A poorly composed focus group produces a façade consensus that a due diligence mission dismantles in one morning. A well-conducted focus group produces verifiable evidence of what communities have understood, feared and requested. The difference does not lie in the budget, but in the field methodology: whom one invites, in which language one facilitates, how one neutralises power relations, and how one documents what has been said. This article details this methodology, step by step, for consultations that withstand lender scrutiny.

Why focus groups, and when they fail

The village assembly brings everyone together. This is its strength for informing, and its weakness for listening. In a large meeting, it is the notables, adult men and the most educated who speak. Women, young people, seasonal herders, landless people remain silent. Their concerns never enter the minutes.

The focus group corrects this bias. It is a structured discussion, in a small homogenous group of eight to twelve people, on a limited number of topics. Homogeneity is the key point. Bringing women together among themselves, young people among themselves, users of a resource among themselves releases speech that the plenary meeting stifles.

It fails when treated as a formality. Single group, participants chosen by the chief, facilitation in the official language alone, duration of twenty minutes before the photograph. The consultation box is then ticked without ever having listened. This is exactly what lenders' E&S teams know how to detect.

This quality requirement flows directly from IFC PS1, which governs stakeholder engagement. The standard distinguishes simple information from genuine consultation. For projects with significant impacts, it requires in-depth consultation, known as informed consultation and participation, which must be "free of external manipulation, interference, coercion, or intimidation" (IFC, Performance Standard 1). A focus group held under the gaze of the village chief does not meet this criterion.

Composing the groups: sampling that builds credibility

The first decision is composition. It is prepared before arriving in the field, from the stakeholder mapping. Without this map, one does not know who is missing.

The principle is to segment the affected population into homogenous groups, then to hold one focus group per segment. The segmentation criteria depend on the project, but several axes always recur.

  • Gender, with at a minimum one women's group facilitated by a woman.
  • Age, with a dedicated group for young people, often those most concerned by employment and most absent from meetings.
  • Relationship to the affected resource: farmers, herders, fishers, forest users, each has a different loss to express.
  • Land tenure status, by isolating the landless, sharecroppers and customary occupants without title.
  • Vulnerability, by providing space for elderly people, people with disabilities, female-headed households.

The classic error is to let the local authority designate participants. It then selects its associates and excludes critical voices. The team keeps control of invitations, relying on multiple intermediaries: technical services, women's organisations, producers' groups, religious leaders. The number of groups is sized to the scale and heterogeneity of the area, not to a number fixed in advance.

Facilitating in local language without losing meaning

A consultation held in a language that half the room barely masters is not a consultation. It is a presentation. Language is the first credibility test of a rural focus group.

Facilitation takes place in the participants' actual working language, not in the official language of the file. This often requires an interpreter. The choice of interpreter is not neutral. A government official or an associate of the project owner inhibits speech. A person external to the village, trained beforehand in the project vocabulary, inspires greater confidence.

Translating technical concepts is an exercise to prepare. Terms such as compensation, easement, environmental flow, economic resettlement or right-of-way do not always have direct equivalents. One must construct, with the interpreter and before the session, a way to explain them through concrete examples. A poor translation creates a misunderstanding that emerges as a complaint two years later.

The facilitation guide remains short. Five to seven open questions, in a logical order, from general to sensitive. One opens on village life and the resource, advances towards the project and its perceived effects, ends on fears and requests. The facilitator probes, reformulates, verifies understanding. He or she does not defend the project. A focus group is not a sales meeting.

Managing power relations in the room

Even in a homogenous group, hierarchies remain. A mother-in-law and her daughter-in-law, a master fisher and his crew, an elder and a junior do not speak freely in front of one another. The facilitator must see these lines of force and defuse them.

The first lever is separation. If tensions are strong between two sub-groups, one holds two sessions rather than one. The second lever is framing. One announces at the opening that no statement will be attributed by name, that disagreement is welcome, that there is no expected correct answer. The third lever is facilitation technique. A systematic round table, an instruction that invites the most silent to speak first, a reformulation that validates minority views.

The venue and timing weigh as much as facilitation. A women's session scheduled at market time or at water-fetching time will be held without women. One sets the time according to their constraints, not those of the team. One chooses a neutral venue, away from the chief's residence and the project owner's office.

One must also accept silence and disagreement as valid results. A room that approves everything, quickly and without nuance, has probably not understood or does not dare speak. A good report captures objections, doubts and precise requests. It is this material that then feeds the grievance mechanism and, in rural areas, its design closest to local practices.

A consultation that never returns to the communities is not complete. Feedback is the stage most often skipped, and the one that distinguishes genuine listening from a communication exercise.

Providing feedback means returning on site to explain what has been noted, what has been integrated into the project and what has not, with reasons. This feedback loop has dual value. It respects participants who have given their time. It proves to the lender that the consultation has influenced decisions, and has not served as window dressing. A concern raised and then visibly addressed is the best sign of a useful consultation.

Documentation, for its part, is built during and after each session. It does not reduce to a photograph and an attendance sheet. A robust file brings together several elements.

  • The facilitation guide used, proof that sensitive topics were raised.
  • The list of targeted segments and the participation rate of each.
  • A report per group, restoring views, including divergent ones, without attributing them by name.
  • Evidence of the facilitation language and recourse to an interpreter.
  • The register of concerns, with the response provided to each.
  • The report of the feedback session.

This documentary requirement is not a lender caprice. The AfDB Operational Safeguard 1, like IFC PS1, requires meaningful consultation and evidence of its conduct (AfDB, Integrated Safeguards System). For projects affecting land-dependent communities or indigenous peoples, this evidence becomes the raw material for demonstrating consent. On this point, the focus group feeds into the logic developed in our dedicated article on genuine and documented community consent.

A credible focus group is not improvised on the morning of the session. It is prepared from a stakeholder mapping, composed in homogenous groups that release weak voices, facilitated in the participants' actual language, protected from power relations, and documented piece by piece.

Three reflexes avoid campaign rework. Segment rather than bringing all together in plenary, to hear those who ordinarily remain silent. Facilitate in local language with a neutral interpreter, so that speech is informed. Provide feedback and record, to transform a discussion into evidence of constructed consent. Consultation is not a box to tick before signing. It is the foundation on which rests the entire relationship with communities during the project's life.

ps1consultationfpicsauvegardes-operationnellesengagement-parties-prenantesfocus-group