So GM wants to avoid becoming the next Circuit City, insofar as not being wiped out by the online sales experience. Dodge the bullet (excuse the pun) of massive layoffs, shareholder wipeout and supply chain trauma. Might be a little late for that perhaps? Well, necessity is the mother of invention, and GM’s decision to partner with eBay to sell new cars has the potential to be a seismic change in the automotive customer experience.

As we have written before, online sales channels have considerable experience advantages over that of bricks and mortar operations. Based on our most recent benchmarks with US consumers (taken, mind you, in December last year in the midst of the economic meltdown), even average performers in online shopping garner NPS® of about 50%, while top-scoring companies achieved scores of 70% or higher. The National Retail Federation rates Zappos, Amazon and Overstock as three of the top four retail experiences (LL Bean keeps physical retail of any kind in the top 4) and yes, we know that the top 3 will be the top 2 very soon. Their advantage springs from several sources. Supply chain efficiencies drive better fulfillment performance at lower inventory levels (the law of large numbers apply). Better data about their customer’s behavior enable them to customize offerings and merchandize more effectively. Oh, and considerably less dependence on "the carbon element" – people, although Zappos has turned that into an asset.

Nothing new to report here. However, the automotive industry had always faced the dilemma of channel conflict. Their customer experience, to a very significant degree, lay in the hands of their channel – the dealers – and their channel was pretty resistant to focusing on customer experience. It didn’t help that JD Powers had built a lock on the industry notion of customer satisfaction which, although a terrific marketing tool, didn’t always line up operational dealer excellence with measured outcomes. Oh, and dealer satisfaction measurements have been the "poster child" for gaming results.

So why not do away with them altogether? Early online efforts around car sales were clearly a compromise. Last year, I was scouting dealers for a new car purchase, without much pricing relief here in the Bay Area (what a difference a year makes). I tried getting a quote from Yahoo Cars, only to find I was referred to – guess who – the same local dealers with the same local prices and inventory. Why couldn't I get a quote from other dealers? Not going to happen, this was a system constructed to preserve a local monopoly. Problem was, sooner or later I found an online service that brokered a better deal with a dealer in LA – who promptly shipped the car right up to my door.

GM’s experiment with EBay leads in only one of two directions. Either the design of the program continues to protect dealers from regional or national competition, in which case people won’t bother to use it, or they will finally open up the door to a national or regional marketplace. If they do – and I hope they do -  the consequences will be profound for the industry. If I can easily get a quote from a dealer in Denver, and shipping costs remain reasonable, it’s possible that premiums in distribution will pretty much evaporate for dealers nationally. This is likely to drive consolidation and efficiency, and a superior customer experience. An easy purchase experience, with a competitive price will improve everyone’s dealer experience, but the dealers will transform into service centers which do need local presence (and are not a bad business to boot). Other brands are likely to follow or face market share loss, assuming other factors (making cars people want to buy) are neutralized.

What about test drives? Clearly autos are product that rarely get purchased “sight unseen” (although in parts of Europe that’s sometimes the case). Firms like GM will need some form of showroom to display their product, but won’t need the expensive immediacy of inventory that bogs down their supply chain. Of course, none of this happens overnight, many car buyers will take time to transition their buying habits. But once the genie is out of the bottle, as bookstores discovered, the outcome has inevitability about it. Better customer experience combined with superior economics win out in the end.

Will GM win this race? Distribution experience is not the only factor, by a long way, in purchasing your car. Nevertheless, when you have little to lose, reinventing the customer experience could prove the most cost effective of the transformative initiatives the company has to take on to thrive in the long run.

How can eBay put you into a GM next week?

Over the weekend, my family and I dropped in to a big box electronics retailer to pick up a copy of the Wall-E DVD. After a fruitless search of the shelves, we finally tracked down an employee who advised us that their system suggested they had two copies in inventory, but he had no idea where they were. Frustrated, we picked up another item we needed and checked out. Nobody asked us if our shopping experience had been a good one. I went home and bought the DVD on Amazon, and therein lies the story of an existential crisis in retailing.  By the way, this wasn't a story of Circuit City, who on Friday announced that they had lost their fight to remain solvent and would liquidate - eliminating 34,000 jobs in the process. In an ironic twist, circuitcity.com is now pointing to a URL with the word "closed" in the title and showing only a letter to customers, not the usual commerce site. It was, after all, the internet that put them out of business.

Back in the early days of the internet, plenty of Silicon Valley types (I include myself in that characterization) predicted the disruption or destruction of numerous industries. Just about every type of business would be turned on its head, we heard. Ten or so years later, it’s coming true. Well, at least in some places.

Sure, the newspaper business has been pretty much pulverized by the internet. But that’s small potatoes compared with the impact on bricks-and-mortar retailing. Over the next 10 years, large sections of the retail industry will face their own “ultimate question”– do we need you at all?

Retailing provides a basic value creating function: distribution – getting products into people’s hands. It’s about breaking bulk.  The problem stems in no small part from the fact that for a significant number of products, online sales has significant advantages in distribution and can neutralize other advantages that physical retailers have in merchandizing or immediacy. Sure, you can’t tell what the fabric feels like on that shirt over the web, but you can read a lot of customer reviews and get detailed specifications from your kitchen table.

Furthermore, retail has its own supply and demand problem. America has been on a building spree, based on a fundamental misjudgment about both the volume and nature of future commerce. We have built malls and we have built retail banks. Everywhere. You know that we are in trouble when cities are forced to regulate the frequency of high street bank branches, Starbucks coffee shops and big box stores.

As I have suggested in the past, a recession has a habit of sorting out winners and losers, and retail has another issue. In too many cases they don’t care about the customer experience, don’t know about it, or both.  The National Retail Federation's survey of customer service puts L.L.Bean at the top, as they have been the last few years. But #2, #3 and #4 are internet-only retailers. And this understates the trend.

Online retailers are rich with behavioral data. They know who you are, what you bought, perhaps even what you liked. They can get feedback from you – both in the form of NPS but also through behavioral data such as when you abandon your purchase.

Retail is data poor. In a worst case scenario, someone walks into a store, pays cash for a product and walks out. The retailer knows absolutely nothing about them.

This data paucity spills over into NPS. Most retailers today simply don’t know if their customers had a good experience or not, and they certainly don’t have the mechanism for taking immediate action around it. You also have to seriously question their intent. Try volunteering information to an employee at your local clothing chain and see what he or she does with it. Is there evidence that they are trained to handle customer feedback? And that’s the easy part. If ONLY customers truly volunteered information. What retailers really need is an ability to measure each transaction.

You might be thinking this is an implementation problem. Sure, multi-branch operations have their unique issues but let’s not forget that Enterprise – that icon of Net Promoter cast studies – solved the problem for frontline employees in car rental.  You have the ask yourself, is it really that hard? Apple has clearly cracked the code in their Stores, partly through a focus on building data about customers and not simply viewing a sale as an anonymous transaction. Verizon Wireless makes it work. Hint: keep your eyes on our NPS benchmark to see how that worked out for them. OK, so theirs is a technology product, but their customers cover some pretty broad demographics – why wouldn’t this work in other operations?

Worse supply chain economics coupled with poor customer experience. If I could buy my latte over the internet, I’d be worried that we could see the end of the high street altogether.