Managing a site day to day

1Planning space according to Sphere

Sphere benchmarks for a camp-type settlement:

• At least 45 m² per person, including household plots, roads, services, communal space and firebreaks (excluding agricultural land); 30 m² if some communal services are provided outside the site.
• 3.5 m² of covered space per person in the shelter, more in cold climates or urban settings (4.5 to 5.5 m²).
• Firebreaks of 30 m every 300 m of built-up area, and at least 2 m between shelters (ideally twice the height of the shelter).
• Gently sloping land for drainage, outside flood-prone areas, at a distance from front lines and borders.

Planning is done in community modules: shelters grouped in blocks around nearby services (water points, sex-segregated latrines, lighting), to preserve social ties and safety. Space is kept for extensions and new arrivals.

2Profiling and registering the population

You can only manage well what you know. There are three levels of data:

• Rapid estimation and counting on arrival (shelter count × average household size, aerial survey, satellite imagery).
• Profiling: number of households, breakdown by sex and age, persons with specific needs (older people alone, persons with disabilities, unaccompanied children, female heads of household), places of origin and return intentions.
• Individual registration by household, carried out by the authorities or by UNHCR in refugee contexts, which determines access to distributions and documents.

The IOM DTM (Displacement Tracking Matrix) produces, site by site, data on population, needs and movements, shared with all clusters. Personal data remain protected: minimal collection, consent, secure storage, aggregated sharing only.

3Monitoring services and identifying gaps

Site monitoring regularly compares, often monthly, the actual situation with the standards: litres of water per person per day (at least 15 l according to Sphere), number of people per latrine (20 maximum), distance to the water point (500 m maximum), lighting, access to health and education, condition of shelters.

Results are presented in a dashboard with colour codes by sector and by block. Each gap is passed on to the service provider concerned and to the cluster, with a follow-up date. The principle: observe, report, follow up, escalate if nothing changes.

4Coordinating actors: meetings and the 4W matrix

The site coordination meeting, chaired by the site manager, brings together service providers, residents' representatives and local authorities every week or every month. It tracks gaps, allocates tasks and checks on commitments made.

The 4W matrix (Who does What, Where, When) lists interventions by block and by sector. It avoids duplication (two distributions to the same block) and reveals neglected areas. It must be kept up to date and shared with the CCCM cluster, which consolidates it for all sites.

5Detailed planning and upgrading an existing site

The standard of 45 m² per person is only a starting point. A professional site plan divides this surface area between different uses, organises services by level and retains the flexibility needed to absorb new arrivals. In practice, most site managers do not open a planned site: they inherit an overcrowded spontaneous site that has to be upgraded without being torn down.

Allocating the surface area: an order of magnitude

Organising services by level

UNHCR's modular approach (family, community of about 80 people, block of about 1,250 people, sector of about 5,000 people, camp module of about 20,000 people) makes it possible to assign each service to the right level. UNHCR recommends not exceeding about 20,000 people per camp: beyond that, a second site is created or the camp is subdivided into self-contained modules.

Sphere benchmarks to build into the plan

• Water: minimum 15 l/person/day; water point less than 500 m away; queuing time under 30 minutes; 250 people per tap, 500 per hand pump, 400 per open hand-dug well with a single user at a time.
• Sanitation: latrines less than 50 m from shelters; 1 per 50 people in the acute phase, 1 per 20 in the medium term; female-to-male ratio of 3 to 1 for public latrines; internal locks, lighting, access for persons with disabilities.
• Solid waste: access to a container or collection point less than 100 m away; benchmark of one 100-litre container per 10 households.
• Shelter: 3.5 m² of covered space per person in warm climates, 4.5 to 5.5 m² in cold climates or urban settings where cooking and living take place indoors.

