Wednesday, August 15, 2012

All Convenience Infrastructures Not Created Equal

This is the second post in a series about deconstructing Amazon's Feedback Loop, an attempt to understand how its components work both as individual units and together as a collective system. See also the previous post, Convenience (and Diaper Stench): Amazon v. Walmart.

All retailers must invest in a convenience infrastructure, plowing cash into those components of their businesses that make it easier for consumers to shop with them. 

But not all convenience infrastructures are created equal. 

For traditional retailers like Walmart, that infrastructure is composed overwhelmingly of stores. And to enhance the convenience it offers customers it must build more and more stores plus staff them, stock them, and maintain them to some reasonable aesthetic and hygienic standard. While this has been a lucrative business for Walmart investors over the years, having an infrastructure rooted in real estate – an inflating asset whose costs increase over time – is pricey. Especially when compared to the alternative. 

For web-based retailers like Amazon, convenience is a much different proposition. It’s driven by technology (software, hardware, internet connectivity, etc.) and the speed with which it can deliver packages to shoppers. 

To build on our Amazon Feedback Loop schematic, here’s the convenience infrastructure addition: 

With such a chunk of its convenience being based on technology, Amazon has a tremendous cost advantage over retailers that are forced to plow so much cash into real estate in order to grow. As we’ll discuss in this article, a convenience infrastructure that depends on technology investment is inherently less expensive and much, much more scalable. 

Convenience through Technology 

We outlined in the previous article how web retailers (and Amazon specifically) use technology to enhance the convenience of their services to shoppers. In my close call with being overwhelmed by diaper stench, I needed bin liners quickly to stave wafting odors from my daughter’s nursery. Amazon made the process of buying them convenient, using technology to: 

1. Provide quick ACCESS to Amazon’s website through various internet-enabled devices. 

I just grabbed the Kindle Fire, pressed the “on” button, slid the Android hibernate bar across the screen, and clicked on the Amazon shopping app icon. This all took about three seconds. 

Amazon could just as easily (and almost as quickly) grant me access through my laptop, iTouch, smartphone, and a host of other devices that connect to the web. Therein lies its commitment to providing the easiest access to its products by taking advantage of any one of the host of rapidly proliferating devices that connect to an ever-faster moving internet. 

A few years ago, the only option would be booting up a computer, clicking on a browser, and typing in the Amazon URL. Which is pretty fast, too, but Amazon wants access to its shopping experience to move as quickly as the fastest device available. And it has spent years investing in that convenience factor. 

To that point, Amazon’s commitment to providing access through a wide variety of internet-connected devices is nothing new. Anyone remember the Palm VII? (Um, for that matter, I should ask whether anyone even remembers Palm now.) It was a digital handheld organizer, a clunk of gray plastic – in terms of design aesthetic, decidedly unsexy – with Palm’s calendar and contact features. This particular model happened to have a flimsy antenna that, when flipped upright, provided spotty mobile access to the web using an even spottier browser. Well, way back in 1999, Amazon was stretching its innovation muscles with a service called “Amazon Anywhere.” (You can read the Amazon press release about it here.) You could actually log-on to Amazon and place a mobile order more than thirteen years ago! 

Amazon has long been prepared for this concept of shopping its stores using apps on mobile devices. 

2. Quickly SEARCH Amazon’s wide catalog for the specific product I needed. 

I knew I needed those specific diaper bin liners, and all I had to do was type the brand name into the search box and hit “Go.” Even with a product catalog that easily tops many millions of items, Amazon served back the option I wanted with sub-second speed. 

One might assume that whether the search results come back in one second or a half-second wouldn’t affect the shopping experience too much either way. Amazon disagrees. The company has invested countless resources organizing its catalog, streamlining its databases, and increasing its server processing power for the sole purpose of shaving milliseconds off your search. It deems search speed that important a convenience factor. 

In 2004, Amazon showed the world how serious it was about investing in heavy duty search capabilities. It took all the algorithms it had built for searching on its site, and offered a service for searching the full web. A9, as it was called, was among Amazon’s first attempts to spin-out its internal innovations for use by wider audiences. It wanted to test whether that market so dominated by Google and Yahoo! was open to an alternative. It wasn’t successful outside of Amazon, but those investments continue to reap benefits by enhancing the convenience factor for customers searching for products on (You can read more about the A9 launch here.)  

3. CHECK-OUT quickly and easily. 

When Amazon sent my Kindle Fire several months ago, the company did me the favor of pre-connecting it to my account. I take it this adds layers of complexity to the various steps of prepping and shipping each of these devices to customers. And while I won’t suggest Amazon is benevolent for choosing to do this, it sure made my life easier. More importantly for them, it reduced the likelihood that I (or any Kindle Fire buyer) would be too lazy (or so lacking in technical skill) to make that connection myself. 

In buying those liners, it made the check-out process quick and easy. Once the product was in my cart, I had maybe two additional clicks until the transaction was complete. No delays, no extra steps, and therefore fewer chances for me to change my mind. 

