Ecosystem building · Research methods

Ecosystem Mapping Handbook: A Practitioner's Guide

A step-by-step handbook for entrepreneurial ecosystem mapping in emerging markets. Use it to define scope, identify actors, collect data, visualize relationships, and translate the finished map into ecosystem development, policy, funding, and program decisions.

1. Define scope and purpose

A map is a decision tool, not a wall poster. Before collecting a single data point, write down the decision it must inform: where a fund will deploy the next $5M, which city the government will pilot a regulatory sandbox in, which sector a support organization will build a new accelerator around. That decision fixes the geography (city, region, country), the time horizon, and the level of resolution.

For our own country profiles in Jordan, Ukraine, and Sierra Leone, we scope at the country level with sector cuts for agri-food, digital services, and light manufacturing, because those are where most cohort founders operate.

2. Choose a framework

Two frameworks dominate the literature. Daniel Isenberg's six-domain model (policy, finance, human capital, markets, support, culture) is the most cited and the easiest to explain to policymakers. Erik Stam's framework separates framework conditions (formal institutions, culture, physical infrastructure, demand) from systemic conditions (networks, leadership, finance, talent, knowledge, support services), which is more useful when you want to isolate what governments can and cannot change directly.

We treat infrastructure as its own component, because connectivity, power, and logistics are binding constraints in most emerging markets. See our full entrepreneurial ecosystem components guide for the definitions we use in every map.

3. Identify the actors

For each pillar, list the actors that operate in your scope. A useful first pass:

• Founders and firms segmented by stage, sector, and founder demographics.
• Funders including angels, VCs, DFIs, banks, grantmakers, and blended-finance vehicles.
• Support organizations such as incubators, accelerators, hubs, coworking spaces, and BDS providers.
• Universities and training providers feeding the talent pipeline.
• Government agencies shaping regulation, taxation, and public procurement.
• Connectors, the individuals who broker introductions across the categories above.

Cross-check the list against our support organization directory. If a category has fewer than five actors, note it as thin: that is a finding, not a data gap.

4. Collect the data

Layer three sources so no single dataset biases the map:

1. Public datasets: World Bank Enterprise Surveys, Global Entrepreneurship Monitor, ILO labor statistics, national business registries, and Crunchbase or PitchBook for funded firms.
2. Program data: Portfolios from accelerators, DFI factsheets, and grantmaker annual reports. These are usually available on request even when not published.
3. Interviews: 20 to 40 semi-structured interviews with founders, funders, and support-organization leaders. Interviews capture relationships and trust, which registries never do.

Record every actor as a row with fixed columns: name, pillar, sector, stage focus, headquarters, founded, active status, and 2 or 3 links to primary sources.

5. Model relationships as a directed network

Actors alone are not an ecosystem. What matters is what flows between them. Model four canonical edge types:

• Capital flows: investments, grants, loans, and revenue contracts.
• Talent flows: alumni moving from firms, universities, or programs into new roles.
• Mentorship and advisory: board seats, advisory relationships, and program cohorts.
• Information flows: co-publications, joint events, shared research.

Each edge should carry direction, weight (dollars, headcount, count of relationships), and a source citation. This is the substrate that makes downstream analysis possible.

6. Visualize the map

Two representations work in tandem. A pillar grid (a simple 3x3 matrix of pillars by stage) communicates coverage at a glance and is what policymakers respond to. A network diagram reveals structural properties: brokers, isolates, dense clusters, and missing bridges. Kumu is the easiest to publish interactively, Gephi is best for quantitative network analysis.

Publish the underlying data as CSV alongside the visualization. Maps that stay in slide decks do not get updated; maps that ship as open data get corrected and extended by the community.

7. Diagnose and act

The final step is where most mapping projects fail. A map that ends with "here is our ecosystem" is a poster. A map that ends with "here are the three interventions the data supports" is a strategy. Run three diagnostics against your finished map:

• Thin components: which component has fewer than five active actors, or no actors at seed stage?
• Broken links: where does capital exist but not reach founders? Where does talent graduate but not stay?
• Missing brokers: which actors would connect two dense but disjoint clusters if they existed?

Each diagnostic implies an intervention. Publish the diagnoses alongside the map so funders, policymakers, and support organizations can each pick up the work that fits their mandate.