Launching PlatformTrack: a weekly-synced streaming availability index for 17 platforms x 57 countries

Streaming availability is a deceptively simple question. Does Netflix carry The Bear in Turkey? How many movies does Disney+ offer in Saudi Arabia versus Germany? Which country has the deepest streaming catalog in the world?

The answers are not where you'd expect. The deepest Netflix catalog in the world isn't in the United States — it's in the Netherlands (9,871 titles), followed by Slovakia, Bulgaria, Spain and Portugal. The US ranks 35th out of 56 markets we track. That's the kind of finding that's hard to surface unless someone is collecting the data systematically, week after week.

We just launched PlatformTrack to do exactly that.

What it is

PlatformTrack is a public index of streaming-platform catalog presence. For every (platform, country) pair we monitor, it records:

Total catalog size (movies + series)
Movie/series split
Popularity-ranked sample of titles
A weekly snapshot for historical trend analysis

Coverage at launch: 17 platforms × 57 countries = 969 pairs, refreshed every Monday at 03:00 UTC.

The site surfaces it across a few page types:

/country/{slug}/ — which platforms serve this country and how deep each catalog is
/platform/{slug}/ — which countries this platform reaches and how its catalog scales
/platform/{slug}/{country}/ — the cross-section: catalog composition, top titles, snapshot trend
/compare/{a}-vs-{b}/{country}/ — head-to-head platform comparison in a market
/ranking/ and /ranking/platforms/ — global leaderboards

The architecture

The interesting build choices were on the data side.

Source: TMDB Watch Providers

All catalog measurements come from TMDB's Watch Providers feed, which is sourced upstream from JustWatch. The canonical query is the Discover endpoint with a watch_region and with_watch_providers filter:

GET /3/discover/{movie|tv}
    ?watch_region={ISO}
    &with_watch_providers={provider_id}
    &sort_by=popularity.desc

The first page returns total_results, which is the catalog size. Subsequent pages are used to harvest the popularity-ranked sample. We don't walk every title; we record the total and snapshot the top ~40 per pair for downstream features (exclusivity calc, trending bands, "top titles" lists).

At 17 platforms × 57 countries × 2 media types × ~2 pages, that's ~3,800 API calls per sync. At 4 req/sec to stay under TMDB's rate limit, the full sync runs in about 17 minutes.

Storage: 5-table schema

pt_platforms     -- tmdb_provider_id, slug, name, logo_url, platform_type
pt_countries     -- iso_code, slug, name, region
pt_availability  -- platform_id, country_id, title_count, movie_count, series_count,
                 -- exclusive_count, last_synced
pt_top_titles    -- the popularity-ranked sample per pair
pt_snapshots     -- the append-only history table for time-series queries

The split between pt_availability (live, overwritten weekly) and pt_snapshots (append-only) is what makes historical analysis cheap. Country pages show the current state; deep-link analyses pull from snapshots.

Frontend: WordPress + virtual URLs

We needed pSEO at scale (hundreds of programmatic pages) without the overhead of real WP posts for each. The plugin (pt-engine) registers rewrite rules that route URLs like /platform/netflix/turkey/ to virtual templates:

add_rewrite_rule('^platform/([^/]+)/([^/]+)/?$',
    'index.php?pt_page_type=platform_country' .
    '&pt_platform=$matches[1]&pt_country=$matches[2]', 'top');

The template_include filter maps pt_page_type to a PHP template file in the plugin. No posts in wp_posts, no editing overhead, instant updates when the data table changes.

The trade-off: WP's native sitemap generator doesn't see these URLs, so the plugin generates its own at /sitemap-pt-{index,countries,platforms,cross}.xml from the data tables.

Sync pipeline

The weekly sync is a Python script on the same host as MySQL:

for country in countries:
    for platform in platforms:
        movie_total, movie_sample = discover_count("movie", country.iso, platform.tmdb_id)
        tv_total,    tv_sample    = discover_count("tv",    country.iso, platform.tmdb_id)
        upsert_availability(platform.id, country.id, movie_total, tv_total)
        replace_top_titles(platform.id, country.id, movie_sample + tv_sample)
        snapshot(platform.id, country.id, today, movie_total + tv_total, movie_total, tv_total)
    conn.commit()  # per-country, so a mid-run failure leaves coherent state
compute_exclusives()

After the sync, an IndexNow ping submits the changed URLs to Bing and Yandex.

Some findings from launch week

A few patterns emerged immediately from the first full sync that are worth reporting:

Europe splits into three tiers. The UK runs 141,533 title-availability records across 13 platforms. The Western Balkans run 11,000–17,000 across 5 platforms. The middle 15 markets cluster between 35,000 and 55,000. It's a sharp three-step distribution, not a smooth curve.

The GCC bundle is real. Bahrain, Kuwait, Qatar and Oman cluster inside a 290-title band — essentially identical supply, despite being separately licensed markets on paper. The UAE and Saudi Arabia anchor the deep end (17,601 and 15,797 titles respectively).

Netflix's catalog is mid-pack in the US. This was the most counter-intuitive finding. The US Netflix catalog is smaller than every European mid-tier market and most of MENA. The deepest Netflix catalogs sit in markets where Netflix is the only major streaming option, because third-party content hasn't been pulled into competing services.

We wrote these up in three research briefs on the launch:

The site is positioned for journalists, academics and analysts — every figure is citation-ready with an "How to cite" block and CC-BY 4.0 licensing on the aggregate data.

What's next

Two things are queued for the next few weeks:

Historical trend reports — once the snapshot table has eight weeks of data, country-level catalog growth/decline becomes a publishable beat.
More platforms — Hulu, Peacock and Rakuten Viki are on the shortlist. The blocker is TMDB watch-provider coverage, not our pipeline.

If you build streaming products, do regional licensing analysis, or just want to know whether your subscription is actually larger or smaller in your country than in your friend's: platformtrack.com.

The underlying availability data is also exposed via the StreamWatchHub API on RapidAPI if you want programmatic access — per-title lookups and bulk endpoints.

Methodology, data dictionary and full caveats in the Research section.

Questions, corrections or custom data requests: research@platformtrack.com.

推荐订阅源

DEV Community