Everything-PR (EPR) confirmed a public notice on the seller's website in 5 of 25 food recall checks (20%) on November 19, 2026. All five matched the FDA-posted notice on which products were affected and what shoppers should do for purchases from that seller.
Original research by EPR Research. Last updated November 19, 2026.
- 5 of 25 checks (20%) led to a matching public notice on the seller's website, and 5 of 5 confirmed notices (100%) matched the FDA version on scope and action.
- 13 of 25 checks (52%) found no matching notice on the site checked.
- 4 of 25 checks (16%) could not be decided because of access limits, and 3 of 25 (12%) had no verified seller website or only off-site material.
- Among the 19 checks with a verified seller website and no access limit, 5 of 19 (26.3%) led to a notice.
In This Study
How often does a retailer site carry the recall notice?
A retailer's own website carried a public recall notice in 5 of 25 seller-recall checks (20%) on November 19, 2026. The other 20 checks did not end in a confirmed notice, for the reasons shown in the table below.
The 25 checks came from nine recalls in EPR's earlier audit of FDA-posted food notices. A single recall could produce several checks if the notice named several sellers. The earlier EPR audit of food recall notices measured which details appeared in the opening of the FDA-posted notice. This study followed the sales channels the FDA notices named and checked what a shopper could find publicly on each seller's site.
| Website result on November 19, 2026 | Checks | Share of 25 |
| Public notice confirmed on the seller's website | 5 | 20% |
| No matching notice found on the site checked | 13 | 52% |
| Access limits prevented a determination | 4 | 16% |
| No seller-operated website verified | 2 | 8% |
| Seller material located off-site only | 1 | 4% |
| All checks | 25 | 100% |
Among the 19 checks with a verified seller website and no access limitation, 5 of 19 (26.3%) led to a matching public site notice. That denominator includes Eden Brothers, whose website was searchable and whose refund form was reached through the FDA-posted Eden Brothers notice. The four access-limited checks covered three Amazon pairs and one TikTok Shop pair.
Which sellers posted a notice, and did it match the FDA?
Five sellers posted a public notice, and all five (100%) matched the FDA-posted version on affected product scope and operative customer instructions for that seller. The table lists each notice, the date it showed and where a shopper finds it.
| Seller | Product | Date shown on notice | FDA first posted | Where the notice sits |
| H-E-B | Backyard BBQ Chop Salad Kit | Oct. 27, 2026 | Oct. 27, 2026 | H-E-B recall page |
| Natural Grocers | Organic mung beans, 2 lb | Oct. 22, 2026 | Oct. 22, 2026 | Natural Grocers recall list |
| WinCo Foods | Hoagie rolls | Oct. 22, 2026 | Oct. 22, 2026 | WinCo food safety page |
| Lidl US | Gelatelli coffee vanilla gelato | Oct. 16, 2026 | Oct. 17, 2026 | Lidl recall list |
| Gardeners Basics | Waltham 29 broccoli sprouting seeds | Oct. 6, 2026 | Oct. 8, 2026 | Gardeners Basics recall posts |
H-E-B reproduced the Taylor Farms notice for its salad kit, including the 13.4 oz size, UPC 041220300457, the October 30, 2026 best-if-used-by date and instructions to discard or return it for a full refund, all matching the FDA-posted Taylor Farms notice. Natural Grocers ran the five affected mung-bean pack dates alongside the refund or credit instructions. WinCo's hoagie-roll notice used the same September 10 through November 3, 2026 date range and return instructions as the FDA version, and its store page showed the same seven states.
The Lidl US gelato notice did not surface in the domain searches. Its public recall index listed the event and linked to a Lidl-hosted PDF with lot 085261, the March 26, 2028 best-by date and the same disposal or refund directions as the FDA posting. The same Lidl notice stood out in EPR's earlier audit, where its lot and best-by date sat in the first paragraph of the FDA version.
Gardeners Basics published its seed recall in a tagged blog post. The product marks and instructions matched the FDA notice for Gardeners Basics' direct sales. The post also added a purchase-date qualifier for Amazon sales that the FDA-posted notice does not contain. EPR logged that qualifier separately and could not verify it on Amazon's public site.
The dates in the table are the dates printed on each notice. They do not show when a page went live, so the audit makes no claim about whether a seller posted before or after the FDA. Lidl's PDF carries an October 16 dateline against an October 17 FDA posting, and the Gardeners Basics post shows October 6 against October 8.
Where did the audit find no matching notice?
