48% better cost per click and 86% higher return on investment were the gains APG and Co reported after standardising product titles across 42 feeds and seven channels. A missing sale price in a product feed does not lower a ranking; it removes the listing from the answer entirely. That is the argument in a peak-season analysis published this week, and it applies to every channel that reads structured data: Google Shopping, Meta, Amazon, eBay, on-site search and the AI assistants shoppers now ask first. As of November 15, 2026, the practical message for ecommerce teams is that price monitoring and product data quality are the same job.
Where promotional pricing breaks
Most catalogues were written for humans browsing a category page, not for systems comparing attributes. Titles were drafted once, attributes were filled for the first channel and never revisited, and the same product carries different descriptions on the site, the marketplace and the feed. During quiet periods this creates a steady drag. During peak it compounds: cost per click rises on the terms that matter most, products get disapproved with no time to appeal, and holiday promotions fail to surface because the sale price is missing from the feed.
Three price states, one feed
Price monitoring usually compares a brand against competitors, which is the easier comparison to run. In feeds the more common failure is internal: the same SKU carrying three different prices across three surfaces, none of them wrong on its own terms. The site is right, the promotion engine is right, the marketplace listing was set by hand six weeks ago, and the feed reflects whichever one was edited last.
| Price state | Where it lives | Typical failure | Check |
|---|---|---|---|
| List price | Product page and feed | Updated on site, stale in feed | Daily diff between site and feed |
| Sale price | Promotion engine and feed | Promotion scheduled but not pushed to feed | Promotion calendar versus feed refresh log |
| Marketplace price | Marketplace listing | Set manually, drifts from site | Weekly sample of top 100 SKUs |
None of these three failures shows up in a conversion report as a pricing problem. They show up as missing impressions, disapproved listings and promotions that look like they underperformed. Separating the three states turns a vague underperformance story into a list of SKUs to fix.
What actually moved the numbers
Standardise before optimising
The Sydney-based group behind Sportscraft, Saba and JAG ran 42 separate feeds across seven channels, most formatted by hand and all pushed through a single all-products campaign. Standardising titles across catalogues and splitting one catch-all campaign into category campaigns produced the CPC and ROI gains reported. The order matters: standardise first, then optimise.
Let spend follow the listing
Brisbane-based City Beach split-tests product data changes and rolls the winners across every feed, labelling SKUs into tiers so media spend follows the listings that earn it. That is price discipline applied to media rather than to competitors: the cheapest lever is often a better title, not a bigger budget.
A feed audit before the peak week
The audit is deliberately narrow. It is not a catalogue rewrite; it is a list of the SKUs that carry the season, checked against the requirements of the channels that matter.
Product data is the least glamorous part of peak-season preparation and the one that decides whether the rest of the spend is seen. A title fix is cheaper than a bid increase, and it compounds across every channel that reads the same feed.
FAQ
Is this a pricing problem or a data problem?
Mostly data. The price is usually correct on site; the failure is that the feed did not receive it, so the promotion never surfaces.
Does better product data help with AI assistants?
Yes. Assistants read titles, attributes and categories before deciding what to show, so a listing that lacks structured attributes is unlikely to be surfaced at all.
Where should a team start?
With the SKUs that carry the season. A 200-SKU audit against channel requirements finds most of the revenue at risk in a day.
