Home / Methodology
Methodology & Trust

How tallywire collects its data

What the data is, where it comes from, how it is normalized, and the privacy line we hold. Data current through 13 July 2026; source observations are collected on a rolling basis.

Sources & coverage

tallywire aggregates publicly listed local-service providers and their published service menus from the major booking platforms. We enumerate the public listings each platform exposes for crawlers, fetch the profile pages, and parse structured data (JSON-LD where present) into one schema.

SourceProvider records · reach
Booksy
226,457
16 countries
Treatwell
119,958
11 countries
Fresha
72,515
89 countries
Planity
57,656
6 countries
Mindbody
2,774
56 countries

Six sources are covered today; the newest (Salonized) is rolling out and joins this table at the next refresh.

Source platforms & affiliation

tallywire observes publicly listed information across the platforms named on this site. It is independent: not affiliated with, endorsed by or partnered with any booking platform. Platform names appear solely to describe source coverage.

Which number means what

Every figure on this site belongs to exactly one of these universes. If two numbers differ, they are counting different things.

TermMeaning
Provider records
Deduplicated per-source records in the core dataset, all verticals, across covered platforms. The sitewide headline count.
Online-bookable venues
One vertical × one metro on covered booking platforms — the universe of the per-vertical market pages (e.g. London beauty).
Map-visible venues
Relevant public map listings inside the market boundary — the whole-market denominator on deep-report metros.
Deep-report metros
Metros with a full map reference pull: denominator, fragmentation, closure share, entity-joined platform attribution.
Monitored markets
Markets with preserved monthly comparison snapshots extending one consistent price and supply series.

Geographic normalization

The same country arrives under several labels (ISO-2, native and English names). We canonicalize them, GB, UK and United Kingdom collapse to one country, so national figures never double-count. City labels mix granularity, so we roll up district-level labels (all the “* London” variants) to their metro, giving one strong market page instead of six thin ones. Regions and départements are kept as honestly-labelled area pages until postcode-level metro resolution lands. Normalization is applied when the site is generated; the underlying records are untouched.

Normalization & dedupe

Every provider record maps into one LocalServiceProvider schema with a guaranteed geo bucket (postcode > city), a controlled vertical, and a list of Service rows (name, price, currency, price_type). Records are deduplicated by name and location within each source. Cross-platform entity joining (one venue across platforms and map listings) runs on the deep-report metros; sitewide provider-record counts stay per-source until that merge is validated everywhere. Prices are provider-published list prices; medians and percentiles exclude values ≤0 or >100,000 to keep parse errors out of the statistics.

Privacy, business data only

We retain business and service data only. No reviewer identities, no personal contact data, no consumer profiles. Sole-trader business names that are inherently public listing data are kept as the business identity, nothing more.

Cadence: rolling collection, preserved snapshots

Local-service prices are sticky, they barely move week to week. That is why the longitudinal series is the asset. Source observations are collected on a rolling basis across the covered platforms and markets; monitored markets receive a preserved monthly comparison snapshot, so the current view compounds into a price history a fast-follower cannot reconstruct retroactively, the AirDNA / Case-Shiller model applied to local services. Current pages show the latest completed collection (current through 13 July 2026); the tracked series accrues from here. We do not publish trend or history figures we have not yet measured.

The aggregate today

450,000+ provider records · ~17M priced service listings · 80+ countries, across the covered sources, as of 13 July 2026.

tallywire · the local-service supply feedMethodology
tallywire is not affiliated with or endorsed by the platforms shown. Platform names and marks belong to their respective owners; data is observed from public listings.