ONS Retail Sales Index Download: What the Record Holds, Delivered

Three records in the catalog answer the query outright, cut from the same measurement program.

The UK ONS Retail Sales Bulletin - Great Britain is the flagship: written commentary beside the part analysts actually consume, roughly 30 tables of value and volume reference series across the store-type panel, with the companion layers attached to every edition. It scores 9 out of 10 on Datadory's quality rubric against a catalog-wide average of 7.81 across 1,744 datasets.

The UK ONS Retail Sales Index - Retail Industry is the same measurement at matrix scale: 622 series keyed by stable four-character CDIDs, value and volume twins side by side, including the textile, clothing and footwear store split apparel analysts reach for first.

The UK ONS Retail Sales Index - Reference Tables carries the multi-table structure behind both, including the CPSA and KPSA contribution series that decompose how much each store type added to or subtracted from the headline move.

One perimeter note belongs up front rather than absorbed silently: the geography is Great Britain - England, Wales and Scotland - because Northern Ireland sits outside the survey frame. Any panel labelled "UK retail" built on these products inherits that caveat, and it belongs in the chart legend before someone else writes it into a comment.

Read straight off the July 2026 edition, laid out the way a delivery resolves:

time_period : 2026 Jul
series      : All Retailing, Including Automotive Fuel
agg_code    : AGG21        dataset_id: J5EK
index_value : 103.8        (chained volume, seasonally adjusted, 2023=100)

series      : Non-Specialised Predominantly Non-Food Stores   <- broadline
agg_code    : AGG12
index_value : 103.3        (All Businesses variant; Large and Small also carried)

series      : All Retailing, Excluding Automotive Fuel
index_value : 104.6        (fuel pulls the aggregate around)

Three readings fall out of one edition. Momentum stayed positive at 1.1% three-month-on-three-month volume growth. The broadline segment printed 103.3 - against 102.7 in June and 101.9 in May - sitting below the all-retailing line: the department-store shape of selling underperforming the whole. And the headline ran ahead of its ex-fuel twin, which is exactly why both variants ship side by side.

Channel mix separates cleanly too. Internet sales accounted for 27.4% of total retail in July 2026 (series J4MC), and textile, clothing and footwear stores alone moved GBP 327.5 million a week online against GBP 23.9 million in January 2008 - a near fourteen-fold shift inside one continuous series, which is precisely why unbroken history beats a snapshot.

The store-type panel makes this a demand grid rather than one headline number, and the second table below maps search intents to the cut that settles them.

Category rotation reads straight off the aggregates: predominantly food stores - GBP 201,285 million of 2023 sales at a 37.57 weight, subdividing into supermarkets (GBP 184,104 million) and specialist food stores (GBP 13,803 million) - set against the non-specialised predominantly non-food segment shows whether spend is moving between groceries and goods. Household goods stores - furniture, electricals, hardware, paints and glass - printed 106.1 in July 2026, the strongest major store group, which is why the cut doubles as Britain's nearest thing to a DIY and building-products demand index.

One discipline note: weights differ between tables, so confirm which weighting applies before rolling sub-sectors into a custom composite. That single habit prevents most aggregation errors people blame on the data.

Deeper than almost anything else monthly in the catalog. Monthly series begin January 1988 - roughly 38 years spanning the early-90s downturn, the financial crisis, the pandemic collapse and rebound and the post-2022 inflation shock - with quarterly and annual aggregations riding alongside. Internet-sales series open in November 2006, capturing the entire meaningful arc of British online retail, and the J4MC online-share ratio starts January 2007.

These are first estimates, so revision is part of the arrangement rather than an accident. Summary and quality tables ship with revision triangles - exactly the material you want when measuring how far a first print moves. One episode belongs in any model's memory: the August 2025 edition slipped to September 2025 while a seasonal adjustment error affecting January-May 2025 was corrected, so extracts taken before the correction disagree with today's figures.

Datadory handles that the unglamorous way: every value travels stamped with its reporting period and vintage, so a restated number reads as a new edition of the same series instead of a contradiction, and cross-vintage comparisons become analysis rather than archaeology.

The American demand analogue sits nearby too: Census MRTS publishes monthly US retail sales by NAICS kind of business from January 1992, including building material and supplies dealers as their own cut. Three mismatches need handling before any transatlantic chart - the UK series run as indices on a 2023=100 base while MRTS prints dollar levels, Great Britain is not the United States, and UK accounting months that sometimes run four or five weeks drift against calendar months enough to shift a month-on-month comparison by several percent.

The trade-off against the storefront layer is history versus granularity: realtime SKU-level texture with no archive, or nearly four decades of monthly observations at national and store-type grain. The first table lines the shelf up side by side.

• Nowcasting and feature stores - seasonally adjusted volume series across roughly 34 store types are ready-made macro features, keyed on identifiers stable enough to train on across decades.
• Market sizing - pounds value series give Great Britain's retail spend directly, with 2023-base weights doing the aggregation arithmetic for every segment model.
• Channel-mix tracking - internet sales by store type quantify the online shift from November 2006 onward instead of inferring it from company disclosures.
• Benchmarking - retailers test like-for-like numbers against the sector index to separate execution from category tide, with the implied deflator splitting price from units.
• Backtesting - investors replay UK consumer exposure across nearly four decades without stitching eras together.
• Citation-grade reporting - accredited official statistics status with a four-character identifier behind every series survives scrutiny in print.

Persona routes with worked examples: data scientists, e-commerce operators, market researchers and investors & quants.

Because six hundred and twenty-two series whose identifiers double as join keys, a layout that reshapes itself across four decades, weights that differ between tables, accounting months that are sometimes five weeks long, and a first-estimate regime that revises its own past - assembling all of that by hand is a project. Receiving it is a query.

Datadory delivers one normalized schema across the bulletin, the index matrix, the reference tables and the US comparators, so a UK-versus-US consumer chart is a join rather than a fortnight. Delivered daily, weekly, or hourly - your call - into your warehouse, your notebook or your dashboard.

Name the store types, the periods and the splits when you request a sample - the food-versus-goods pair back to 2010, the household goods basket with its hardware sub-cut, internet-sales shares by store type for the last five years - and the extract comes back cut to that scope, with the complete field dictionary and coverage statement attached. You keep all three regardless of what happens next.

That is the point of sample-first: the thing you evaluated is the thing that ships, drawn from the same delivery path as the ongoing feed rather than a fenced-off demonstration surface. Fields beyond the verified core get confirmed cell-by-cell against your nominated scope before the pipeline locks naming for your side.

Every record referenced here is indexed with field dictionary, sample rows and coverage in the broadline retail data hub, and the scorecard lives at best broadline-retail datasets.

Product pages worth opening: UK ONS Retail Sales Bulletin - Great Britain, UK ONS Retail Sales Index - Retail Industry and UK ONS Retail Sales Index - Reference Tables.