so that an implausibly fast survey stands out; and duplicate mobile numbers are screened to catch the same respondent appearing more than once. Anything that fails a check is flagged and sent back to be collected again.
Before a single response counts, it passes through layers of supervision and back-checks; before the data is read, it is weighted so the sample reflects the population it stands for.
Quality is managed in tiers rather than at a single gate. Field supervisors watch collection as it comes in, a dedicated quality desk reviews incoming records in real time, and a share of interviews is back-checked against the items a respondent's answer can be verified on. Records that do not hold up are returned and re-collected, so the dataset that reaches analysis has already passed through more than one pair of eyes.
Once the field is clean, responses are weighted and calibrated to recognised population frames, so that groups which are easier or harder to reach are represented in their true proportion rather than as they happened to fall in the sample. Where a study estimates voting behaviour, a consistent screen is used to focus on those likely to vote, applied the same way each time so results stay comparable across studies.
A central web platform turns incoming data into live monitoring, disaggregated analysis and district-level reporting—and a standing feedback loop keeps the method sharpening from one study to the next.
The same platform that receives the data also acts as a command centre. From it, the team and its stakeholders can follow a study's progress, the profile of who has been reached, and the trends emerging in the responses. Analysis is built to be cut finely—by district, region, urban and rural segment, and demographic profile—and written up as district-level reporting with recommendations addressed to the departments and agencies that can act on them.
How a finding is expressed depends on the work. Where a study measures a programme, the reporting maps awareness, coverage, satisfaction and how much of the intended benefit is reaching the last mile. Where it estimates an outcome such as an election, figures are presented with their uncertainty rather than as a single, falsely precise number.
The method itself is treated as a living thing. Observations from surveyors, stakeholders and analysts are gathered after each study and folded back into the design, so the apparatus stays aligned with conditions in the field rather than frozen at the day it was built.
The same apparatus runs from a single constituency to a multi-state, year-long programme—staffed in defined roles, supervised throughout, and buffered for the way fieldwork actually unfolds.
Large assignments are scoped as a structure, not a crowd. Field surveyors are organised under supervisors, supported by a quality and data desk and held together by coordination—so that adding states or sample does not dilute the checks that make the data trustworthy.
Plans are built with room for reality. Leave, weather, electoral restrictions, public holidays and local festivals are anticipated rather than discovered mid-field, so a programme can hold its standard from the first week to the last.
For government-scheme work, the same machinery is pointed at a clear set of questions: who knows about a scheme, who it actually reaches, how satisfied beneficiaries are, where implementation falls short—and what a publicity campaign returned for what it cost.
Standards exist so that a firm cannot quietly grade its own homework. We work to documented procedures, keep client and respondent data confidential, and stay open to outside review.
Studies can be commissioned at the constituency, state or national level—from opinion and exit research to impact assessment of public schemes and communication campaigns. Method notes and supporting documentation are available on request.
Every interview is geo-tagged, time-stamped and open to back-check from the moment it is recorded.
How a study was sampled, validated and analysed is written down and travels with the work.
Findings reflect what the field returns—not the interests of whoever commissioned the study.