Objective.
Provide an independent, reproducible ranking of the world’s leading EdTech and online learning companies using only publicly available digital signals and open sources. No paid submissions, private surveys, or non-public datasets were used.

Company Universe Construction.

• Initial pool (>2,500 companies): compiled from industry research and public directories, including EdTechIndex, HolonIQ, and broad visibility across learning ecosystems.
• Coverage expansion: additional organizations identified via sector overviews, media coverage, ecosystem partner listings, app store discovery, search presence, and LinkedIn.
• Data hygiene: normalization of company names and domains, consolidation of global vs. local entities, removal of duplicates, and exclusion of entities without a core EdTech/learning product.

Sources and references (public).
The universe and validation signals were built from public materials, including industry overviews, EdTechIndex.org, HolonIQ.com, and LinkedIn.com, alongside other openly available web signals.

Important note on global platforms.
The EdTech market includes many global platforms that operate local brands, country subdomains, reseller networks, or regional subsidiaries. In some cases, a single global website and shared social profiles can inflate signals for local units. To reduce distortions, duplicates were removed and digital weights were attributed to the primary global entity. Local entities are included independently only when they demonstrate a clearly separate digital identity.

Key Index Factors (public signals).

• Website traffic based on public estimations of monthly visits (e.g., Similarweb).
• App footprint measured through public signals such as mobile app presence and estimated installs (where applicable).
• Media activity reflected by public news coverage and media visibility signals.
• Search visibility approximated via public indicators of brand mentions and search interest.
• Backlink profile measured through publicly observable link signals and referring-domain strength.

Statistical Treatment.
All variables are normalized to reduce skew from heavy-tailed distributions, standardized into comparable scales, and combined into a weighted composite index. The final score is scaled to 0–1000, where 1000 represents the strongest overall public digital footprint within the dataset.