That’s present day e-commerce shopping. Let your mind wander back nearly 20 years to the dawn of web retailing. While internet usage was famously growing at the breakneck annualized rate of 2300 percent, it was far from clear that it would be a medium consumers would trust for shopping. At that point, shoppers guarded their credit cards as if they were cash. I remember traditional stores often required an ID for a credit card purchase, verifying the identity of the buyer each time out of fear of penalty from Visa or American Express if fraud occurred on its watch. This paranoia with person-to-person transactions was amplified when making a catalog purchase over the phone. It was not at all clear that consumers or card issuers were going to be comfortable with the risk of punching their credit information into a keyboard, transmitting them across a dial-up modem into the great unknown of the internet. 

What nefarious agents might be lurking in the web’s shadows, eager and ready to nab your credit card digits and run up a big bill on your tab? 

In this brave new world, Amazon managed to convince shoppers to store their credit information on its servers, to link it to their usernames, and to keep a shipping address on file. Amazon called it the “1-Click” process, launching it in 1997 and patenting it in 1999. (Reference Amazon's press release about it here.) All so Amazon could help them check out more quickly, thereby increasing the convenience factor and losing fewer sales to the dreaded abandoned cart. 

How Moore’s Law Makes Amazon More Convenient for Less Money 

Despite all these investments in technology improvements to make the shopping experience more convenient, the very nature of technology as a driver of convenience (and hence a driver of growth) makes the process of growing much, much less expensive. 

Convenience is different for web retailers than it is for traditional stores. As we’ve discussed, stores rely so heavily on location to customers as their primary means of being “convenient.” Real estate is inherently expensive, its price tag tends to expand with time, and each new store brings with it the need to constantly stock it, staff it, and maintain it. 

Not so on the internet. 

For web retailers, convenience is more a matter of the factors we explored above. How quick and how easy it is to access the web store? How quick and how easy is it to search for the product the shopper wants? And how quick and how easy is it to check-out? 

Each of those is a function of technology, and herein lies an advantage for web-retailers over those operating out of stores. While traditional retailers are pouring cash into real estate as they push the convenience growth lever and seek more shoppers, web-retailers are enhancing their convenience factors by investing in technology. 

To that point, Jeff Bezos sat down with Charlie Rose in 2001 and had this insight to share: (You can watch the interview here.) 

One of the things that’s totally different about e-commerce from physical world commerce is that real estate doesn’t obey Moore’s Law. Moore’s Law says that microprocessor performance doubles for the same price point every 18 months. That’s held true for more than a decade. What you’re finding now is disk space is getting twice as cheap every 12 months. And bandwidth is getting twice as cheap every nine months. So if you take the bandwidth doubling rate of nine months and assume it holds constant for the next five years, that means that we can spend the same amount of money on bandwidth per customer that we spend today five years from now but use 60 times as much bandwidth. That’s a big deal!* 
As microprocessor speed doubles every 18 months, it powers the Amazon technology for even easier product searches and faster check-outs. As disk space is getting twice as cheap every 12 months, Amazon can provide more rich content supporting its products and still help customers search through all the information quickly. And as bandwidth is getting twice as cheap every nine months, Amazon is ensured that more customers get online and get easy access to its website. No matter if they’re at home, work, or out about, the internet is nearly ubiquitous and fewer shoppers are ever without some sort of device that connects them to the web. 

To harness technology, Amazon must invest in software developers, database designers, system architects, and the like. These professionals are expensive. But the work they do is scalable. A single piece of well-written code can perform its function for all of Amazon’s 200 million customers with the same amount of effort and investment as it could to do the same thing for one customer. That makes the first customer very expensive to Amazon, but the additional 199 million quite cheap. 

Contrast that to traditional retailers. While Walmart’s hourly workers at each store might be inexpensive on an individual basis, the company requires a lot of them in order to serve customers. And a single worker can only help so many customers in any given period of time. Unlike that piece of code sitting on an Amazon server, that worker is decidedly un-scalable. 

It’s a big deal, indeed, when an asset that helps drive growth depreciates in cost over time (technology) rather than appreciates (real estate). To reach more customers, retailers like Walmart must constantly build new stores. The cost of which only grows over time. For Amazon to reach more customers, it must only make sure the bandwidth is sufficient, server processor speeds fast enough, and disk storage space deep enough to handle the exchange of data. That particular cost of growth is significantly lower for a web-based retailer. 

But Wait, Says Walmart, We Have an Ace up Our Sleeve…the Convenience Barrier 

In my experience shopping for diaper-bin liners, I mentioned one convenience advantage that Walmart held over Amazon: If I wanted those liners immediately, if I couldn’t postpone the gratification of holding my new purchase in my hands immediately, then Walmart would have won my business. 

This is the convenience barrier, and it has been the most important piece of protection the traditional retailers have had to keep Amazon and its ilk at bay. The need for immediate gratification – to get what you want now versus waiting three to five business days – is a big deal to shoppers. 

Amazon, however, has counted delivery time as part of its convenience infrastructure, investing heavily in it over the past several years. In the next article we’ll discuss how these investments have whittled away at Walmart’s convenience barrier advantage and what that might mean for the future of both companies. 

* For an interesting read on Moore’s Law, see the article Was Moore’s Law Inevitable?