Creators express their location in dozens of formats: city names, country names, flag emojis, abbreviations, foreign scripts, and more.
Signals are drawn from multiple sources across each profile and weighed by reliability.
The result is a single, clean Location with country, region, and city.
Coverage
Share of profiles for which we have a resolved structured location.
Coverage by Platform
The Priority Cascade
Sources are checked in a specific order, chosen for the best balance of accuracy and execution time. Each source returns its own confidence weight; the final output combines whatever signals are found.
Sources in Priority Order
When a platform exposes structured location data, we treat it as the primary country signal. State and city are then enriched from other sources. This is the highest-confidence path.
If the platform exposes a structured location field, we resolve it to a clean, structured place.
Free-text addresses are parsed to extract place names, then geocoded. Noise words are stripped and multiple substring combinations are tried until a valid place resolves.
The bio is the richest fallback. We scan it for six different kinds of signals, each with its own confidence weight (see breakdown below).
If the bio is written in a non-English language (e.g. Cyrillic, Chinese), US-only matches get demoted automatically to avoid false positives where, for example, "Georgia" could refer to either the US state or the country.
When a contact number is publicly listed, the country code provides a location signal. For US and Canada numbers, the area code also gives us a state.
A country-code TLD on the email domain (.uk, .de, .br) maps to a country.
How Confidence Adds Up
Every source produces a result at its own confidence tier. Multiple findings are combined: signals that agree increase confidence, signals that conflict reduce it. Final confidence has an upper limit so any single noisy signal can't dominate the result.
Source Confidence at a Glance
The pipeline blends the confidence of every source that found a location. Each additional source that agrees raises confidence. Each source that disagrees lowers it. The combined result has an upper limit, so even multiple agreeing signals can't claim absolute certainty. This stops any single noisy signal from dominating the final answer.
Location Resolution for Two Illustrative Profiles
Bio: "Lisbon-based traveller exploring Europe on a budget"
Post content: Mentions Lisbon landmarks, local cafes, Bairro Alto neighborhood
Phone prefix: +351 (Portugal)
Bio: "Documenting life one vlog at a time"
Post content: Mentions Berlin, Amsterdam, Prague across different vlogs
Phone prefix: Not available
When bio, content, and phone prefix all point to the same place, the algorithm resolves location down to city level with high confidence. When signals are mixed or sparse, the pipeline may only resolve to country level. The output always includes a confidence indicator so customers know how precise each location is.
Cross-Platform Signal
When we match a creator's profiles across platforms, location data from one platform can fill gaps on another. If a creator lists their city on one platform but not another, the matched profile carries that signal over.
How a Location Fills Across Platforms
Previously empty on Instagram, now filled from matched YouTube profile
Our profile matching system links accounts that belong to the same creator. When a location is available on one platform but missing on another, the matched value fills the gap. The most specific location (city over country) takes priority.
Validation 
Every accuracy number on this page comes out of a two-track validation process. We compare at least two independent AI sources on tens of thousands of records, and we manually validate at least 1,000 records per subset, with subsets defined to cover every meaningful use case. Location is fully re-validated after each major revision.
