Define success before optimising the programme
A public agency or social organisation may be trying to improve employment, reduce waiting times, expand access or strengthen educational outcomes. Revenue and cost still matter, but they do not fully define success.
Quick Prep’s social and public sector preparation teaches you to translate a mission into a decision that can be analysed. Establish the target population, the desired outcome, the time horizon and the budget. Then compare credible ways of achieving that outcome while recognising delivery constraints and distributional effects.
The OECD’s evaluation criteria consider whether an intervention is appropriate, coherent, effective, efficient, impactful and sustainable. They are a useful reference for broadening the discussion beyond expenditure and activity counts. OECD: Purpose and role of evaluation criteria
Distinguish inputs, activities, outputs and outcomes
A training programme’s funding is an input. Delivering workshops is an activity. Completing courses is an output. Sustained employment is a possible outcome. The long-term contribution to participants’ wellbeing or local economic conditions is a broader impact question.
This distinction changes the analysis. A programme can exceed its attendance target without improving the outcome it exists to achieve. Conversely, a more intensive intervention may serve fewer people but create larger benefits for a group with greater needs.
Sketch a simple theory of change: the resources provided, the services delivered, the behaviour or condition expected to change, and the ultimate objective. Identify assumptions between each stage. For example, interview training may improve job-search skills, but employment also depends on suitable vacancies and access to transport or childcare.
Understand the delivery system
The funder, operator, participant and evaluator may be different organisations. Map responsibilities and incentives before recommending a change. A central agency may control funding while local partners control delivery, and another institution may hold the data needed for evaluation.
Segment the population in ways that matter to access and outcomes. Geography, service need and ability to use digital channels can create different delivery requirements. A lower average cost can conceal exclusion of harder-to-serve groups.
For a public-service operations case, examine demand, eligibility, applications, assessment, delivery and appeals or follow-up. Ask where queues build and whether the constraint is staffing, information, decision authority or process design.
| Measure | Useful question |
|---|---|
| Cost per participant | What resources does the service consume? |
| Completion rate | Who receives the intended service through to the end? |
| Outcome rate | What changes for participants? |
| Incremental outcome | What changes because of the programme? |
| Waiting time and access | Who can obtain the service, and how quickly? |
| Outcomes by group | Are benefits and barriers distributed unevenly? |
Worked exercise: compare employment programmes
Illustrative interview exercise. Figures and evaluation assumptions are fictional.
A funder compares two programmes, each costing 1 million and serving 500 people. Programme A reports that 60% of participants achieve sustained employment. A credible comparison estimate suggests 40% would have done so without the programme.
Estimated incremental outcomes are 500 × (60% − 40%) = 100 additional sustained jobs. Cost per additional sustained job is 10,000.
Programme B reports 55% employment, with an appropriate comparison estimate of 25%. Its estimated incremental outcomes are 500 × (55% − 25%) = 150, implying a cost of approximately 6,667 per additional sustained job.
Programme B has a lower headline employment rate but a larger estimated incremental effect. The counterfactual changes the interpretation.
The arithmetic alone does not establish which programme should receive more funding. Check whether the comparison groups are credible, employment is defined consistently, outcomes are sustained for the same period and the estimates have material uncertainty. Participants may face different barriers, and neither programme’s result automatically scales to a much larger population.
A recommendation could fund a staged expansion with evaluation milestones rather than assume constant performance at scale.
Structure the recommendation around public value
A useful decision framework has four parts. First, assess the magnitude and distribution of the need. Second, compare interventions and the strength of evidence. Third, test delivery capacity and whole-programme cost. Fourth, define how progress and unintended effects will be measured.
For digital public services, include assisted access and the consequences of mistakes or service outages. For infrastructure, consider the beneficiaries, usage assumptions and operating obligations after construction. For a nonprofit, distinguish fundraising efficiency from mission effectiveness; low administration spending alone does not demonstrate impact.
Avoid presenting stakeholder alignment as a vague final bullet. Identify who must approve, fund, deliver, use or maintain the intervention, and what could prevent each group from participating.
Practise explaining a funding choice
Choose a public problem and write an objective that names the population, outcome and timeframe. Build two plausible interventions and list the data needed to compare them. Explain what you would recommend if the evidence were incomplete.
Aim for a decision that is measurable and implementable while acknowledging uncertainty. Continue with Market Analysis for population and demand estimation, and Fit Interview Preparation for discussing stakeholder influence and responsible judgment.
