BARB Main Site - Viewing Data Hub Data

It is the public face of Barb Audiences Ltd - the United Kingdom's independent television audience measurement body, one of its Joint Industry Currencies alongside ABC, RAJAR, PAMCO and ROUTE, self-described as "the industry's standard for understanding what people watch". Broadcasters commission it, agencies plan against it, and streamers that subscribe get measured by the same national panel. Within Datadory's fifteen-dataset broadcasting slice it is the umbrella record: where its sibling BARB Weekly Top 50 Shows and Viewing Data drills into one weekly dashboard with device breakdowns, the hub exposes four viewing-data products - Weekly top 50 shows, the Total Identified Viewing summary, the Monthly Viewing Summary and Archive viewing data - each a different cut of the same currency.

The archive is the deep end: weekly top-10 and top-30 programme rankings running July 1998 to September 2018, weekly viewing by channel group and TV-set viewing summaries from January 1992 onward, four-screen viewing summaries covering September 2018 to January 2022, monthly channel viewing up to September 2025, and As Broadcast and As Viewed top-50 series through December 2025. It scores 8 out of 10 on Datadory's quality rubric.

Get a sample of this dataset cut to your channels, weeks and programmes.

One row per programme per reporting week in the flagship ranking; one row per channel or group per week or month in the summaries. Three straight from broadcast week 202631:

week           : 202631
Broadcaster    : BBC
Programme Title: Ann Droid: Series 1, Episode 3
mbid           : s86p2q
seven_day_value: 5051.700  ($thousands)
rank           : 1

week           : 202631
Broadcaster    : BBC
Programme Title: The Great British Sewing Bee: Series 12, Episode 3
mbid           : s8zwzw
seven_day_value: 4248.800  ($thousands)
rank           : 2

week           : 202631
Broadcaster    : ITV
Programme Title: Coronation Street: Series 67, Episode 147
mbid           : s8t4zw
seven_day_value: 3679.200  ($thousands)
rank           : 3

Three rows, three different stories the same shape tells: a drama premiere clearing five million viewers in seven days, a craft competition holding 4.2 million, a sixty-year-old soap still pulling 3.7 million. Note mbid doing quiet work - Barb's own identifier for the transmitted programme instance - which means a week's rows key into your own title database without fuzzy matching. The archive layers carry the same discipline back through time: a 2003 channel-group summary and last month's Total Identified Viewing figure sit in compatible frames, so a thirty-year trend needs no re-plumbing between eras.

The core row shape carries four load-bearing fields, defined in the dictionary below: the reporting period, the broadcaster or channel group, the programme title, and the audience figure in thousands. Around that spine, individual products add their own columns - rank position in the weekly rankings, channel-level rollups in the monthly summary, device-family splits in the four-screen series.

Additional fields on request. The wider delivery carries more than the core row: 7-day and 28-day consolidation windows and per-device components on current records, the mbid programme identifier for join keys, SVOD flags distinguishing streaming averages from consolidated broadcast figures, and the archive series' own column sets. Each gets defined and exemplified against live rows before your pipeline locks a schema - because the honest caveat about this hub is that no single published field dictionary covers all four products, so the sample is where the dictionary gets proven rather than assumed.

• Geo: The United Kingdom, measured as one national panel. Barb reports the country that watches together; these products carry no sub-national split.
• Temporal: Two clocks side by side. Current weeks flow continuously - live products carry the latest reported periods - while the archive stretches from January 1992 to December 2025 depending on product: channel-group viewing from January 1992, programme rankings from July 1998, four-screen summaries September 2018 to January 2022, monthly channel viewing before September 2025, top-50 series through December 2025. Any current figure drops into a three-decade context without changing source.
• Granularity: Weekly and monthly aggregates at three levels - individual programmes, channels, and broadcaster groups. The finest cut is one row per transmitted programme instance per week, ranked into a top 50; the coarsest is a monthly channel-level summary.

That span is the pitch. A ratings currency usually gives you now; this hub gives you now plus the whole modern history of British television measurement behind it.

API, files, or your warehouse. Daily, weekly, or hourly.

Match cadence to the decision being fed. Ad-sales desks want each new reported week landing fast, because pricing conversations happen while the audience is fresh; strategy and commissioning teams take the archive backfill once and append as periods close; analysts building long-run genre curves pin a fixed historical snapshot so a chart built today reproduces next quarter. Because every record carries its reporting period explicitly, incremental pulls key cleanly without double-counting - and mixing a current extract with archived years stays a join on period and channel, not an archaeology project.

The sample comes first: name the products, weeks and channels you care about, and real rows come back cut to that scope before anything reaches production.

• Market researchers and consultants anchor media-market sizing and channel-share analysis in the UK's own currency rather than vendor estimates - the market researchers use cases page shows the workflow.
• Journalists, academics and students treat it as the citation-grade answer to "what did Britain watch": any claim about UK audiences resolves to a Barb figure, and the archive supplies the historical curve behind the headline - see journalists academics use cases.
• Ad sales and media planning price slots against what programmes actually delivered, week after week, across every rival channel at once.
• Commissioning and scheduling track genre performance over time; the archive turns "reality is saturating" from a hunch into a thirty-year curve.
• Competitive intelligence teams benchmark streamer performance against linear hits in shared frames - patterns collected under competitive intel product teams use cases.

Persona fit has edges: sales-growth teams find no prospecting signal here, and e-commerce operators read it only indirectly as an ad-market demand proxy. This hub measures audiences, not buyers.

Three notes worth having before you request a sample. First, this is the umbrella, not one flat table: four products with different column sets and different temporal windows, so verify the window of the specific product you join against before building a trend series - four-screen summaries end January 2022 and old-form top-10/top-30 rankings end September 2018. Second, the sibling record drills deeper on one dashboard: BARB Weekly Top 50 Shows and Viewing Data carries ten verified fields including 7-day and 28-day consolidation and device components, scored 9 out of 10; start there if your question lives entirely in current weeks. Third, it is demand-side: pair it with supply-side registries when the question needs infrastructure too.

Notes and neighbors that pair well with this page:

• The broadcasting data hub collects all fifteen datasets in the industry, this one included.
• Sibling deep-dive: BARB Weekly Top 50 Shows and Viewing Data - the same currency at finer grain for current weeks.
• UK regulator complement: Ofcom research and data - Broadcasting and Media - market context around the measurement itself.
• Head-to-head: BARB Viewing Data Hub vs RabbitEars.info - UK audiences against US station registries.
• Terminology: audience measurement - how panel-based currencies like Barb work.