Upgrading a spontaneous site without tearing it down

1. Map what exists: satellite or drone imagery, survey of shelters, paths, latrines and low points; each shelter is assigned a code.
2. Identify safety emergencies: shelters in flood-prone areas, densities that would make a fire uncontrollable, lack of access for emergency services.
3. Design a target plan with the Shelter Cluster: layout of firebreaks and access roads, locations of services, decongestion area.
4. Negotiate every relocation: the households concerned are informed individually, choose their new plot where possible, receive support (materials, transport, labour) and are only moved once the receiving plot is ready.
5. Proceed in phases, block by block, avoiding the rainy season, and starting with voluntary relocations.
6. Lock in the gains: marking of firebreaks, site rules validated by the committees, monitoring of re-occupation.

A firebreak that is not maintained fills up with shelters within a few months. The site plan is not a document: it is a collective rule that the committees and the maintenance team enforce every week.

6Population data: counting, profiling, registering, tracking movements

Every decision on a site rests on a population figure: water volumes, number of latrines, rations, school places. An overestimated figure wastes resources and fuels fraud; an underestimated one creates invisible gaps. The site manager needs to know which method produces which figure, with what margin of error, and what it is used for.

Choosing the right method

Conducting a site headcount: the method

1. Preparation: agreement of the authorities, residents informed the day before through the committees (date, purpose, no direct link with a distribution so as to limit fraud), site map divided into zones, teams of two (one man, one woman), half-day training.
2. Cordoning: the count takes place early in the morning, in all zones simultaneously, to avoid double counting between zones.
3. Data collection: for each occupied shelter, number of members present by sex and age group (0-4, 5-17, 18-59, 60 and over), persons with specific needs by category, marking of the shelter or handing over of a token.
4. Quality control: a supervisor checks 5 to 10% of shelters; discrepancies are explained.
5. Analysis and feedback: aggregated figures by block shared with the cluster and service providers; no individual lists circulated.

The site profile: what it should contain

• Population and households by block, breakdown by sex and age.
• Persons with specific needs: persons with disabilities (the Washington Group questions are the reference for identification), isolated older persons, female-headed and child-headed households, unaccompanied or separated children (reported immediately and confidentially to child protection actors), people with chronic illnesses, pregnant and lactating women.
• Areas and communities of origin, date of arrival, languages spoken.
• Short- and medium-term movement intentions.
• Skills and livelihoods, to prepare paid activities and durable solutions.

Protecting data

The IASC Operational Guidance on Data Responsibility in Humanitarian Action, the data protection policies of UNHCR and IOM, and the ICRC Handbook on Data Protection in Humanitarian Action converge on a few rules:

• Minimisation: only collect what serves an identified decision.
• Purpose and informed consent: the person knows why data is being collected and what will happen to it, and can refuse without losing access to life-saving assistance.
• Security: encrypted devices, passwords, role-based access, no paper lists left lying around or sent via personal messaging apps.
• Sharing: aggregated data by default; individual data only under a data sharing agreement that specifies the purpose, the security measures and the retention period.
• Sensitive data: ethnicity, religion, political affiliation, survivors of violence, children: never in a shared spreadsheet. Cross-tabulating small categories by block can be enough to identify a person ("the only female head of household with a disability in block 4"): small cells are grouped or suppressed.

Tracking arrivals and departures

A site is alive: births, deaths, arrivals, departures, pendular movements (people who go to farm their fields during the day and come back). The site manager keeps a movement register at the reception desk, cross-checks this data with that of the block committees and updates the figures every month. A full verification (recount or registration verification) is useful every six to twelve months, or as soon as discrepancies between sources exceed about 10%.

When two sources give different figures (authorities, DTM, food distribution, the site manager's headcount), do not pick the figure that suits you: document the methods, explain the discrepancies to the cluster, and propose a joint verification exercise.

7Service monitoring: indicators, thresholds and escalation protocol

Service monitoring is the technical core of the site manager's job. It turns scattered observations into comparable evidence, makes it possible to prioritise gaps and gives the cluster the material for its advocacy. It must meet three requirements: indicators common to all sites (defined by the cluster), a regular frequency, and a response pathway that leads to action.

The indicator set for monthly monitoring

The amber and red thresholds above are examples: in each country, the cluster defines them together with the sectoral clusters. What matters is that all site managers use the same ones.

