01 · What breaks
Design for the failure modes, not only the happy path.
Household surveys are easy to start and hard to justify. When the data never comes back to the village, or reveals the wrong thing to the wrong person, trust falls and the next round of answers gets worse.
Extractive surveys with no return value
A family that answers forty questions and hears nothing again learns to give short answers. Plan what goes back before the survey goes out: a hamlet summary at the next meeting, a list of scheme applications, a referral that actually happens. If nothing is planned, cut questions until something can be.
Sensitive labels visible to neighbours
Field teams often come from the same village, and phones get passed around. Labels such as widow, debt, alcohol use or migrant status should be visible only to the roles that act on them. Show a volunteer the task (visit, refer, follow up) rather than the reason behind it.
Old profiles treated as current
Households migrate, marry, split and recover. A profile collected two years ago and shown as if it were today sends help to the wrong door. Date every answer, set a review window for each type of field, and grey out anything past its window on dashboards and lists.
Loudest voices represented as consensus
A priority list from one meeting often reflects who could attend: men, landowners, the hamlet nearest the road. Record attendance by group at every priority-setting meeting, and flag any priority set without women, a hamlet or a marginalised group present, so the team knows whose view the list holds.
02 · Minimum viable model
Connect each household to its place, its priorities and what was done.
Community profiles fail when they are built as a one-time census. These six records let a team move from a household's answers to a community priority, an owned action and a review date, and each field is named against the decision it serves.
| Field you are tempted to collect | Ask first | Keep it only if |
|---|---|---|
| Caste category | Which scheme or service will use it? | A named scheme application needs it, and only the role filing that application can see it |
| Aadhaar number | Does any service here need the number itself? | Rarely. A yes or no on whether the household holds the document usually does the job |
| Household income | Would a rough band change a decision? | A band, not an exact figure, decides eligibility or priority |
| Photos of the home | Has a funder required it, and has the family agreed? | A funder requires it and photo consent is recorded separately |
| Details of children in the household | Which children's service depends on them? | A service uses them, and consent is recorded from a parent or guardian |
Under India's Digital Personal Data Protection Act, 2023, processing the personal data of anyone under 18 needs verifiable consent from a parent or lawful guardian (as of October 2026). Most of these duties apply from around May 2027, as of October 2026. When the person answering is not the child's parent or guardian, record that and confirm the approach with a legal adviser.
03 · Implementation sequence
Move through reviewable gates.
Start with the questions the team and the community will act on this season, and let the survey grow only when a new decision needs it. For the field form itself, see our guide to a data collection app for nonprofits.
Agree which decisions need data
List the decisions the programme will make with the profile in the next twelve months: which households to help with scheme applications, which hamlets need a water intervention, where to hold a health camp. Every proposed question must point to one of them. Questions that cannot are dropped before the form is drafted.
Co-design categories and consent
Test the answer options with community members and field teams in the local language. Categories such as landless or single woman mean different things in different places. Write the consent script in plain words, record who consented and to which uses, and keep photo consent as its own question.
Establish household and place identity
Build master lists of villages, hamlets and landmarks before the first visit, so nobody types three spellings of one place. Give each household a stable ID and check duplicates on the head of household, a second member and the hamlet. Link members to the household instead of copying them.
Connect priorities to actions
Each community priority becomes a record with an owner, a target date and a status, linked to the households and assets it concerns. A scheme referral is tracked to an outcome: applied, sanctioned, received, or rejected with a reason. Without that link the profile shows need but never what was done about it.
Review representation and missing voices
Before each priority review, the system lists households not yet visited, hamlets with low response and groups missing from meetings. A field coordinator decides whether to revisit or hold a separate meeting. The gaps go into the meeting notes, so the decisions are read with their limits.
Retire stale or unnecessary fields
After one full cycle, the programme lead reviews every field and asks whether it was used in a decision, a referral or a report. Unused fields leave the next round, and expired answers are archived. A shorter survey is the clearest sign to a community that its time is valued.
04 · Evidence
Measure whether the work became clearer.
Measure whether the profile is fresh, fair and acted on. A high household count with no linked actions is a survey, not a system. The wider programme model sits in our community development playbook.
05 · Responsible automation
Let systems prepare. Keep consequence with people.
Systems can support
- Village and hamlet master lists
- Duplicate household checks
- Freshness flags on dated answers
- Lists of unheard groups and hamlets
People should decide
- Community priority setting
- Handling disputed information
- Sensitive household referrals
- Publishing or sharing local data
06 · Common questions
Practical questions teams ask.
Questions programme leads ask before a household survey goes into the field.
What is a household profiling system for a nonprofit?
It is a dated record of households, their members, needs and strengths, linked to the village institutions they use and to the actions taken on what they said. It replaces one-off survey spreadsheets with a profile that can be reviewed, corrected and followed through to scheme applications, referrals and community priorities.
How many questions should a household survey have?
As few as the decisions require. Every question costs minutes of a field visit and a family's patience, and answer quality tends to fall as forms grow long. Start from the decisions the programme will make this year, keep only the questions that feed one, and move anything else to a later round or a smaller sample.
How often should community profiles be updated?
It depends on the field. Household composition and migration can change within a season, so confirm them at each visit. Housing and assets change slowly and can be checked once a year. Set a review window for each field type, show the date beside every answer, and treat anything past its window as unconfirmed rather than true.
Should we collect Aadhaar or caste in a household survey?
Only if a named service or scheme application needs it, and then only visible to the people filing it. For most profiling, recording whether the household holds the document is enough. Collecting sensitive identifiers just in case adds risk without adding value, and families notice when questions seem unrelated to any help they will receive.
How do you make sure community priorities are not set by a few voices?
Record who attended each priority meeting by group and hamlet, and compare it with the households in the profile. Flag gaps before decisions are made, and hold separate meetings with women, distant hamlets or excluded groups where needed. Keep the attendance record attached to the priority so anyone reading it later knows whose view it reflects.