2022 | StartupBlink View the Global Startup Map p. 11 Methodology This section explains our methodology for producing the Global Startup Ecosystem Index. We start with guiding principles, provide relevant definitions, and then discuss the elements comprising each section of the total score. Guiding Principles To ensure that the rankings are as accurate as possible, we have based our algorithm on objective, quantifiable data that can be comparatively measured across regions, countries, and cities. We refrained from using subjective tools such as surveys and interviews, and instead utilized data that was either accumulated directly from the StartupBlink map or has arrived from integration with a reliable global data partner. We allow as few assumptions as possible regarding cause and effect and focus on one thing: measuring results. We avoid relying on any theoretical models assuming the causes of success for startup ecosystems. Our experience in ecosystem consulting shows that no two ecosystems are alike; policies and practices that are successful in one ecosystem can be disastrous in another. It should be noted that the index does not measure urban innovation or implementation of advanced policies related to city development. It focuses instead on the output of entrepreneurial innovation developed in each location. Most annual rankings face a trade-off between maintaining the consistency of the algorithm and innovating on new elements to improve or adapt the algorithm to changing business environments. We have always leaned toward constantly improving our algorithm so stakeholders can rely on our results to make informed decisions across the board. An algorithm cannot remain unaltered forever; since reality continually changes, so do startup ecosystems. Every year our algorithm is more accurate, and it should be noted that the momentum change of each ecosystem is not only influenced by its achievements over the last year, but also by these algorithm improvements. We have been sampling startup ecosystem data on the curated and interactive StartupBlink Global Map, which enables us to test and perfect our algorithm based on vast sets of data. We estimate that the core map dataset has a representative sample covering 10-15% of total relevant entities in global startup ecosystems. In addition, hundreds of thousands of entities and data integrations are taken into account via our global data partners. Each location’s final score is based on the exact same algorithm. However, we are aware that our sample size fluctuates depending on location and data sourcing. Our only intervention in the score is discounting locations where we determined that the sample size of the entities is higher than average. In order to solve issues with lower than average sample size, we have partnered with approximately 100 Ecosystem Partners, most of which are Government agencies, many of them in locations where our data is limited. We offer all governments administrative access to curate the dataset of their ecosystems, at no cost, granting them complimentary access. For more information, please contact us at Feedback@StartupBlink.com.