The social impact assessment (SIA) process varies by sector and country, but most well-run assessments follow the same sequence. This guide walks through each step as it applies to NGOs, social enterprises and their funders. For what an SIA is and when it is needed, see Social Impact Assessment (SIA): A Practical Overview.
1. Screening
Screening decides whether a full assessment is needed and how deep it should go. A small training program may need only a light review. A project that changes land use, displaces activities or affects large populations needs a full SIA. Funder and national requirements often set the minimum.
2. Scoping
Scoping defines the boundaries: which communities and groups are affected, which types of effects matter most, the geographic area and the time frame. A stakeholder map is built at this stage, including groups that are easy to overlook, such as women, young people, migrants or people with disabilities.
Scoping is also where the assessment’s questions are set. Who is the target population and how is it made up? What barriers do they face? Which groups might be left out? What local institutions and services already exist? What could go wrong?
3. Baseline Study
The baseline describes current conditions using the indicators that will be tracked later. Before collecting new data, gather what already exists: national statistics, census figures, previous evaluations, partner reports and local studies. This shows where the real information gaps are and keeps fieldwork focused. New data then comes from surveys, interviews and group discussions, using methods that suit literacy levels, languages and connectivity in the area.
Choosing Indicators You Can Actually Collect
Start from outcomes: what should be different for people if the project succeeds? Keep the list short and specific. “Participants find work” is a start. “Participants hold a paid job six months after training” is something you can measure. For each outcome, choose one or two indicators and check that the data can realistically be collected with your team’s time, tools and access to participants. An indicator that requires follow-up visits the budget cannot cover will produce gaps, not evidence. Where funders use standard metrics such as IRIS+, align with them to make reporting easier.
4. Stakeholder Engagement
Engagement runs through the whole process, but it is most intensive here. Affected people are consulted on how they expect the project to affect them and what matters most to them. Use a mix of methods suited to the setting: interviews with local leaders and service providers, group discussions with different parts of the community, and short surveys where useful. Hold separate sessions for groups that may not speak freely in mixed settings, and work in the languages people use day to day.
Good practice includes informed consent, safe ways to raise concerns, and feedback to participants on how their input was used.
5. Impact Prediction and Analysis
The team assesses likely effects, positive and negative, and how significant each one is. Significance usually depends on the scale of the effect, how many people it touches, how long it lasts, whether it can be reversed and how vulnerable the affected group is.
Bring the findings together around the project design. Which groups are likely to benefit most and least? Are there social tensions the project could worsen? Are there local strengths, such as existing savings groups or community organizations, it could build on?
Combine Numbers and Voices
Quantitative data shows how many people changed and by how much. Qualitative data, from interviews, open survey questions and group discussions, explains why. Both are needed. Ask participants and partners about unexpected effects too. A skills program might raise participants’ confidence in ways no indicator captured, or it might take time away from family income during training. Both belong in an honest assessment.
Be Careful About Cause and Effect
Other factors influence outcomes: the local economy, other programs, family circumstances. Unless you have a comparison group, describe the project’s contribution to the change rather than claiming it caused the change on its own.
6. Mitigation and Enhancement
For each significant negative effect, the assessment sets out how it will be avoided, reduced or compensated. For positive effects, it identifies ways to strengthen them, for example by prioritizing local hiring or adjusting delivery so that harder-to-reach groups benefit.
An assessment is only useful if it changes something. Translate each key finding into a decision: adjusting eligibility criteria, changing delivery times so that women can attend, adding a translation budget, or partnering with a local organization that already has community trust.
7. Management and Monitoring Plan
Commitments only matter if someone tracks them. The plan names indicators, data sources, frequency, responsibilities and budget. It also defines how a grievance mechanism will work and how findings will reach decision-makers.
8. Reporting and Review
The final report presents the baseline, predicted effects, mitigation measures and monitoring arrangements in language decision-makers and communities can follow. A fuller report for funders should describe the methods, the main findings, the design changes made and any limitations in the evidence. Share a short summary with the communities and partners who took part, in local languages, explaining what was heard and how it shaped the project.
Periodic reviews during implementation compare actual effects with predictions and update the plan. Over time, this record becomes the strongest evidence an organization has.
A Lighter Version: Social Assessment for a Program
A social assessment for a program is lighter than a full SIA for a large project, but it serves a similar purpose: understanding the people a program will reach, the context it will operate in, and the risks and opportunities that should shape its design. It is most useful before a program starts or before it expands to a new area.
The same sequence applies in compressed form: set the questions, review what is already known, talk to the people involved, analyze risks and opportunities, turn findings into design choices, set up a few indicators to monitor, and report back. For a small program, this can be completed in a few weeks with a small team.
Keeping the Process Proportionate
Not every project needs every step at full depth. The aim is an assessment that matches the project’s scale and risk, produces evidence the team will actually use, and treats affected communities as participants rather than data sources. The goal is not a long document but a project that fits its context and a team that understands the people it serves.
Frequently asked questions
Most well-run assessments follow eight steps: screening, scoping, a baseline study, stakeholder engagement, impact prediction and analysis, mitigation and enhancement, a management and monitoring plan, and reporting and review. How deep each step goes depends on the sector, the country and funder requirements.
Screening decides this. A small training program may need only a light review. A project that changes land use, displaces activities or affects large populations needs a full assessment. Funder and national requirements often set the minimum.
Only with care. Other factors, such as the local economy, other programs and family circumstances, also influence outcomes. Unless there is a comparison group, describe the project’s contribution to the change rather than claiming it caused the change on its own.