How to collect

• Direct observation by the team, with one form per block: count functioning taps and latrines yourself, do not settle for reported figures.
• Service provider data, requested before a fixed date every month.
• Rapid survey of a sample of households, quarterly, on perceptions of services and safety.
• Focus group discussions separated by sex and age to understand why a service is not being used (latrines avoided at night, unsuitable distribution times).

The escalation protocol

An alert is only valuable if it is precise: "block D has 28.6 people per functioning latrine, 3,000 residents, 45 latrines missing, reported on the 3rd, no response as of the 17th" is usable. "The latrines are a problem" is not.

8Worked example: the Bourou-Est site dashboard

Fictional context: the Bourou-Est site, opened 14 months ago in the Haut-Sirané region, hosts internally displaced persons from an intercommunal conflict. Site management agency: the fictional national NGO "Entraide Sirané", supported by an international NGO. Surface area: 48 hectares. Six blocks, A to F. A single WASH partner, funded until the end of the quarter.

1. The month's raw data

Other data: volume of water delivered, 165,000 l/day according to the partner; 60% of the solar lamps around the latrines are working; 9% of shelters damaged; 186 pieces of feedback received during the month, including 41 on water and 23 on night-time safety.

2. The calculations

• Surface area per person: 480,000 m² ÷ 14,600 = 32.9 m², against the expected 45 m²: the site is dense; increased risk of fire and epidemics, expansion or decongestion to be considered.
• Water: 165,000 ÷ 14,600 = 11.3 l/person/day. To reach 15 l: 14,600 × 15 = 219,000 l/day, i.e. a shortfall of about 54,000 l/day.
• Latrines: 14,600 ÷ 520 = 28.1 people per latrine on average; requirement at 20: 730 latrines, i.e. 210 missing.
• Taps: 14,600 ÷ 38 = 384 people per tap; at 250 per tap, about 59 taps are needed, i.e. at least 21 more (23 if calculated block by block, which is the correct method since residents do not use taps in a distant block).

3. The dashboard by block

Interpretation: the site average masks two critical blocks, B and F, which together account for 120 of the 210 missing latrines. Block D, the most populated, is red for water. 70% of the 23 complaints about night-time safety come from blocks B and F: the lack of latrines, faulty lighting and distance force women to walk further at night. Cross-analysis turns a WASH problem into a protection problem.

4. The three alerts presented at the coordination meeting

5. Follow-up the next month

One month later, the WASH partner has desludged and brought back into service 60 full latrines and built 18 new cubicles in blocks B and F (30 more latrines in block B, 48 in block F); 70% of the lamps are working. Block F moves to 2,300 ÷ 103 = 22.3 people per latrine (amber). Block B reaches 2,600 ÷ 100 = 26 (amber). Water remains at 11.8 l: alert 2 is escalated to the WASH Cluster with the history. The dashboard thus shows not only the state of the site but also the speed at which gaps are resolved, an indicator donors increasingly look at.

9Coordinating and maintaining the site: meetings, 4W, referrals and maintenance

A site coordination meeting that produces decisions

A useful site meeting lasts 90 minutes at most, follows an agenda sent in advance and ends with a list of time-bound commitments. Standard agenda:

1. Follow-up of commitments from the previous meeting (action tracker, line by line).
2. Population trends (arrivals, departures, movement alerts).
3. Service dashboard: red and amber gaps, by block.
4. Residents' feedback: trends of the month, presented by the committees and the accountability officer.
5. Risks and preparedness (season, epidemics, tensions).
6. Updates from service providers (new activities, project end dates).
7. Decisions and commitments, read out loud before closing.

Residents' representatives (including women) take part, at least for part of the meeting. Individual protection cases are never discussed in plenary.

The action tracker

The site 4W matrix

It is maintained by the information management officer, updated every month, and submitted to the cluster in the cluster's format. Minimum columns: organisation (and implementing partner), sector, activity, block or zone, number of people targeted and reached, start and end date, status, donor, contact. The end date column is the most useful for anticipation: it reveals upcoming service disruptions.

