The Basket Nobody Sells

When a buyer describes a situation instead of naming a product, nineteen of the thirty pages captured carried no commerce at all, and where commerce did appear outside apparel, eight of the eleven products found were pre-bundled kits.

Twelve items named. Nothing sold.

Type “what do I need for a long haul flight with a toddler” into Google and you get a complete shopping list. The AI Overview names twelve items. A spill-proof straw cup. Kid-friendly headphones. Mess-free coloring pads. Suction spinners. A baby carrier. Disposal bags. Two full outfits plus a spare shirt.

A parent is standing there with a credit card and a departure date.

Nothing is sold against any of it. No sponsored results. No shopping grid. No product module of any kind. Mommy blogs and YouTube own the page.

If you sell any of those twelve items, that page is open ground and no merchant has taken it.

That is not one unlucky query. Across thirty situation queries captured on 2026-08-04, nineteen returned exactly that: a complete commercial vacuum.

The vacuum is where the money is

The list of vacuum queries reads like a retail calendar.

“What do I need for a college dorm room” returns zero commerce. That is one of the largest retail moments of the year.

“What do I wear apres ski” returns zero commerce. The AI names eight categories: sweaters, thermal leggings, jeans, insulated winter boots, a ski jacket, fleece, a puffy vest, winter accessories. Vogue and Saks Fifth Avenue own that page through editorial.

“What do I need for a first apartment kitchen” returns zero commerce. The AI names a chef’s knife, a frying pan, and dinnerware, and then goes further. It advises the buyer to skip large pre-packaged sets and buy individual high-quality essentials. Google is telling the buyer exactly how to shop. One Reddit thread on this need carries 269 answers. No merchant is standing in front of any of it.

“What do I need for a newborn in summer” returns zero commerce, and People Also Ask on that page literally reads “What to buy for a baby born in summer?” The word “buy” is on the page. The buying is not.

The rest of the vacuum list: ski trip, host thanksgiving, start a garden, home office setup, snowstorm, job interview, moving house, picnic, funeral, cross country move, summer wedding guest, hiking in summer, beach camping trip, winter hike.

Those are not obscure needs. They are the moments when a household buys eight things at once.

This is not a merchant failure

You did nothing wrong. Search changed underneath you.

Current taxonomies store what a product is for as a single word. Mountain. Road. Recreational. Sea. One word, chosen from a short list, standing in for every purpose that product could serve.

A situation is not a word. It is a sentence with a person in it, and usually a dozen products. There is no word for “first apartment” and none for “flying with a toddler,” because those are not properties of any single product. A generator has a published wattage, so shopping carries a power output facet and the buyer gets an answer. A dorm room has no spec sheet and no word, so the retrieval layer has nothing to match on and returns no commerce at all.

The buyer typed “what do I wear apres ski.” The AI answered with insulated winter boots, a fleece, and a puffy vest. Your listing for that boot carries a brand, a model, a material, and a size, because that is what a product title has to do. The right product. Zero overlap.

Buyers search in one language. Products are listed in another.

The market already found a workaround, and it is a kit

Eleven individual non-apparel products turned up across all thirty queries. Eight of the eleven are pre-bundled kits:

• Sirius Survival Pre-Packed Hurricane Kit
• Newborn Basics Kit
• Starter Pack #24
• Fridababy Baby Basics Kit
• Hape Beach Basics
• FUN LITTLE TOYS Kids Camping Set
• Ozark Trail Saltwater Kit
• KastKing SteelStream Fishing Tool Kit

When a situation query does surface a real product, the market answers with a box.

The reason is structural. Retrieval cannot assemble a basket. It can match one query against one SKU. So a merchant pre-assembles the basket into a single SKU that attribute matching can actually retrieve, and that SKU becomes the only thing on the page. The kit is a findability decision, not a product decision. Those merchants did not conclude that buyers want a box. They found the one shape of inventory that a situation query can reach.

If you sell those items loose, the basket on that query is still yours to assemble.

When commerce does appear, it collapses to one category

Commerce appearing is not the same as the need being served.

“What do I wear to a black tie wedding” returned nine products. Every one of them is a dress. Black tie for men means a tuxedo. Zero tuxedos on the page.

The need is named in full. One category is sold against it. That is the same shape as the vacuum, just with a storefront attached.

What a merchant can actually do

There are three options for serving a situation query today, and only three.

Bundle the basket into one SKU, the way eight of those eleven products did, so a single retrievable product covers the whole moment.

Write the buying guide that assembles the basket, the one Vogue and the mommy blogs are writing now, so the page that owns the moment is yours. On these thirty queries, editorial owns the ground almost everywhere. That ground is unoccupied by anyone who sells the goods.

Or leave the ground to someone else.

Nothing in the retrieval layer will assemble the basket for you. There is no facet coming. There is no algorithm update that turns “what do I need for a college dorm room” into a shopping grid built from your loose inventory. The assembly is your job now. The merchants who worked that out are holding the few product slots that exist across thirty queries.

The situation language exists. Buyers type it every day, and the AI Overview enumerates it item by item. Your products exist. The connection between them is the piece nobody has built yet, and you can build it. Open the graph and see the intersection: the situation words buyers actually use, resolved against the products you already have in stock.

Methodology

n=30 situation queries, captured 2026-08-04 by direct inspection of live Google search results, query by query. No SERP vendor was used. Commercial-module presence was detected by element inspection of each results page, a method validated before use against a known positive page and a known negative page to confirm it discriminates correctly. A query was scored a commercial vacuum only if the page carried no sponsored results, no shopping grid, and no product module of any kind. Sponsored listings, shopping grids, and related-product modules count as commercial. Editorial and forum results do not.

Queries were limited to the situation form, meaning a buyer describing a circumstance rather than naming a product with a qualifier. Results from other query forms are reported separately and are not included in the 63 percent figure.

Conditions, stated because they bear directly on a finding about module absence. All captures used one browser profile on a United States connection, signed in to a personal Google account, so results carry whatever personalization that account attracts. A shopping browser extension was active and injected product panels into several pages; those injected modules are not Google results and were excluded from every count here, including from the vacuum scoring.

This is a dated snapshot, not a permanent property of search. Several of the products found in this study sat inside the AI Overview rather than in a separate shopping module. That was observed on the page, not announced by anyone, and it suggests the placement of commerce on these queries is changing. The 63 percent figure has a short shelf life and should be re-measured.