Blog / Product evidence

Shopping Spreadsheet Link Freshness: Prices, Stock & Listing Changes

Understand why a working product URL can still contain outdated prices, sizes or listing details, and what a useful shopping record should preserve.

The central idea

Link health, listing identity and commercial availability are three separate tests of a shopping reference.

A product link can still open after its price, available sizes or listing identity has changed. This article examines those different forms of record freshness and the maintenance decisions behind them. The distinction helps explain why a saved reference needs context as well as a working address.

The wider evidence for digital decay

Pew Research Center's May 2024 study found that 38% of sampled webpages from 2013 were no longer accessible at the study's observation time. For pages collected in 2023, the figure was 8%. The research drew on Common Crawl samples; it did not examine OSSBuy or measure shopping-spreadsheet failure rates.[1]

Those figures provide context for a familiar problem, not a prediction for a particular list. Older web references can disappear, but the study does not establish how long a seller link will last or what share of this catalog will change. Transferring the percentages directly to shopping entries would turn useful research into an unsupported statistic.

The editorial lesson is narrower and more durable: publishing a link is not the same as maintaining a reference. A directory accumulates a responsibility to describe when its information was observed, even if every destination still loads on the day a reader visits.

A live page is only one kind of freshness

Consider three different states. A URL might return no useful page at all. A page might open but show a different item. Or the same item might remain visible with the desired size unavailable. These failures affect different parts of the shopping decision, even though a simple link checker may flag only the first.

Google Merchant Center's availability documentation separates in-stock, out-of-stock, preorder and backorder states. That distinction illustrates why 'the page exists' and 'the item can be supplied now' are not equivalent. It is a terminology reference here, not a statement that this directory participates in Merchant Center.[2]

The guide to browsing saved product records explains where to check reference details while using this site.

Three layers of record freshness — our editorial framework
LayerWhat it establishesWhat remains unknown
Link healthA destination can be openedWhether it is still the same listing
Listing identityThe title, image and variation matchWhether an order can be fulfilled
Commercial availabilityThe current seller accepts the selected optionThe buyer's eventual item and delivery outcome

The date belongs to a specific observation

Our catalog snapshot was fetched on 16 November 2026. An article's later source-review date does not refresh that saved product data. A price estimate, a stock observation and a QC photo can also originate at different times. Combining them into one undifferentiated 'updated' label would conceal those differences.[3]

This is why a dated reference can still have value after an item sells out. It may document a silhouette, an earlier price estimate or the existence of a photo group. Its historical value does not make it a present offer. The reader should be able to tell which of those roles the record is serving.

A replacement link introduces another distinction. Preserving a category is not the same as preserving the original listing. If a reference is replaced with a similar item, the new record deserves its own identity and evidence rather than inheriting every claim attached to the old one.

Content drift behind a working address

A 2016 PLOS ONE study by Jones and colleagues examined changes in web resources referenced by scholarly articles. It distinguishes missing destinations from content drift: the address can still resolve while the material at that address changes. The study concerns scholarly references, not retail listings, but its distinction gives a precise name to a problem that a working-link indicator cannot settle.[4]

In a shopping context, the relevant change might be a variation, price basis, seller description or the product shown on the page. A link can lead somewhere useful while no longer supporting a particular old claim. The question is therefore not just whether a page opens, but whether it still provides the evidence for which the reference was saved. That is an editorial application of the research distinction, not a measured shopping-link failure rate.

Imagine a saved record describing a navy hoodie while the destination now opens with a black version selected. The page has not necessarily disappeared, and the shop may still offer both colours. Yet a reader comparing the saved photograph with the default destination is already working across two states. A stable product family does not remove the need to preserve the exact variation attached to the original observation.

Different fields run on different clocks

A catalog record contains information that can age at different rates. A model name may remain meaningful for years. A size list can change with a seller's assortment. A displayed reference price may change when an option changes, while an inspection photograph remains an accurate historical image. The same record can therefore be fresh enough for visual research and too old for a claim about immediate availability.

An article review adds another clock. Correcting the wording of a shipping explanation on Tuesday does not remeasure a package recorded on Monday, and checking a source publication does not refresh a product's stock. An undifferentiated update date can accidentally collapse those events. A more useful editorial account identifies the observation being dated: publication review, catalog fetch, stock check or photographed reference.

The original snapshot can remain valuable after later changes precisely because it preserves an earlier state. A historical price may help explain a past comparison without becoming today's offer. The limitation is not that historical data has no value; it is that its time frame has to remain attached. Removing the date creates an appearance of immediacy that the underlying evidence cannot supply.

