Needs assessment: start with the community

A needs assessment is a structured look at a community’s problems, strengths and priorities, carried out with the people affected and with the best data available, before anyone designs or funds a programme. It answers three questions: what is wrong, for whom, and what would make the biggest difference.

Done well, it saves money and goodwill. It stops a company building a clinic that families won’t use, or an NGO running evening classes in harvest season. It also gives you the facts that a baseline will later measure precisely, and the problem statement your theory of change starts from.

This guide is for CSR teams, company foundations and NGO programme staff: the data to read first, how to listen to a community and work with local government, ethics, and how to turn findings into priorities, with a template at the end.

Why a needs assessment matters

A needs assessment isn’t a wish list or a single survey. It brings together what the data says, what people say about their own lives, and what services already exist. It records strengths as well as problems: an active self-help group or water committee is something to build on. For companies, it is also part of how the law expects CSR to be planned.

The words “if any” matter: the Rules don’t require a needs assessment for every project. But a plan that must describe need is hard to write honestly without one, and a company that later commissions an impact assessment will be asked what the project set out to change. Our guide to the CSR policy and the annual action plan covers the plan itself.

There is a second reason. The Ministry of Corporate Affairs’ FAQs say CSR “should not be interpreted as a source of financing the resource gaps in Government Schemes” (FAQ 3.17). A needs assessment that maps what the government already provides helps you design work that complements a scheme instead of paying its bills. See CSR and government schemes.

Five steps, from desk to village

Budget time for step 5: it is the one most often skipped.

Start with public data

India publishes a great deal of free data down to district level and below. Read it before you go into the field, so your questions are sharper and you don’t ask families for facts the government already holds.

A few notes on the most useful sources:

• NFHS-6. The sixth National Family Health Survey was carried out between May 2023 and December 2024 in 715 districts, with the International Institute for Population Sciences (IIPS), Mumbai as the nodal agency. Its India and state fact sheets were released on 29 May 2026 and its district fact sheets on 20 and 21 August 2026. They can be downloaded from the NFHS website with a guest login. Compare them with NFHS-5 to see which way things are moving.
• The Census. It is the only full count of every person and household, village by village, but the latest results are from 2011. Census 2027, the first digital census, is under way: houselisting was scheduled for April to September 2026, and population enumeration for February 2027 (September 2026 in Ladakh and the snow-bound areas of Jammu and Kashmir, Himachal Pradesh and Uttarakhand). Until its results are out, use 2011 figures for village detail and say how old they are.
• UDISE+ and PARAKH. The Ministry of Education’s UDISE+ 2025-26 report came out on 7 July 2026. NCERT’s learning survey of 4 December 2024 tested about 21 lakh students in grades 3, 6 and 9 across 781 districts, which shows whether children in school are actually learning.
• Mission Antyodaya. An annual survey of gram panchayats that supports their development plans.

Collect your own information

Public data tells you what; people tell you why. Use more than one method: each hears different voices.

The participatory methods, often called participatory rural appraisal or PRA, work well in villages. A map drawn on the ground by residents shows which hamlets the handpumps don’t reach; a seasonal calendar shows when families migrate for work.

Make sure you hear from the people least likely to come to a meeting: women, older people, people with disabilities, Scheduled Caste and Scheduled Tribe households, landless families, seasonal migrants and the hamlets far from the main village. Meet at times and places that suit them, in their language. For a household survey whose numbers you will rely on, work out the sample size properly: our guide to baselines and evaluation explains how.

Involve the community and local government

A needs assessment done to a community is weaker than one done with it. Involve these people from the start:

• The gram sabha and gram panchayat. Every gram panchayat prepares a development plan through its gram sabha, the meeting of all adult voters in the village. Ask to see the current plan and the panchayat’s Mission Antyodaya gap report.
• Frontline workers. ASHAs (accredited social health activists), auxiliary nurse midwives, anganwadi workers, teachers and self-help group leaders know which families are missed and why.
• Block and district officials. The block development officer, the district education officer, the child development project officer who oversees the block’s anganwadis, and the medical officer at the primary health centre. They can tell you what is planned and funded, and where delivery falls short.
• In towns and cities, ward committees, municipal officials and urban health centres play the same part.

The aim is work that fits with what government provides, adds what is missing and can be sustained later, rather than a parallel service that stops the day the CSR money does.

Ethics and consent

People give you their time and often personal information. Treat both with care.

• Explain and ask. Tell people who you are, why you are asking, how the answers will be used and who will see them. Ask for their consent, in their language, and make clear they can refuse or stop at any point.
• Don’t promise what you can’t deliver. A needs assessment raises hopes. Say plainly that you are listening, not yet committing.
• Children need extra care. Ask a parent or guardian for consent and the child for their own agreement, use trained staff, and never interview a child alone. Follow your child protection policy.
• Ask sensitive questions only if you will use the answers. For caste, disability, religion, health or violence, explain why you are asking, let people describe themselves or decline, and know where to refer someone who tells you they are being harmed.
• Protect the data. Collect the minimum, store it securely and remove names before sharing. The Digital Personal Data Protection Act, 2023 and its 2025 Rules apply to NGOs too, with most duties starting on 13 May 2027; see data protection for NGOs and consent, photos and dignity. Never publish figures for groups so small that people can be recognised.
• Health measurements need more. If you weigh children or test for anaemia, use trained staff, arrange referrals for anyone who needs treatment, and consider review by an ethics committee.

Turning findings into priorities

By now you will have more needs than money. A simple scoring grid, the one the Academy’s foundation course uses, makes the choice explicit. Score each option from 1 to 5 on six questions. Need: how severe is the problem, on the best data? Fit: does it connect to the company’s places, people or skills, without serving its business? Evidence: is there good evidence about what works? Partners: are there capable NGOs and government partners? Scale: can the budget make a visible difference within five years? Voice: do communities themselves rank it as a priority?

The foundation dropped the clinic. It now funds a local NGO to restore village ponds and wells with the panchayats, and an early-learning programme that trains anganwadi workers with the child development project officer’s support. The grid didn’t decide for it. It made the reasons visible, so the CSR committee and the villages could argue about the right things.

A simple needs assessment template

Keep it short, with detailed tables in annexes: a CSR committee reads ten pages, not a hundred.

Common mistakes

• Deciding the project first and using the assessment to justify it.
• Talking only to leaders: the sarpanch, the head teacher or the company’s own local staff.
• Asking leading questions, such as “Would you like a clinic?”, which almost everyone answers yes.
• Collecting data nobody analyses, or ignoring what already exists and duplicating a scheme.
• Never going back to tell people what you found and what you decided.

Once the priorities are agreed, the next steps are a theory of change for how the work will lead to change, and indicators to measure it.