Words to the wise for data-crazy online marketers: Databases – the operational backbone of e-business – are great tools for simultaneously structuring and storing sets of information and describing the relationships among those sets in almost infinitely complex ways. But they don’t define a representation of an item or object in a conceptual manner that matches a consumer’s desires.
That might seem obvious at first glance, but most e-commerce sites today treat the database as the alpha and omega of their existence: The data they present to users are week-old mussels slathered in light oil, gently tossed, and presented as bouillabaisse.
Almost no online stores, and certainly none of the major brand names, leverage the power of the information relationships contained in their product databases in a way that benefits the consumer, or even reflects the manner in which consumers make choices.
What’s to blame? That friendly metric of marketing science, the stock-keeping unit, or SKU. Whether it’s a UPC code, an ISBN, or an internally created number or code databases use to store product information, most e-commerce databases are still highly SKU-centric. For generating invoices, looking at aggregate patterns of customer orders, or paying sales tax by state, an SKU-focused database is fine. SKUs reside in one table of a relational database, or perhaps in an entirely separate product database. (The customer order database, for example, has tables for orders and customer identities, among other things.) So finding and identifying your niche market is key!
Individual consumers, however, generally do not visualize products they wish to purchase as SKUs. They think of the object of desire: something that holds some set of connotations in their mind. It’s the rare consumer who can recite an ISBN or other product code. They might be able to remember “Van Camp’s Pork and Beans” because of expensive branding and marketing, but that’s the extent to which we can rely on a consumer’s individuation of the object of desire: the item they want to purchase.
There’s a concept known as perfect knowledge: the notion that consumers could be presented with a level playing field on which they could easily compare the price and value of an item uniformly. In the case of consumer choice, I like to think of “perfect representation”: how information could relate to other information in a commerce milieu where the customer’s internal representation of the object of desire (a book, a spatula, a marked-down bike seat, a bottle of ibuprofen) would match the method by which the site offers up that information. The corollary: “perfect relationship,” where the up-sell, cross-sell, or even down-sell are built into that object’s representation as part of the object’s “self-knowledge” (or even better, “sell knowledge”).
Unfortunately, the structure of your typical e-business’ database-the manner in which it’s fielded and keyed-currently limits the creative thinking that turns listings of products into an e-commerce site. “Perfect representation” of a book would provide a structure for finding that book and getting more information about it in the manner in which people think of a book.
Say you want to purchase Edith Wharton’s The House of Mirth. You probably do not visualize the physical characteristics of a specific edition of the work. Instead, you think of the work itself; the more knowledge you have of the work, the more likely you will drill down to a more specific conceptualization of it, which might include a hardcover edition or a release with critical commentary. More sophisticated bibliophiles will drill down to a particular edition. They’re relying on booksellers’ descriptions of items as listed on Bibliofind and elsewhere.
Got all that? Now compare these notions with the method in which all online bookstores present information about books: sucked directly out of their ISBN-oriented databases without any thought to building a relationship around the work, the author, or even the physical characteristics. Everything attached to the display of search results or a specific book page at these stores is an embellishment to the all-powerful ISBN.
A few months ago, I set up ISBN.nu, a Website that does ISBN-based price comparisons among several online bookstores. Included is a page on the novel Ethan Frome that I created to show how I think information about a specific work could be presented, so a user would better understand it. Check out also this post on niche marketing with stadium cups.
The point? If you create a hierarchy of the most abstract to the least abstract constituents of a book, like a set of interlocking spheres, you start with the work itself. You could then catalog physical characteristics at the next layer: hardcover, paperback, hand-bound, and so on. Next, you have a specific edition of the title, such as any printing of the book in hardcover from a certain press; this is barely distinguishable from individual printings of that edition, though it is a distinction.
Finally, you could distinguish between actual books; one copy of Moby Dick locked in a cabinet in Portland, Ore., is a truly different animal from another locked in a drawer in Portland, Maine.
Online marketing efforts like these – to provide some semblance of related product information – are portentous for e-commerce. They’re an inkling of things to come. Eventually, many bookstores will take the big step of reorganizing themselves to help people find what they want rather than just a book. For now, keeping the fire stoked in the current system is a full-time job for dozens or hundreds of people at every consumer product-driven Website.
Technology-wise, there’s nothing keeping any product store from building relationships, flat or hierarchical, between their products.
As consumers start to demand more or, better yet, respond to more than just arbitrage in comparison shopping-“Quick, buy a PalmPilot for $79! The store made a mistake!”-the kind of “object-knowledge” I’m describing could make a store stand out.
A store that provides product information and categorization as a function of how users conceptualize what they want, or the many ways in which that visualization takes place, might actually be the place that helps you find that song running through your head, the book you actually want to buy, or all the mystery novels Rex Stout ever wrote. It might be that this notion separates the future has-beens from the future success stories.