Skip to content
Sign in

Materials and methods

How the study behind the ranking was built. The study rests on three public datasets, collected and verified in November 2026: the 151 finalists of 2026, the champions since 2017, and 50 regional finals. Every figure is sourced in the study’s atlas or its list of sources.

151

Finalists of the 2026 Grand Finale

74

Countries represented

611

Public sources cited

50

Regional finals, 2023 to 2026

2.1

Study population

The 151 finalists of the 2026 Grand Finale

San Francisco, 4 to 6 November 2026. Built from the three official lists displayed on site: the Top 10, then the 141 other finalists. They come from 74 countries.

The champions from 2017 to 2026

From the competition’s official page, victory press releases and the press. There was no Grand Finale in 2020 and 2021.

50 regional finals from 2023 to 2026

20 of them publish a comparable number of applications.

2.2

Sources

Public sources only: startups’ official websites, press releases, the business and trade press, public pages of funding databases (Crunchbase, Tracxn, PitchBook, Dealroom), company registers, founders’ public profiles, and pages of the Startup World Cup and Pegasus Tech Ventures.

The study’s atlas cites 611 sources for the 151 profiles. The research was carried out by AI agents working in batches, each required to cite every source. Their output was then reviewed and corrected by hand.

2.3

Identification and cross-checking

Each name was searched on its own, then together with “Startup World Cup 2026” and the name of its regional final, to rule out namesakes.

A second, independent pass reviewed 150 profiles; 112 of them were corrected or completed.

Identification confidence

High
93
Medium
58
2.4

Variables measured

Variables measured, their definition and coverage among the 151 finalists
VariableDefinitionCoverage
Country and continentCountry of the operational headquarters151 startups
Sector15 harmonized sectors151
Year foundedLegal incorporation, otherwise the declared launch; as a last resort, the website’s first archive, flagged as an estimate134
Funds raisedEquity and grants in $M; debt, in-kind prizes and unclosed targets are excluded76 disclosed amounts, 55 with no funding
Funding stageLast known round: pre-seed, seed, Series A, Series B and later, self-funded102 determined
Product maturityR&D, prototype, pilot, product on the market151
TractionNone known, pilots or letters of intent, first customers, strong traction151
Use of AIAI at the core, AI integrated, no documented AI151
LLM providerOnly on public evidence34
Investor ratingFour criteria from 1 to 10 and an overall score151
Addressable market (TAM)2025 global market of the targeted segment, in $bn of revenue151
ScalabilityEase of replicating the model in other countries, from 1 to 10151
2.5

Calculation rules

Conversion to dollars at fixed rates

  • €1 = $1.08
  • £1 = $1.27
  • CHF 1 = $1.12
  • AUD 1 = $0.65
  • ¥100 = $0.67
  • ₩1,000 = $0.72
  • CNY 1 = $0.138

When sources diverge, the best-documented amount is used, with a range.

Comparisons

The Top 10 is compared with the 141 other finalists. Amounts and ages, which are widely dispersed, are compared by their median; scores by their mean.

Economic data

2025 GDP, population and GDP per capita from the World Bank (IMF for Taiwan).

Scale of the competition

The median number of applications to comparable regional finals, multiplied by the number of regional finals announced by the organizer, then cross-checked against the figure the organizer publishes. National contests, student registrations and figures shared between two regional finals are excluded.

2.6

Scoring grid

Each startup is scored as an investment bank analyst would, on four criteria from 1 to 10. Each score is justified in one sentence in the study’s atlas.

Scoring grid: criterion by score band
Criterion1–34–67–89–10
DisruptionIncremental improvementReal innovation in a known categoryNew approach that can change practicesPlatform or technology breakthrough
Business model strengthNo proven revenue or customersPilots or first paying customersRecurring revenue or solid contractsProven profitability at scale
Market potentialTAM below $1bnTAM of $1bn to $20bnTAM of $20bn to $100bnTAM above $100bn, scalable and international
Risk10 = very riskyRisks under controlUsual risksSignificant risksMajor risks

Risk combines technology, regulation, execution and team, cash relative to needs, competition, and data uncertainty. It is the only criterion where a higher score is worse.

2.6

Market score rule

Starting point: the addressable market (TAM)
TAMBase market score
Below $1bn3
$1bn to $5bn4
$5bn to $20bn5
$20bn to $100bn6
Above $100bn7

Then adjusted

  • +1if scalability reaches 8
  • −1if scalability is 4 or less
  • +1if the startup sells in at least two countries
2.6

Overall score and grades

Overall score

Risk is inverted (11 − Risk) so that, like the other three criteria, a higher value is better.

  • A7 and above
  • B6 to 6.9
  • C5 to 5.9
  • DBelow 5

Limit. These scores are an analyst’s judgment on an explicit grid. They do not reproduce the jury’s assessment.

2.7

Fairness adjustments to the addressable market

  1. 1When a startup covers several segments, their markets are added together.
  2. 2When the published figure measures a volume rather than revenue (invoices financed, sales, outstanding loans), it is converted into revenue.
  3. 3Very broad segments are flagged as upper bounds.

Source: section 2 of the open study Startup World Cup 2026 by TIBOK. Scores apply to the 151 SWC 2026 finalists only.

Materials and methods · Champions Network