Referral pathways

Site management staff encounter situations every day that fall within the remit of other specialists: gender-based violence, a separated child, a sick person, a threat against a family. They do not manage these cases: they refer them, safely and confidentially, with the person's consent (except for the exceptions provided for children at risk). To this end, the site manager keeps an up-to-date written referral pathway with the clusters and areas of responsibility: available services, opening hours, contacts, criteria, and trains staff in the first approach (listen, inform, refer, do not investigate).

Site maintenance and paid activities

A site deteriorates quickly: blocked drainage, re-occupied firebreaks, gullied roads. The maintenance plan lists the communal infrastructure, its maintenance frequency, the person responsible and the budget. Part of the work is carried out by residents through paid activities (cash for work). Rules to follow:

• Pay rates harmonised within the cluster and with other sectors, to avoid competition between organisations and distortion of the local labour market.
• Published selection criteria, rotation of participants, access to suitable tasks for women and persons with disabilities.
• No child labour, protective equipment, insurance or coverage of accidents.
• Supporting documents: signed attendance lists, payrolls, acceptance of works.

The 4W reveals who is present; the action tracker reveals who keeps their word. Together, they allow the site manager to move from observation to holding service providers accountable.

10Variations, common mistakes and the donor's perspective

Adapting day-to-day management to the context

Common mistakes

• Site averages that hide critical blocks: always present data by block.
• Unverified reported figures: the number of latrines "built" is not the number of functioning latrines.
• Monitoring without a response pathway: a dashboard that feeds neither the site meeting nor the cluster is useless.
• Site meetings that are too long and produce no decisions, where each organisation reads out its activity report.
• Personal data shared via messaging apps or in unprotected spreadsheets.
• Frozen population figures: using the figure from the first headcount for a whole year.

What donors and evaluators look at

• The regularity and coverage of monitoring (number of cycles completed, sites covered) and its consistency with the cluster's tools.
• The resolution rate of reported gaps and the average time to resolution.
• Data quality: documented methods, disaggregation by sex, age and disability, data protection measures.
• Evidence of use: decisions taken on the basis of dashboards, evidence-based advocacy.
• For paid activities and small works: complete supporting documents and harmonised rates.

11Interview questions and key references

Interview questions

1. "A site of 12,000 people has 300 functioning latrines. Analyse."

A good answer immediately calculates 40 people per latrine, compares this with the acute-phase (50) and medium-term (20) benchmarks, asks for the breakdown by block and by sex, checks locks and lighting, and links the issue to protection and health (diarrhoea). It proposes a quantified alert to the WASH partner with a deadline.

2. "The DTM, the authorities and your own headcount give three different figures. What do you do?"

It compares the methods and dates, explains the discrepancies (empty shelters, pendular movements, multiple registrations), refuses to choose arbitrarily, and proposes a joint verification exercise under the auspices of the cluster, while protecting personal data.

3. "How would you build a site dashboard?"

It cites indicators aligned with the cluster and Sphere, presentation by block with colour-coded thresholds, sources verified by observation, a monthly frequency, presentation at the site meeting and the link with an action tracker.

4. "A service provider has been ignoring your alerts for a month."

It describes the protocol: written, quantified reporting, an item in the meeting, a deadline, then escalation to the sectoral cluster and the CCCM cluster with the history, without publicly attacking the service provider, and seeking a continuity solution with the cluster.

5. "What data in a site profile must never be shared at individual level?"

It cites unaccompanied children, survivors of violence, ethnic or religious affiliation, medical information, and explains the risks of re-identification by cross-tabulating small categories.

Key references

• Camp Management Toolkit (NRC, 2015): chapters on information management, coordination and site maintenance.
• Minimum Standards for Camp Management (CCCM Cluster, 2021).
• The Sphere Handbook (2018), WASH and Shelter and Settlement chapters.
• UNHCR Emergency Handbook: site planning and related standards.
• Displacement Tracking Matrix (DTM) methodologies: IOM.
• Operational Guidance on Data Responsibility in Humanitarian Action: IASC.
• Handbook on Data Protection in Humanitarian Action: ICRC.
• Washington Group questions on disability; IASC Guidelines on the Inclusion of Persons with Disabilities in Humanitarian Action (2019).
• Site monitoring and 4W tools published by the Global CCCM Cluster and national clusters.