How loyalty‑card prices and scanner data affect UK inflation
The ONS tested replacing shelf prices with loyalty‑card prices and began using grocery scanner data in 2026; it found no material change to headline CPI.
Originally published:
Quick answer: did loyalty discounts change the published CPI?
Short version: the ONS tested swapping shelf prices for loyalty‑card discount prices in an internal 2023 exercise and reported that producing indices with the discounted prices did not produce a material difference to published headline CPI or to broad divisions. [1]
What exactly did the ONS do and when?
According to an ONS freedom‑of‑information release, the statistics team began collecting loyalty‑price data at the end of 2022 and ran a simple internal exercise in 2023 that replaced shelf prices with loyalty prices for discounted items to see how an index built from those prices would move. [1]
Separately, the ONS published an overview in January 2026 explaining that groceries scanner data will be introduced operationally from February 2026, with the first scanner‑based results published in March 2026. [2], [1]
Why use scanner data? What does it add?
Scanner data are records captured at the point of sale that include every product sold and the revenue and quantity sold, so they can show how often promotions, multibuy offers and loyalty discounts were actually taken up by shoppers rather than only listing advertised shelf prices. [2]
The ONS says scanner records let it move from a sample of about 25,000 monthly price quotes collected by humans to roughly 300 million price points from over a billion product units per month for about half the grocery market, improving timeliness and representativity. [2]
If scanner data show discounts, why didn’t inflation numbers move much?
Two reasons the ONS gives: first, the 2023 exercise showed that simply substituting discounted loyalty prices for shelf prices didn’t materially change headline or divisional CPI; second, scanner data let statisticians weight prices by actual sales volumes and promotional uptake, which tends to produce a more accurate but not necessarily very different headline rate. [1], [2]
Put another way: discounts matter for some products and shoppers, but when indexed across the whole basket and adjusted for quantities bought, the net effect on the headline figure was small in the ONS test. [1], [2]
What this means for shoppers and for trust in inflation numbers
For most readers the practical takeaway is that official headline inflation measures remain based on methods the ONS has tested and is refining; adding scanner data is meant to make the published series reflect actual spending patterns more precisely rather than to produce sudden large revisions. [2], [1]
The ONS also notes it cannot publish the raw discount fields because that would risk identifying individual businesses, so some details in its internal slides were redacted under data‑disclosure rules. [1]
A consumer angle: do loyalty cards help you?
Independent reporting reminds readers that loyalty schemes are widespread and that whether a card is ‘good value’ depends on individual shopping patterns and which promotions you use, which is why the ONS wants data that captures actual take‑up rather than advertised prices alone. [3], [2]
Bottom line
The ONS tested loyalty‑price substitution and is introducing scanner data in early 2026 to better measure promotions and sales; its published position so far is that these changes improve accuracy but did not produce a material change to headline CPI in the internal test. [1], [2]
Sources & further reading
- Effect of loyalty card discount schemes on inflation - Office for National Statistics ↗ons.gov.uk · Primary source
- Overview of how we use scanner data in consumer price inflation statistics - Office for National Statistics ↗ons.gov.uk · Primary source
- Monday briefing: Do loyalty cards offer customers a good deal? | The Guardian ↗theguardian.com
What do you think?
Join the conversation with your Ali Pro account. Keep it kind, stay on topic, and leave personal details out.