Datasets:

crawlora-net
/

streaming-fragmentation

Tags:

License:

Dataset Viewer

The dataset viewer is not available for this subset.

Cannot get the split names for the config 'default' of the dataset.

Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 290, in _generate_tables
                  pa_table = paj.read_json(
                      io.BytesIO(batch), read_options=paj.ReadOptions(block_size=block_size)
                  )
                File "pyarrow/_json.pyx", line 342, in pyarrow._json.read_json
                File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
                  return check_status(status)
                File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
                  raise convert_status(status)
              pyarrow.lib.ArrowInvalid: JSON parse error: Column() changed from object to string in row 0
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                                         ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 101, in _split_generators
                  pa_table = next(iter(self._generate_tables(**splits[0].gen_kwargs, allow_full_read=False)))[1]
                             ~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 304, in _generate_tables
                  batch = json_encode_fields_in_json_lines(original_batch, json_field_paths)
                File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 111, in json_encode_fields_in_json_lines
                  examples = [ujson_loads(line) for line in original_batch.splitlines()]
                              ~~~~~~~~~~~^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 20, in ujson_loads
                  return pd.io.json.ujson_loads(*args, **kwargs)
                         ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
              ValueError: Expected object or value
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 66, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                               ~~~~~~~~~~~~~~~~~~~~~~~^
                      path=dataset,
                      ^^^^^^^^^^^^^
                      config_name=config,
                      ^^^^^^^^^^^^^^^^^^^
                      token=hf_token,
                      ^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
                  info = get_dataset_config_info(
                      path,
                  ...<6 lines>...
                      **config_kwargs,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Streaming Fragmentation Index — data

Where the most-watched movies and TV shows in the US actually stream — and how scattered, exclusive, and expensive that has become. Two open snapshots (June 2026):

Popular-100 — the 100 most-popular titles (top 50 movies + 50 shows), the demand-weighted head of the market.
Broad ~5,000 — a 5,047-title cross-section pulled across every genre and decade.

It is the dataset behind the Crawlora study "How Many Streaming Subscriptions Do You Need in 2026?". Every title's full list of ways-to-watch (offers) was pulled via Crawlora's JustWatch endpoints and each offer mapped to the service behind it.

Files

file	rows	description
`data/popular-titles.csv`	100	per-title availability for the 100 most-popular US titles
`data/broad-titles.csv`	5,047	per-title availability for the ~5,000-title cross-section
`data/popular-summary.json`	—	aggregates for popular-100 (fragmentation, exclusivity, monetization, overlap, cost)
`data/broad-summary.json`	—	aggregates for the broad corpus

CSV schema (one row per title)

column	meaning
`id`	JustWatch title id (`tm…` movie / `ts…` show)
`title`, `year`, `type`	title, release year, `movie` or `show`
`num_subscription_services`	# of distinct standalone streaming services carrying it
`has_subscription`	on ≥1 subscription service
`exclusive`	on exactly one subscription service
`rent_buy_only`	no subscription — rent/buy only
`cinema_only`	no subscription/rent/buy — in theaters only
`subscription_services`	pipe-joined service brands (standalone OTT; ad-tier/quality/reseller variants collapsed to parent)
`monetization_types`	pipe-joined offer types seen (`FLATRATE`/`ADS`/`RENT`/`BUY`/`FREE`/`FAST`/`CINEMA`)

Headline findings (June 2026, US)

Popular-100

14 distinct streaming services carry the 100 most-popular titles; you'd need 11 subscriptions to watch all 74 that are on a subscription — and 26 of 100 aren't on any subscription (rent/buy or still in theaters).
The Big Four (Netflix, Disney+, Max, Prime Video) cover 51%. 51% of titles are exclusive to a single service.
All 11 services ad-free ≈ $152/mo (~$1,829/yr); even the Big Four ad-free ≈ $70/mo.
Shows 98% on a subscription vs movies 50% (the movie gap is new-theatrical release-window timing).

Broad ~5,000

Content scatters across 93 distinct streaming services; 60.7% of titles are exclusive to one. A concentrated head (6 services cover 80% of what's streamable) and a very long tail (65 services to cover all of it).
Netflix dominates exclusives — 1,220, more than the next four services combined. Big Four cover 59.5%.
Movies 70% / shows 89% on a subscription; ~21% on no subscription.

Method (brief)

For each title, JustWatch offers were pulled — justwatch_popular for the head; justwatch_discover fanned across genre × year for the cross-section (each call ≤50 results, no offset, deduped by id). Each offer's provider is classified into a service class — standalone OTT subscription / live-TV bundle (vMVPD) / free ad-supported channel (FAST) / single-network cable-login app / rent-buy store (TVOD) / cinema — and ad-tier, quality (SD/HD/4K), and reseller-channel ("… Apple TV/Amazon Channel") variants are collapsed to the parent brand. "Subscriptions needed" counts only standalone retail streaming services; a greedy set-cover gives the fewest services to reach 50/80/100% of the streamable titles. Subscription prices are US list prices (mid-2026), the one input not from JustWatch. The data comes from Crawlora's JustWatch API; the study writeup documents the full method.

Caveats

Point-in-time snapshot (June 2026), US storefront — offers change weekly and by country.
Popularity-stratified, not a census — the broad corpus is the popular head across genres and decades, so every concentration figure is a floor; the true long tail is even more fragmented.
The movie / no-subscription gap is largely release-window timing (new theatrical), not permanent lock-out.
~29 obscure/regional providers in the broad corpus are left unclassified (they appear in the data and affect a small number of titles).
Subscription prices are hand-maintained list prices; treat dollar figures as current-as-of-June-2026.
No quality claim — only where, and at what cost, you can watch what's popular.

How to cite

This repo ships a CITATION.cff, so GitHub shows a "Cite this repository" button. Plain text:

Crawlora (2026). Crawlora Streaming Fragmentation Index: where the most-popular US titles stream (v1.0.0) [Data set]. https://github.com/Crawlora-org/streaming-fragmentation-data

Data collection & ethics

Public availability metadata only. Each record is the public "where to watch" listing for a title — the same information a viewer sees. No page content is republished and no authentication was bypassed.
No personal data — titles, services, prices, and availability flags only.
Point-in-time. Treat as a snapshot, not a live availability feed.

License

Data is licensed CC BY 4.0 — free to use, share and adapt with attribution to Crawlora (https://crawlora.net), a link to this repository, and an indication of changes. See LICENSE. Generated with Crawlora's JustWatch API.

Downloads last month: 11

Total file size:

1.11 MB