Size-out personalization: recommending in-stock alternates when a size sells out

The shopper found the perfect item and their size is gone. What you show next decides whether they buy something or leave. How to personalize the recovery when sizes sell out.

Why the default sold-out experience fails

Most stores show the same dead end to everyone: a grayed-out size selector and a suggestion to check back later. That treats a high-intent shopper the same as a casual browser. But the shopper staring at a sold-out size is one of the most qualified visitors on your site. They have already decided they want the product. The only problem is inventory, and inventory problems have inventory solutions.

The cost compounds during peak season. When popular sizes sell out across a category, the default experience does not just lose one sale; it trains shoppers that your store cannot fulfill their size, and they start their next search on a competitor's site. Every size-out is a small trust event, and the recovery is where trust is either rebuilt or lost.

What a good recovery knows about the shopper

The minimum viable recovery knows three things: the shopper's size, the product they wanted, and what is in stock right now. From there, the ranking writes itself: same product in the closest available size first, then the same silhouette in their size, then stylistically similar items they have shown interest in. Each step down the list should feel like a thoughtful alternative, not a random clearance rack.

Fit history makes the recovery dramatically better. If the shopper's profile says they size up in this brand's trousers, the closest-available-size recommendation should account for that instead of blindly offering one size down. Returns data is fit data, and the size-out moment is exactly where it pays off.

Ranking the alternates

Not all alternates are equal, and the order matters. Start with the same product: a different colorway in their size, or their size in a sister product with the same fit block. Then widen to the same category and aesthetic. Only at the end should you show the generic bestsellers, and even those should be filtered to their size. An alternate they cannot wear is not an alternate.

Availability must be real-time. Nothing destroys trust faster than recommending an alternate that is also sold out in their size. The recommendation call has to check live inventory at render time, not rely on a cached feed from this morning. During high-velocity sales, inventory changes by the minute.

The messaging around the swap

How you frame the alternate changes whether the shopper accepts it. Be explicit about why you are showing it: your size in this color is gone, but it is available in these options. Transparency about the substitution beats pretending the alternate was the plan all along. Shoppers accept a good Plan B when they understand why Plan A failed.

Back-in-stock notifications are the safety net, not the strategy. Offer the notification, but do not make it the only action on the page. A shopper who leaves with only a notification set is a shopper who will probably buy it from whoever has it in stock today. The alternate recommendation is the sale; the notification is the consolation.

Measuring whether it works

Track the recovery rate: of sessions that hit a size-out on a product page, what share convert on a recommended alternate within that session? Segment by how far down the alternate list the purchase came from. If most recoveries come from the first two recommendations, your ranking is working. If shoppers routinely scroll past ten alternates, the ranking needs retraining.

Also watch the return rate on recovered orders. An alternate that gets returned at twice the normal rate is not a real recovery; it is a delayed loss with shipping costs attached. The goal is not just to save the session but to save it with something the shopper keeps. Fit-aware ranking is what separates the two.