BCG and the Gaza Humanitarian Foundation: The Aid Model

A humanitarian objective does not complete an engagement review

An engagement can have a stated humanitarian purpose and still require difficult decisions about authority, safety, access and participation. The purpose describes an intended benefit; it does not establish that the operating model delivers it or that the people affected have been adequately considered. Those are separate questions that need evidence and accountable decision-makers.

For consultants, the acceptance brief should identify the service being provided, the actual counterparty, the intended beneficiaries and the institutions whose involvement the plan assumes. Record which facts are confirmed and which are representations made by a sponsor. If the legitimacy of an arrangement depends on multilateral participation, the existence of that participation needs verification rather than repetition in a reassuring project description.

The South Africa state-capture case also asks what a professional-services firm should understand about a client’s institutional setting. The cases have different facts and legal histories. The comparison is useful at the decision-process level: technical competence cannot substitute for an accurate understanding of the engagement’s authority and consequences.

Follow the counterparty as well as the fee

The organisation signing a contract, the sponsor directing work and the people affected by the output can be different. A useful engagement map shows those relationships separately. Otherwise, a team can describe itself as working for one initiative while the contractual route and operational instructions come through several other organisations.

Financial analysis needs the same care. An intended fee, a signed commitment, an invoice, an invoice cancellation, internal labour cost and cash received are different measures. A model that combines them into a single revenue number does not become more informative because the number is prominent. The page retains the firm’s fee position rather than turning reported contract values into an asserted payment.

The Puerto Rico case provides a parallel lesson about tracing entities and monetary measures without assuming what the trace will prove. A clear diagram can show several relationships and leave the conclusion proportionate. It should not turn the existence of a commercial chain into an unsupported claim about who controlled every downstream activity.

Pro bono work still uses authority and institutional reputation

No-fee work consumes people, time, management attention and an organisation’s credibility. It can also produce recommendations that a recipient uses to justify decisions. The absence of an invoice therefore cannot be the reason to relax acceptance, evidence retention or escalation. It changes the commercial question, not the fact that an institution is participating.

In a fictional pro bono project, a team might agree to assess the feasibility of a proposed service. Later it is asked to design procurement, develop operating assumptions and prepare a presentation for a sponsor. Each request may sound adjacent to the original brief, but the cumulative change can be substantial. A review process should ask what the team is now doing rather than whether the billing code still says pro bono.

This is a practical matter for staffing and governance. Name the person who can approve follow-on work, identify what changes require reapproval and give junior team members a route for raising a mismatch between the approved brief and the actual tasks. These are proposed controls, not a reconstruction of every internal conversation in this case.

Treat a scenario model as a set of choices

A model is not neutral simply because it uses a spreadsheet. It contains choices about the alternatives considered, costs included, outcomes valued and constraints treated as fixed. A comparison can look precise while leaving the most consequential assumptions outside its boundary. Reviewers need to understand the question the model was permitted to answer before interpreting the numerical result.

For a sensitive population-related scenario, distinguish an assumption about a voluntary choice from evidence that people can make that choice safely and freely. A cost difference does not establish desirability, legitimacy or implementation feasibility. The analyst should record whose perspective defines cost and whose losses may be excluded. These questions apply to model design; they are not findings that a particular published scenario was implemented.

The Saudi report case examines another boundary between an analytical output and the people described by it. The shared question is what a report makes possible and how a team considers foreseeable misuse. That comparison does not establish a causal connection between the cases or resolve disputed facts in either one.

A stop instruction needs an operational record

An instruction not to proceed is effective only if the relevant people receive it, understand its scope and know what to do with work already underway. A robust process identifies which work must stop, which materials must be retained and who can authorise any restart. It should not leave the team to infer whether a narrower task is still permitted.

The record should distinguish refusal, conditional approval and approval of a different service. If an exception is granted, state the conditions and evidence required. If the scope changes again, the exception should not silently expand with it. A decision log allows later reviewers to see what was known at the time rather than reconstructing permission from conflicting recollections.

Communications retention serves that same purpose. The relevant issue is whether instructions, model versions and approvals can be reconstructed when needed. Teams should use the channels and retention arrangements appropriate to their obligations. The case does not justify assuming that every informal message is improper, but it illustrates why an evidence trail cannot be an afterthought.

Keep humanitarian consequences and consultant responsibility distinct

An engagement can deserve scrutiny because it operates in a setting where the stakes are grave. That does not remove the need to establish who did what, what influence an output had and what evidence connects a particular action to a particular outcome. Humanitarian context and causal attribution belong in the same account, but they should not collapse into one assertion.

The ICE detention-budget case similarly asks whether an efficiency objective leaves important consequences outside the model. A consultant should examine the objective and its constraints, while a writer should distinguish a proposal, an adopted measure and a documented outcome. The requirement for evidence becomes more important when the consequences are serious.

A clear article can therefore present company explanations, parliamentary questions and the broader operating context without declaring that a corporate review is a trial. It can explain why a reported scenario attracted criticism without treating the scenario as an implemented policy. This is a standard of attribution and reasoning, not a reason to minimise the subject’s human importance.

Build an acceptance decision that survives a change in the story

The strongest engagement review can be revisited when a new counterparty appears, a service becomes operational or an assumption about participation proves wrong. It gives the team a practical route to pause, escalate or decline work. It also makes clear which conditions are essential, so that losing one of them changes the decision rather than merely changing the wording of a slide.

Use the management consulting presentation guide to make that decision explicit: the proposed scope, the evidence supporting it, the people affected, the unresolved questions and the accountable owner. The useful lesson from The Aid Model is to connect those elements before the organisation’s name, tools and effort become part of a project it does not fully understand.