Every click on a product that is out of stock is wasted budget; so are some of the clicks on a product with a broken size run, because the visitor cannot find the size they are looking for and leaves. The solution has two layers: updating the stock information in the feed frequently and accurately (the availability field), and keeping the products you do not want shown in ads out of the feed with a rule. For products that are completely sold out, the first is usually enough; a broken size run, however, shows as "in stock", so it can only be caught by a rule that looks at size availability.
Why does a sold-out product still appear in ads?
Ad platforms read stock information from the feed. If the product is listed as in_stock in the feed, the ad can keep running even if it is sold out on the site. Common reasons for this:
- The feed is updated infrequently (e.g. once a week), or the platform fetches the feed infrequently.
- Stock information is kept at product level and not sent at variant level.
- The stock field is not used in the feed at all, and every product is written as "in stock" as a fixed value.
Platforms such as Google can crawl the product page and detect a mismatch, but rather than relying on that you need to keep the feed accurate. Stock mismatch also features prominently among the reasons for Merchant Center disapprovals.
What is a broken size run, and why is it a problem?
A broken size run is when some of a product's sizes have sold out and only a few remain. For example, if only the largest and smallest sizes are left of a shirt that comes in six sizes, the product is technically "in stock", but most visitors cannot find their own size. These products:
- Get clicks, but their conversion rate drops.
- Send the ad algorithm a "this product does not sell" signal; the product's performance may be affected later too, when the sizes are restocked.
- Pull the average down in target ROAS campaigns.
The practical way to measure a broken size run is the size availability rate: the ratio of the number of sizes in stock to the total number of sizes. If two out of six sizes are in stock, the rate is about 33 percent.
Step 1: Send stock information at variant level
For sized products such as apparel and shoes there are two approaches:
- Sending variants as separate rows: Each size becomes a separate product row, carries its own
availabilityvalue and is grouped under the sameitem_group_id. The platform knows which size is in stock. - Sending the product as a single row: This is a simpler feed; but the stock information comes from the total, and a broken size run is invisible.
Whichever you choose, it is important that stock information is updated close to real time.
Step 2: Speed up the feed and the fetch frequency
Stock sync has two stages: from the store to the feed, and from the feed to the ad platform. Both need to be frequent. Check the scheduled feed fetch frequency in Merchant Center and Meta Commerce Manager; consider increasing the frequency during promotion periods when products sell out fast.
Step 3: Exclude with a rule
The availability field flags a sold-out product but does not catch a broken size run. For that you need rules that determine which products enter the feed. A typical setup:
- If total stock is below a certain quantity, remove the product (e.g. fewer than 3 units).
- If the size availability rate is below a certain threshold, remove the product (e.g. below 50 percent).
- Exception: Do not apply the size availability rule to one-size products (accessories, bags); separate these products with a category or size rule.
Set the thresholds according to your own data. The threshold may be different for low-priced, fast-moving products and different for high-priced ones; after the initial setting, compare conversion rates for a few weeks.
Step 4: Consider separating instead of removing entirely
You do not have to throw every product with a broken size run out of the feed. The alternative is to mark these products with a custom label (e.g. broken-sizes) and move them to a separate, low-budget campaign, or to direct them to an end-of-season sale. This protects the main campaign's budget while the remaining stock is sold off. For the setup, see the custom label strategy guide.
Step 5: See the stockout risk in advance
Alongside reactive filtering, seeing in advance the products with high demand and falling stock also helps with supply planning. Products with high views and sales whose size availability is dropping fast are an early warning for both the advertising and the purchasing team.
Stock and size rules with HexFeed
HexFeed regularly syncs product and stock data from your connected e-commerce platform (T-Soft, Ticimax, Shopify, ikas or XML) and generates the availability value in the feed from stock. For each feed you can choose whether variants are sent as separate rows; when variants are sent, each size carries its own stock status and the shared item_group_id value.
In feed rules you can use conditions such as stock quantity, size availability rate (stock ratio), size, category and price; a product enters the feed when it matches all the rules. On the analysis side, Stock Risk Analysis monitors the ratio of in-stock sizes for products that are in demand and classifies the stockout risk as critical, warning and watch; the Dead Stock report shows products that have stock but do not sell. For details see the stock reports and feed filtering pages; for setup, see the Meta catalog with T-Soft page.
Checklist
- The
availabilityvalue in the feed comes from real stock and is not fixed. - I decided whether to send stock at variant level for sized products.
- I checked the store → feed and feed → platform sync frequency.
- I set a lower threshold for total stock.
- I set a lower threshold for the size availability rate and made one-size products an exception.
- I considered running products with broken size runs in a separate campaign instead of removing them entirely.
- I made a note to review the thresholds against conversion data after a few weeks.
Frequently asked questions
Is it better to delete a sold-out product from the feed or to send out_of_stock?
For short stockouts, keeping the product in the feed as out_of_stock preserves the product ID and its historical data; when stock arrives, the product returns to ads quickly. For products that are permanently discontinued, removing them from the feed is cleaner.
What should the size availability threshold be?
There is no single right value; it varies by category, number of sizes and price level. The soundest way is to start with a reasonable threshold and adjust it by comparing the conversion rate of products with low size availability against the others.
Does sending variants separately solve the broken size run problem by itself?
Partly. Sold-out sizes are marked as out_of_stock, but the rest of the product family keeps receiving ads and the visitor may still not find the size they are looking for. A rule that looks at size availability solves the problem at product level.
Should I take a best seller with falling stock out of ads?
Not right away. A best seller whose stock is falling but whose sizes are still available can stay in ads; for these products the real precaution is supply planning. When size availability starts to drop, your rule separates the product on its own.
