Evan Schuman's StorefrontBacktalk

The retail industry is well known for its contrarian tendencies. After all, “be where everyone else is not” was pretty much the original strategy of Wal-Mart when it was setting up shop in rural communities that chains of the day wouldn’t go near. This concept is at the heart of the gold mine of information that today lies just out of reach, contained in the local inventories of millions of shops across the country. No other current retail or e-tail effort—with the possible exception of mobile—has anywhere near the potential of local inventory search.

But as has been reported, today’s efforts—by companies including Google, TheFind and Milo—are horribly inaccurate when it comes to searches looking for the specific products being sought. To be clear, even the early drafts of these engines find a lot of products. But they’re only finding products that already can be found easily by existing engines.

No kudos are merited for finding big-ticket televisions or couches that could be found just as easily via Bing or Yahoo or even Google’s regular search. If the consumer already knows the manufacturer’s name—or a specific model and make—the manufacturer’s site is generally quite helpful. Certainly if the store is already known, that store’s site can deliver those answers.

The value—and extreme value it is—comes from an engine finding products that simply cannot be found otherwise. For example, a 5-year-old receiver’s volume knob falls off two hours before an important show or sporting event. “I need to find somebody nearby that has that part in stock, probably in their storeroom covered in cobwebs,” the consumer would think. “There are dozens of independent TV and audio repair shops within 30 minutes of me. Is there an easy way for me to search and see if any of them have this item in stock?”

Similarly, a plate falls and smashes on the morning of a big dinner party. The consumer needs to find any rose-patterned plate that is the same size and looks somewhat like the broken one. The consumer doesn’t care about the brand—unless it happens to be an exact match. The only concern is which is the closest shop with that plate in stock. Or, in an example we’ve cited before, a parent needs to find not merely a stuffed animal but a very specific kind, such as a rhinoceros. In all the stuffed animal bins in every store that sells fuzzy critters and is within an hour’s drive, the consumer needs the search engine to reveal the closest shops with a rhino.

To make this product location happen, serious changes have to be made on multiple levels. Suppliers need to label and code their items much more specifically. Retailers—especially smaller chains and one-location merchants—need to program their POSs much more specifically and then support the sharing of that data. The search engines must be programmed for a different kind of search, one where specific searches and distance-sensitive results are key.

The search approach is arguably the biggest change. Today’s engines can handle specific and general searches but nothing in between. A search for a “Panasonic Viera VT25 Plasma 3D HDTV” will likely deliver very satisfactory results. And if someone goes more general and asks for any HDTVs, those results will also likely work. But if the desired item is a .75mm drillbit, an engine that answers such a query, provides accurate inventory data and comes up with some obscure hardware store four minutes away, now that’s a game-changer.

We say game-changer because such an engine has the empowering potential to reward some of the shops that choose to have a large inventory of hard-to-find items. In a community-supporting way, it can point customers to local retailers and it will promote results solely by proximity. These kind of changes get us to what eBay accomplished in its early days, where it allowed someone seeking an obscure item to find someone who happened to have that item.

It’s a matter of thinking small. It’s easy to find a “tie clasp,” but what if the search is for a “tie clasp shaped like a barnyard animal” or “an earring in the shape of a triangle”?

It’s going to take years for true local-inventory-level searches—ones that are truly different and useful—to materialize. But they will make quite a noise when they do. Today, though, we’re not even close. (That’s a challenge, by the way. If someone out there can create such an engine that works at this level, we’d be thrilled to test it out.)