Startup Ecosystem Simulator

Startup Ecosystem Simulator

Choose a city baseline and adjust ecosystem levers. The model predicts a 3‑year average of annual investments and shows multipliers and non‑linear interplay (synergy).
Phase 1 — Inputs

Structural conditions

Leading tech firms (proxy: unicorn offices)
LowHigh
STEM + developer talent pool
LowHigh
This slider drives both STEM students and developers.
International attractiveness (rank-based, reversed)
LowHigh
Industry connections rating
LowHigh
Selected city index vs best predicted city
Phase 2 — Founder base

Founder base & pipeline

Accelerator participants
LowHigh
Scenario multiplier vs baseline
Baseline predicted investment
Baseline observed investment

Model fit (all cities)

Source: startmapheatmap.com
Phase 3 — Working modules

Startup support

Program managers (innovation staff)
LowHigh
Program success rate (%)
LowHigh
Implied accelerated startups raising >$100k (numerator)

Community

Meetups activity
LowHigh
Co-working spaces
LowHigh

Market & investment climate

Seed deals
LowHigh
Exits (sum)
LowHigh
Used together with seed deals as “investment climate”.
Startups reaching €1bn revenue (count)
LowHigh
Synergy uplift (interplay)

Cutting edge

Cutting edge (proxy)
LowHigh
Phase 4 — Outcome

Predicted investments

Scenario predicted investment (annual, 3y avg)
Scenario index vs best predicted city
Source: startmapheatmap.com

Cascade view (what drives the result)

Local decomposition vs the selected city baseline (largest contributions shown).
Source: startmapheatmap.com

Module multipliers (vs baseline)

Source: startmapheatmap.com

How calculations work

1) Slider → value: counts use a log mapping anchored on benchmark percentiles; ratings use linear mapping.
2) Program success rate: derived as (accelerated startups raising >$100k) / (accelerator participants).
3) Features: computed mostly as ln(1+x), then standardized to z-scores.
4) Prediction: log(1+Investment) = intercept + Σ(coef·z) + capped synergy uplift; Investment = exp(·)−1.
5) Recommendations: “Recommendation” buttons apply the highest local gain (+10 slider points) within the section.