The audit found no matching notice on the checked site in 13 of 25 checks (52%): nine frozen-pizza outlets, three sellers of the Gardeners Basics seeds, and Stewart's Shops. Each check used the seller's recall hub or a comparable page plus product-specific searches of the seller's domain.
| Group | Checks with no match | What the checked sites showed |
| Saint Francis Apizza frozen pizza outlets | 9 | Homepages, shop pages, menus, contact and return-policy pages. One indexed product listing was for the correct brand and had no recall notice. |
| Gardeners Basics seed sellers: eBay, Etsy, Walmart | 3 | eBay linked regulators from a general product safety policy. Etsy and Walmart searches surfaced unrelated seed listings or a food recalls index with no event-specific pointer. |
| Stewart's Shops, Heavenly Medley and vanilla chocolate strawberry ice cream | 1 | The site map, news and recall queries and flavor listings held no matching notice, although an FDA posting and local television coverage exist. |
What counted as a match
A retailer-hosted supplier release or PDF qualified. Unrelated listings, generic policies and off-site forms did not. Golden Gelato's site mentioned a Saint Francis Apizza relationship in its gelato distribution list, and that mention did not describe the pizza recall, so it did not count.
The FDA-posted Saint Francis Apizza notice directs customers to discard the four frozen pizza varieties or return them for a full refund, with no code required. None of the nine outlets with a verified site repeated that instruction in a place the audit could find. The Stewart's Shops result stands out because the FDA notice comes from Stewart's Shop Corp itself and names its stores in New York, Vermont and New Hampshire, yet the audit found no notice on stewartsshops.com.
What happened in the seven other checks?
Four checks were inconclusive because of access limits. In two checks, EPR could not verify a seller-operated website. In one, EPR found an Eden Brothers refund form off-site through the FDA notice. Eden Brothers had a searchable website, so it remains in the 19-check comparison.
Amazon's help page identifies a public recall list as well as a separate page for a shopper's own purchases. EPR's access to the Amazon product safety page returned an error during the check, and domain searches were restricted. That left three Amazon pairs undecided, one each for Gardeners Basics seeds, Eden Brothers seeds and HOME GROWN Calabrese broccoli seeds. TikTok Shop showed a verification screen, which left the fourth access-limited check undecided.
Two pizza sellers, Harvest Market and Madison's at Findlay Market, had no verified seller-operated website. Findlay Market runs a merchant directory and an ordering collection, and the Harvest Market record pointed to a social media page, which was outside the scope of the audit. Eden Brothers has a searchable website, but the notice that matters sits in a refund form linked from the FDA posting, and no route from the retailer's own domain was found.
Does one recall skew the 25 checks?
One recall supplied 11 of the 25 checks (44%): the Saint Francis Apizza frozen-pizza recall, which named local outlets. All 11 of those checks ended without a confirmed notice, and EPR counts that result as part of the sample.
Removing the pizza recall leaves 14 checks, and 5 of those 14 (35.7%) led to a confirmed notice. EPR calculated this figure from the scorecard as a sensitivity check. It is not a separate finding, and the 20% result stays the headline because the 25 pairs were frozen before any retailer outcome was checked.
The sample includes 23 distinct sellers, because Amazon appears in three pairs. The five confirmed notices came from five different recalls out of nine.
How did EPR choose and check the 25 pairs?
EPR built the sample from the 40 FDA-posted notices in its original study, kept the first 25 seller-recall pairs by posting date and froze that list before checking any retailer. The screen produced 48 eligible pairs in total, and the 25 newest were scored.
- Screen the 40 FDA-posted notices for explicitly named U.S. consumer sellers, including marketplaces and direct sellers.
- Rank the resulting pairs by first FDA posting date, newest first, using the canonical FDA URL rule for same-day events and sorting by seller name within each event.
- Count multiple locations of one banner once, then freeze the first 25 pairs.
- On November 19, 2026, review public recall hubs or comparable site surfaces and run product-specific domain searches.
- For each confirmed notice, compare the affected product identifiers and instructions with the most recent available FDA-posted notice for that event.
The scorecard records the individual results, sources, search terms and selection rules. The results describe what the audit could verify on one date, and they confirm public availability rather than proof of publication at the time of recall. The checks did not cover purchase-account messages, emails, app notifications or signs in stores, and the sample is not an estimate for all U.S. food recalls. The sprout-seed recalls are also covered by an FDA Salmonella outbreak investigation.
For how recall notices fit into wider crisis response, see EPR's food crisis communications guide, and for animal-food notices, which this audit left out, see the pet food recall tracker. More studies are on the EPR Research hub.
What should a retailer recall page include?
A retailer recall page should be a public index of recalls that each link to a notice listing the affected products, the codes and the action a shopper should take. Lidl US is the straightforward example in this audit, because its public index listed the gelato event and linked to a notice with the lot, the best-by date and the refund directions.
- Keep one public recall index on the retailer's own domain, reachable without a login or a store selection.
- Give each recall a searchable brand and product name, so a domain search for the product returns the notice. Lidl's did not in this audit, and its index was the only route to it.
- Publish the product size, UPC, lot or date range and the discard or return instruction on the page or in a linked PDF.
- Name each retailer-specific rule, such as a purchase-date window for one sales channel, in the same notice that carries the main instruction.
Nine of the 25 checks involved pizza outlets with no findable notice, and the FDA notice for that recall needed no code to act on. A short notice naming the four affected pizza varieties and repeating the FDA-posted discard or return instructions would have met both parts of the match test.