Tuesday, September 26, 2017

Fighting Gresham's law of journalism -- more "yes, it is just a god damn vending machine" blogging

Just to review, a few days ago, there was a great deal of fanfare around an article by Elizabeth Segran from Fast Company magazine entitled:

Two Ex-Googlers Want To Make Bodegas And Mom-And-Pop Corner Stores Obsolete

At the risk of belaboring the obvious, the article was a bad piece of business and technology journalism. It credulously accepted what should have been incredible claims from an entrepreneur with an enormous interest in hyping the story. The response was largely divided between writers who were understandably offended by the cultural insensitivity and those who unquestioningly accepted the idea that a line of low functionality vending machines (no refrigeration, no hot food or beverages, not set up for cash transactions) presented an immediate threat to convenience stores and other small retail outlets. More than a few commentators managed to fall into both camps.

The problem here is not that all of the coverage of the Bodega Vending Machine Co. was bad; it is that the bad got most of the oxygen. This is primarily a business story (the technological aspect is trivial), and it has produced some excellent business writing, but it appears that the coverage is another example of Gresham's law of journalism: the crappy crowds out the good.

The best of the clear eyed analyses probably comes from  Helen Rosner, a smart, knowledgeable writer who explained in crushing detail the major flaws in the Bodega business plan.
Bodega’s product is, fundamentally, a vending machine. (Well, maybe it’s a mini-bar — open access to product, in fancy places, with a presumed audience that’s affluent and design-minded.) Vending machines are a unique form of commerce, mostly defined by the lack of human interaction at the point of transaction. This kind of unmanned retail operation has a long history (the vending history timeline on the website of NAMA, the National Automatic Merchandising Association, kicks off with Egypt’s Ptolemaic dynasty; it is delightful), with efficiency as its primary appeal. A tiny, self-contained store without an employee saves all sorts of overhead: Less required real estate, lower payroll, shockingly greater likelihood that shoplifters will be crushed to death.

These efficiencies aren’t gravy, though; they’re essential. They work by way of a simple economy of scale: If you run a few dozen machines (or a few thousand), it becomes possible to buy your products at a discount, to warehouse the products more effectively, and to both fill and repair your vending units in a more streamlined way. These businesses live or die by logistics.

This is where things seem likely to fall apart for Bodega. Even with their wifi connections and app-connected camera sensors, the units themselves are still just offering consumers a basic model of unmanned commerce — only with smaller, fancier machines to process the transactions. What Bodega does offer as a differentiator are the number of unique products per unit (100, the average vending machine has 20-40), and the promise that the products will not just be tailored to their general environments — protein bars in the gym, tampons in a sorority house — but to their specific users. A promise of “machine learning” will, as Fast Company explains, “constantly reassess the 100 most-needed items in that community.”

At 100,000 units — the scale McDonald and Rajan envision — that’s ten million items that are active at a time, plus reserve products for restocking, plus new products to introduce as the “machine learning” (I’m sorry, I just can’t) cycles out low performers. Across specialized markets and user-informed preferences, the number of SKUs (industry shorthand for a stocked product, rather than an individual unit of that product) that Bodega would be dealing with would quickly climb into the thousands.
Labor is not a minor issue, with a company like this one. “Unmanned retail” isn’t a precisely accurate phrase: There may not be a person ringing up your transaction, but there are plenty of people working to maintain a system that allows that absence — even the famous midcentury automats were just the outward storefront of a working, fully-staffed kitchen. Bodega’s warehouses will need to be staffed. The trucks will need to be driven. The Bodegas themselves will need to be manually restocked — each can, bottle, and box placed one by one onto each unit’s shelves. Many traditional vending machine companies employ restockers who double as machine repairers. Will a Bodega restocker be trained to fix a busted computer-vision camera?

Rosner closes with a wonderfully pithy and honest summary of what is probably Bodega's real business plan.
In Silicon Valley, Bodega’s success will not be measured by how well it truly replaces the stores it wants to eliminate — by how many lives it makes better, how many jobs it creates, how many communities it strengthens, or how many families it serves. Like most startups, its success will depend on whether its founders and investors make money, either by cashing in with an IPO or selling to a bigger company for a tidy profit.

1 comment:

  1. From the linked article I found this juicy quote from someone hyping this business plan: "Curious why Bodega is eliciting such strong negative reactions? If it works, you’ve proven it’s an audacious idea."

    Wow. Just wow.