Retail sales monitor card transactions: the CNBC/NRF card-panel guide

A retail sales monitor built on card transactions turns anonymized payment flows into a demand read that lands well before official statistics. In the United States the reference source is the CNBC/NRF Retail Monitor, published through the National Retail Federation (NRF) Research & Insights Hub and computed from anonymized Affinity Solutions debit and credit card transactions covering 150 million+ cards.

The monitor reads two headline series - total retail sales and core retail sales, where core strips out autos, gas and restaurants - and breaks both into nine sector series. Each monthly report publishes roughly 10-14 days after month end and is never revised. That combination of speed and stability is why analysts use it as an early read on US consumer demand instead of waiting for government releases.

The Retail Monitor lands 10-14 days after month end and publishes without revisions. The July 2026 reading showed core sales up 0.30% month over month and 4.72% year over year - a figure you can act on while most official series are still in collection.

Compare that clock with the statistical agencies in the same slice. Eurostat's sts_trtu_m retail turnover and volume indices run monthly for 41 European geographies from January 1991 (424 months per series), but the record's coverage runs "to roughly six weeks before present" - a verification query on 21 August 2026 returned a 20 August refresh stamp with June 2026 as the newest observation. The card-based monitor trades that depth for about a month of extra lead time.

One archive caveat matters for backtesting: reports before January 2026 are unlinked from the current Retail Monitor page, and online ranking archives reach only to roughly 2020. A model needing a deep card-series history will not find it here - which is exactly why analysts pair the monitor with longer official series rather than replace them.

Each of the 12 reports a year carries the same skeleton: a total retail sales read, a core retail sales read excluding autos, gas and restaurants, and nine sector series underneath, expressed month over month and year over year. Because every vintage stands unrevised, the number you see in week two of a month is the number your audit trail keeps.

For calendar alignment, the hub's roughly thirty 4-5-4 merchandising calendar PDFs (back to the 2009-2011 cycle, restated variants for 53-week years) let you map the monitor's months onto retail fiscal weeks before comparing against company reporting.

A five-step workflow gets you from monitor release to a defensible demand call:

1. Pull the latest Retail Monitor report from the NRF Research & Insights Hub within two weeks of month end; log total and core MoM/YoY readings (July 2026: +0.30% MoM, +4.72% YoY core) into your series store with the release date attached.
2. Align periods with the 4-5-4 calendar using the hub's roughly 30 PDFs back to the 2009-2011 cycle, restating 53-week years so comparisons do not silently gain or lose a week.
3. Anchor the trend on the open series: pull Eurostat's sts_trtu_m through its keyless JSON or SDMX API (commercial delivery terms 4.0, 1,483 dimension-combination rows per monthly vintage, 41 geographies from January 1991) whenever the question touches Europe or needs a decades-long baseline.
4. Go local with Data.gov: cursor-walk the unauthenticated catalog API - three pages exhausted all roughly 264 'retail' matches inside the 552,271-dataset catalog - and pick up series such as Washington State taxable retail sales by county and NAICS industry (1994-present, issued 2026-07-15).
5. Reconcile and date-stamp: treat the monitor as the leading indicator and the official index as the benchmark that confirms it weeks later, stamping every figure with its as-of date before it reaches a memo or model.

The monitor measures what cards capture. It is US-only, monthly, and reported at national-and-sector level - no company row, no store row, no basket detail - so it cannot rank chains the way the Kantar-compiled Top 100 and Hot 25 do (125 ranked US companies a year on the same hub). Cash-heavy categories settle outside the panel, and none of the underlying input is exposed for reweighting.

This page is the high-frequency chapter of a broader pool. Start with the other specialty retail data guide, which maps all 20 pooled datasets for the industry, then go deeper on the official side with European retail sales data by country and the modeling perspective in how data scientists use other specialty retail data.

Record-level detail lives on the National Retail Federation (NRF) Research & Insights Hub, Eurostat - European Statistical Office Data Portal and Data.gov Retail Catalog (552k+ Datasets Federated Search) pages. For shortlists, keep the best other-specialty-retail datasets ranking and the free other-specialty-retail datasets list handy, or browse the full inventory in the other-specialty-retail data hub.