Project cycle: design, planning and performance management

1The phases of the cycle and the decision gates

The project cycle has five phases: identification, design, implementation and monitoring, evaluation, closure and learning. The PMD Pro guides (PM4NGOs) add a logic of decision gates: at the end of each phase, a body decides whether to continue, adjust or stop. In practice, many organisations never stop a project during design because a donor is interested; this is one of the main causes of failure.

The three classic constraints (scope, time, cost) are complemented in humanitarian work by quality (Sphere standards, CHS), risks (security, protection, fraud) and accountability. Any change to one constraint affects the others: reducing the timeline without reducing the scope raises the cost or degrades quality.

2From the problem tree to the logframe

The problem tree organises the causes (roots) and effects (branches) of a central problem. Converted into an objective tree, it reveals several possible pathways; the options analysis chooses the one that matches the priority needs, the organisation's added value, what other actors are doing, cost and risk.

The logical framework then organises the project into four levels: overall objective (impact), specific objective (outcome), results (outputs) and activities. Its vertical logic is checked with the "if... and... then" test: if the activities are carried out and the assumptions hold, then the results are achieved. The horizontal logic associates indicators, means of verification and assumptions with each level.

Three mistakes get a proposal rejected by a demanding reviewer: a specific objective that repeats an activity ("distribute kits") instead of a change ("households apply key hygiene practices"); assumptions that are in fact risks the project should manage ("staff are recruited on time"); indicators with no baseline or target.

A good assumption is an important external factor, likely but not certain, that the project does not control. If it is very unlikely, it is a killer assumption: change the design, do not write it in the column.

3Planning: breakdown, critical path and resources

The work breakdown structure (WBS) breaks each result down into work packages precise enough to be estimated, assigned and tracked. From the WBS, you estimate durations, identify dependencies (for example, the baseline survey must precede targeting, and targeting must precede distribution) and calculate the critical path: the longest sequence of dependent tasks. Any delay on a critical-path task delays the whole project.

Example: recruiting enumerators (10 days), then baseline survey (15 days), then analysis and targeting (10 days), then distribution (20 days) = 55 days. In parallel, procuring the kits takes 45 days but must be completed before distribution, which starts on day 35: it therefore has a negative float of 10 days. Procurement becomes critical: it must be launched before the survey ends, based on estimated quantities, or through a framework agreement.

Resource planning must include constraints specific to humanitarian work: rainy seasons and access, the agricultural calendar, religious holidays, import and customs clearance times, donor approval times (often 4 to 8 weeks for a modification), and the real workload of teams already committed to other projects.

4Steering with earned value and burn rate

Simple budget monitoring (spent versus budget) is misleading: a project may have spent 50% of its budget at mid-term while having completed only 30% of the work. The earned value method compares three figures:

• Planned value (PV): budgeted cost of the work scheduled to date.
• Earned value (EV): budgeted cost of the work actually completed.
• Actual cost (AC): what has actually been spent.

From these come the schedule performance index SPI = EV / PV and the cost performance index CPI = EV / AC. An index below 1 signals a delay or an overrun. The estimate at completion is often calculated as EAC = total budget / CPI, if the observed performance continues.

Worked example. Total budget USD 800,000 over 12 months. In month 6: PV = 400,000, EV = 320,000, AC = 360,000. SPI = 320 / 400 = 0.80 (20% behind the planned work). CPI = 320 / 360 ≈ 0.89 (you pay USD 1.12 for USD 1 of work). EAC = 800,000 / 0.89 ≈ USD 900,000: without action, the project will overrun by about USD 100,000 or will not complete all the work.

Manager's reading: a low SPI with a CPI close to 1 signals a capacity or access problem (less work done, but at the right price). A low CPI signals a problem of price, productivity or, sometimes, fraud. The two call for different actions.

5Managing changes, stakeholders and closure

No humanitarian project unfolds as planned. Change management requires a log in which each change is described, assessed (effects on scope, cost, time, quality, risks), decided at the right level and, where necessary, approved by the donor. Most donors set a flexibility threshold between budget headings (for example 15% for several European donors, under conditions), beyond which an amendment is mandatory. Exceeding this threshold without agreement exposes the organisation to ineligible costs.

Major modifications (area, target group, duration) are requested before being applied, with a data-based justification (monitoring, security, market), a revised budget and a revised logframe. A no-cost extension request must arrive at least one to two months before the end date, depending on the donor.

Closure is often rushed. It includes: checking commitments, the final narrative and financial report, inventory and disposal of assets according to donor rules, ending or transferring contracts, archiving (often 5 to 10 years depending on the donor), informing communities and implementing the exit strategy, and a lessons-learned workshop.