A base built to be replicated.
Entrepreneurship research is starved of consistent, longitudinal data on support structures. OSE provides it — with a public API, exports, a research tier and replications documented against the original papers, discrepancies included.
Surveys are not censuses, and snapshots are not panels.
Most ecosystem datasets are self-reported, single-year, or limited to a handful of countries. Reviewers ask for citable methodology and replicable measures; the field rarely has either at continental scale.
- No cross-country, longitudinal data on support organizations at the level of services delivered.
- Measures published once, never recomputed on fresh data.
- Network-level constructs (communities, complementarities) without the underlying edge lists.
- Institutional variables (WGI, gender statistics, AEEI) kept in separate silos.
Measured on the live base — never declared.
The base
60,600+ qualified organizations, 150+ countries, 14 indicators, 5 layers, 14 canonical boosts (+14 sectoral v2), 10+ years — re-qualified continuously.
Meta-organizational measures
Coverage Index, Functional Differentiation (Shannon), Membership Turnover, network density, VRIO coverage, Service Redundancy Index, Ecosystem Effectiveness Score — all computed live and exportable.
COTEE replication set
Typology (43 apparatus-builders / 68 endowment-based; paper 58/55; bootstrap ARI 0.631), cultivation interaction (β = 0.23, p 0.076; paper 0.396), necessity analysis, within-country variance, press panel H5.
Cross-lenses
Aligned Gender Index (World Bank Gender Statistics) and AEEI (20 indicators, 54 countries) joined to the OSE base by country.
Three results that change the decision.
Published or replicated on the OSE database — the same numbers you will find in OSE Expert.
Replicated, with the discrepancies published.
COTEE replicated on the OSE base with WGI 2023 and World Bank outcomes: typology 43/68 vs. paper 58/55 (bootstrap ARI 0.631); interaction β = 0.23 (p 0.076) vs. 0.396; within-country variance 68.4% vs. 72%.
A press panel of 154 country-years to test interface visibility.
H5: 19 countries, 2014–2023, 154 country-years. Visibility of interfaces +0.12 (n.s.) under weak enforcement, −0.46 (significant) under strong; interaction β −0.341 (t −3.82).
Every published measure is recomputed live and exportable.
Guéneau et al. (2025) three-scale framework and Guéneau et al. (2026) meta-organization measures are computed on the current base — not frozen at publication. Exports and the public API let you re-run them.
Sources: COTEE replication on the OSE database (WGI 2023, World Bank outcomes); Guéneau et al. (2026), Journal of Business Venturing Insights; Aligned Gender Index (World Bank Gender Statistics); AEEI (africa.ecosystem.build). Measures are computed live in OSE Expert.
What you get.
Research tier on OSE Expert
Full country diagnostics, benchmark of up to 6 countries, network views, CSV and PNG exports for every published measure.
Public API
api.ose.africa — public statistics, country map, rankings, layer coverage, AEEI; documented endpoints for reproducible pipelines.
Data partnership
Embedded access for centres focused on entrepreneurship, innovation policy and ecosystems; co-authored replications welcome.
Three steps, no commitment before the first.
- 1
Scope the question
Tell us the construct and the countries; we check coverage on the live base in 30 minutes.
- 2
Access
Research tier or data partnership; API keys and exports for your pipeline.
- 3
Replicate and publish
Documented methodology, scientific-council contact, co-authorship where it makes sense.
Other audiences.
- Governments & Ministries
- Donors & DFIs
- Gulf Sovereign (GCC)
- Foundations
- Embassies & Bilateral
- Investors & VCs
- Founders & Entrepreneurs
Galaxy OSE observes; ADALIA trains; EERS advances the science. See how they combine →
Run the measures yourself.
Thirty minutes to check coverage for your research question on the live base.