Senegal open data: a gap to close in 2026
Senegal launched its national open data portal data.gouv.sn in 2019, but it remained largely underused (380 datasets by end 2025, 60% obsolete or undocumented). The "Senegal 2050" strategy relaunched the effort with a target: 2,000 national datasets + 50 municipal open data portals by end 2027.
On the commune side: only 4 have a functional open data portal in June 2026 (Dakar Plateau, Pikine, Mbour, Saint-Louis). Huge field to conquer.
Why does municipal open data matter? Three reasons:
- Transparency and accountability — every CFA spent is traceable
- Economic innovation — startups, journalists, researchers create value from data
- Public efficiency — staff themselves use their own data (frequent paradox: a city hall doesn't know how many streetlights it has)
ROI measured in equivalent cities (Nairobi, Cotonou, Tunis): 2-5 FCFA generated in economic value per 1 FCFA invested in the portal. 2024 World Bank study on 12 African cities.
H2: Standards and formats to respect
CKAN (Comprehensive Knowledge Archive Network). Most used open source software in the world for open data portals. Used by data.gouv.fr, data.gov (USA), data.europa.eu, data.gouv.sn. Python + Postgres + Solr. Installation cost: 4.5-9.5 M FCFA. Annual hosting: 600 KFCFA-1.8 M FCFA.
dataset.json schema. W3C DCAT-AP standard to describe each dataset (title, description, license, format, update frequency, contact). Essential for interoperability.
Accepted formats.
- CSV (mandatory) — universal, readable in Excel, Python, R
- JSON (recommended) — for APIs and mobile apps
- GeoJSON (for geographic data) — roads, parcels, facilities
- PDF (only as complement, never as primary format — a PDF is NOT open data)
ODbL license (Open Database License). Global open data standard, authorizes commercial reuse with attribution. Adopted by data.gouv.fr, OpenStreetMap. Alternative: CC-BY 4.0.
Open REST API. Each dataset must have a JSON API for querying: https://data.commune.sn/api/3/action/datastore_search?resource_id=... (CKAN standard). Allows developers to build apps in real time on data.
H2: The 20 essential municipal datasets
Governance
- Municipal council minutes — PDF + metadata (date, agenda, attendees, votes)
- Deliberations — CSV with date, subject, vote (for/against/abstention per elected)
- Primary budget and administrative account — CSV line by line (cf article 3)
- Public procurement — CSV (subject, amount, awardee, procurement mode)
Territory and urban planning
- Local urban plan (PLU) — GeoJSON of zones (UA, UB, UC, UH, N, A)
- Cadastral parcels — GeoJSON (Senegal cadastre reference)
- Public facilities — GeoJSON (schools, health posts, markets, mosques, churches, stadiums)
- Roads — GeoJSON with hierarchy (primary / secondary / tertiary), surface, condition
Public services
- Public lighting — GeoJSON of light points + status (functional / defective)
- Public water points — GeoJSON standpipes, SDE / Sen'Eau status
- Waste collection — GeoJSON gathering points + SONAGED passage hours
- Stormwater sanitation — GeoJSON canals, basins, flood-prone areas
Local economy
- Building permits issued — CSV (date, parcel, surface, type, applicant)
- Business licenses and patents — CSV (activity category, zone, amount — anonymized)
- TOM (Household Waste Tax) collected — CSV by neighborhood
Demographics and social
- Population by neighborhood — CSV (ANSD census)
- Social aid beneficiaries — CSV aggregated by neighborhood (never nominative)
- School enrollments — CSV by establishment, level, sex
Environment
- Air quality (if sensors installed) — real-time JSON (PM2.5, NO2)
- Green spaces and trees — GeoJSON parks, tree alignments, One Million Trees project
Need a professional website?
Kolonell builds websites that attract clients, optimized for the Sénégalese market. Free quote in 2 minutes.
H2: Dataset publication workflow
Step 1 — Source identification. Which internal database contains the data? Often: a staff Excel spreadsheet, more rarely a structured base.