Editorial meaning of different dates in a product-reference article
Date typeWhat happenedWhat it does not refresh
Article reviewThe analysis or sources were checkedSeller stock and product measurements
Catalog fetchSaved product fields were collectedChanges made after collection
Stock observationAn option's status was recordedFuture fulfilment availability
Photo contextA pictured reference existed at that timeEvery later item or variation

Useful history and misleading replacement

A discontinued destination creates a choice for an editor: preserve the historical reference, retire it from discovery, or introduce a new candidate. Those choices have different meanings. Replacing the URL with a visually similar product can keep a page populated, but it also creates a new relationship between the old description and the new destination. That relationship deserves scrutiny rather than automatic inheritance.

Consider an illustrative sneaker entry originally supported by photographs of a white-and-green pair. A replacement listing showing a white-and-blue pair may fit the same broad category while differing in the feature that interested the reader. Transferring the old QC photos to that new destination would make the page look more complete while weakening its evidential connection. Surface similarity is insufficient justification for carrying over historical claims.

By contrast, a clearly separated new record can preserve useful continuity: the old reference explains what was previously documented, while the new one records its own title, options and images. This is the difference between maintaining access to a topic and pretending the underlying item never changed. It is especially important when a reader returns through an old bookmark rather than encountering the catalog for the first time.

Maintenance priorities beyond a green status light

Not all stale fields have the same consequence for a reader. An outdated category description may cause confusion; an outdated price basis can change a comparison; a mismatched variation can cause someone to research the wrong item. A maintenance system that treats all successful page loads as equally healthy misses this hierarchy. The most consequential discrepancies concern identity and the meaning of the displayed offer.

A hypothetical maintenance report makes the distinction concrete. 'All links opened' is a statement about reachability. 'All links opened and still displayed the recorded product family' adds identity checking. 'All selected variants could be ordered at the recorded price' makes a much stronger commercial claim. Each result would require different observations, so the first should not be used as shorthand for the last.

For readers, the benefit of a well-maintained spreadsheet is less detective work when returning to a reference. They can see what the record means, which part was checked and where a change may have occurred. That makes maintenance part of the editorial service rather than a hidden technical chore. A trustworthy list does not freeze the market; it makes its own observations and revisions understandable.

Maintenance is part of the editorial product

For an OSSBuy spreadsheet, trust is built less by the promise that nothing will change than by a clear account of what has been checked. That includes the scope of the list, the date of saved details and the difference between an unavailable destination and an unavailable size.

Our interpretation is that freshness should be treated as a set of observable claims. The web research establishes that disappearance is real; the catalog example shows why timestamps need context. Neither supports a guarantee of permanent links. A useful directory can acknowledge change while still preserving the information that makes a reference worth finding.

When applying those checks to a shortlist, browse current product categories and compare each saved record with its matching listing.

Sources & editorial notes 4 references

Sources checked 17 Nov 2026. Published findings and catalog observations are attributed below; interpretations and illustrative calculations are our own.

  1. Pew Research Center · Published 17 May 2024When Online Content Disappears

    Common Crawl-based analysis of web persistence. The 38% figure concerns sampled pages from 2013, and the 8% figure pages from 2023, at the study's observation time; neither measures shopping spreadsheets.

  2. Google Merchant Center HelpAvailability [availability]

    Distinguishes in-stock, out-of-stock, preorder and backorder statuses. A reference for the meaning of availability, not evidence of Merchant Center participation by this site.

  3. OSSBuy SheetOur catalog: dated editorial snapshot

    Local catalog fetched 16 November 2026: 343 product records, 103 distinct non-empty brand labels, 326 records with QC groups and 2,059 source QC groups. The Jordan White Oreo example records 1,720g and 36 × 25 × 13cm with sampleCount 1. These are catalog observations, not live inventory or a representative market sample. Additional local audit: source categories include 56 sneakers, 28 bags/backpacks, 25 jeans, 22 hoodies/sweatshirts and 13 sweaters/knits; Other is a brand-field placeholder on 39 records. No source QC group is marked preferred. A second single-sample record, Nike Light Blue Swoosh Tank Top, has a 207g package measuring 25 × 19 × 3cm. All observations refer to the same dated snapshot.

  4. PLOS ONE · Jones et al. · Published 2 Dec 2016Scholarly Context Adrift: Three out of Four URI References Lead to Changed Content

    Original research on changing web resources referenced in scholarly publications. Used for the distinction between disappearing links and content drift; its numerical findings are not applied to shopping links.

More from the journal

Another perspective.

All articles