Step 2 — Cleaning and standardization. Duplicate removal, code harmonization (commune in ANSD postal code), anonymization if necessary. Tools: OpenRefine, Python pandas, dbt.
Step 3 — Documentation. Clear description, column dictionary, license, contact, update frequency.
Step 4 — CKAN publication. Upload + metadata filling + automatic API activation.
Step 5 — Promotion. Announcement on city hall networks, newsletter to registered developers / researchers / journalists, quarterly datalab / hackathon.
Step 6 — Maintenance. Update according to promised frequency (monthly, quarterly, annual). Response to reuser questions within 7 days.
H2: Investments
| Item | Cost |
|---|---|
| CKAN installation + commune theme | 4,500,000 to 9,500,000 FCFA |
| ETL pipeline for 20 datasets | 6,500,000 to 14,000,000 FCFA |
| Initial cleaning and documentation | 3,500,000 to 7,500,000 FCFA |
| 4-staff training (data steward) | 850,000 FCFA |
| Annual hosting | 600,000 to 1,800,000 FCFA |
| Maintenance and new datasets / year | 4,800,000 to 9,600,000 FCFA |
| Annual hackathon / datalab | 3,500,000 to 8,500,000 FCFA |
Initial total: 16.5-39.8 M FCFA. Annual recurring: 8.9-19.9 M FCFA. Civic ROI: 18 months to 3 years by usage.
H2: Use cases emerging from data
- Mobile apps: "Mon Quartier Pikine" (water outage alerts, waste collection, streetlight status)
- Data journalism: Le Quotidien, Senego, Walf: budget vs realization analysis
- University research: UCAD, ESMT, IFAN: urban planning, demographic, local economy studies
- Civic tech startups: KaayMa, Mbal'It: open data-based services (cf previous batch article on civic tech)
- Real estate investors: analysis of available parcels and PLU to identify opportunities
- Citizens and associations: tracking electoral promises, comparison of neighboring communes
FAQ
Doesn't open data create a security risk?
Risk mainly comes from nominative data (identity, precise address). As long as we aggregate (by neighborhood, age range) and anonymize, risk is nil. Transparency on public procurement, budget, PLU exposes nothing — on the contrary, it protects (a public public market is less falsifiable than a secret market).
How to publish without demotivating staff?
Gradual approach: start with 5-7 most consensual datasets (public facilities, roads, council calendar). Show value quickly (positive press, startup uses). Gradually extend to sensitive datasets (budget, markets). Reward staff who publish well (incentive, training, mobility).
Without dedicated IT, is it manageable?
Yes for a commune of 50,000+ inhabitants with trained IT staff and communication agent. Below: go through a provider that hosts and maintains (mutualization possible between 5-10 neighboring communes, "Données du Saloum" GIP model launched in 2025).
What precise economic ROI?
2024 World Bank study Cotonou (1.5 M inhabitants): open data portal costs 32 M FCFA / year, generates 86 M FCFA in economic value (internal city hall savings + data startup activity + journalism + research). Ratio 2.7. Beyond purely economic, democratic gain hard to quantify but real.
What if a neighboring commune refuses to publish its data?
Pressure comes from associations (Forum Civil, OSIWA), media (Le Quotidien, RFI, Jeune Afrique), donors (EU, World Bank condition part of funding on transparency) and citizens via elections. Dynamic is clear: opaque communes lose attractiveness (investments, projects, talents). The 2027 standard will be: no open data portal = donor doesn't invest.
Let's discuss your project
If you are mayor, CIO, NGO supporting data opening in Senegal, we can scope the open data portal and strategy. WhatsApp +221 77 596 93 33.
Mohamed Bah
Fondateur, Kolonell
Passionate about digital and entrepreneurship in Africa, Mohamed has been helping Sénégalese businesses with their digital transformation since 2020. Founder of Kolonell, he believes every SME deserves a professional and accessible online présence.
