People continue saying Bezos is secretive, and I continue to contend that he simply hides his secrets in places where everyone can find them...but leaves the thinking part (to understand the secrets) up to them. The result? People continue saying Bezos is secretive.
Amazon provides a lot of information about how it works inside of its SEC filings. If you're interested in understanding what the business is doing over the long haul, the filings are very informative. If you're just eager to get a scoop on the next piece of technology, next partnership, or next earnings results...you're going to be disappointed. Bezos continues to be very hush on specifics.
Here's a piece included in the 2011 10-K that provides a high level view into the strategy of running Amazon.
We seek to reduce our variable costs per unit and work to leverage our fixed costs...Our fixed costs include the costs necessary to run our technology infrastructure and AWS; to build, enhance, and add features to our websites, our Kindle devices, and digital offerings; and to build and optimize our fulfillment centers. Variable costs generally change directly with sales volume, while fixed costs generally increase depending on the timing of capacity needs, geographic expansion, category expansion, and other factors. To decrease our variable costs on a per unit basis and enable us to lower prices for customers, we seek to increase our direct sourcing, increase discounts available to us from suppliers, and reduce defects in our processes. To minimize growth in fixed costs, we seek to improve process efficiencies and maintain a lean culture.
Very dry stuff, right? Yes, but in the appropriate context, it's incredibly meaningful. The first thing to understand - points driven home by Bezos on any interview that includes the topic of Amazon's business model - is that ecommerce is a scale business. (See what he said here in a 2001 Charlie Rose interview.) Businesses operating on small- or medium-scale cannot compete against those operating on a large-scale. Scale comes into play with the size and complexity of the software, the purchasing power of the business, the distribution capabilities (among other factors). These are fixed costs.
Growth is a tension between timing your increase in fixed costs (in a way that is necessarily messy as you increase headcount, expand your fulfillment centers, improve your software, etc.) with the additional profits those investments should bring. Proponents of Economic Value Add (EVA) insist that growth must show a quick return by way of earnings boost that demonstrates higher value added than the cost of the capital invested to generate it.
It makes me think of Yogi Berra's "In theory there's no difference between theory and practice. In practice there is."
Messy growth requires extended time frames. That's not to let management off the hook. They still must make good decisions with allocating capital. But it would be plain silly to avoid investments that take several years to develop and mature if, at the end of the investment period, they provide strong returns and are protected by some sort of competitive advantage. That's the whole idea of having a franchise.
The interesting part of this comment in the 10-K is that it tells anyone willing to pay attention, exactly what happens as Amazon starts cleaning up the messy part of its growth. When it grows into a new geography or with a new category of products, Amazon does it with a mind to win customers. It builds selection quickly. It prices the products competitively. It uses sheer effort to make up for what it lacks in systems. It markets more heavily than usual, rewarding its affiliates with a higher percentage of sales for referrals. All of this to build the customer base quickly and help them create the habit of buying those specific products from Amazon. This is the messy part.
Then, as the main thrust of the growth subsides, Amazon circles back and cleans up the mess. Among the most important things it does is cozies up to product suppliers. Where it has been going through middlemen to source new products, it hammers out deals directly with the manufacturer to cut its price. Where it has been buying in small lots as it attempts to learn its customer's demand for products, it aggregates its purchases and demands better pricing for the higher volume. And then it just reduces defects.
When we look at the Amazon numbers, we must ask: what is hidden in the expenses? Amazon is that rare business that has long demonstrated an ability to grow customer demand as quickly as it expands its own capacity to service that demand. And if that growth is messy and costly (showing up in the expense category, thereby reducing earnings; or showing up on the balance sheet, thereby increasing invested capital)...just how profitable would this business be if it slowed down its growth?
I don't know the answer. It's probably impossible (even for Bezos) to come up with a precise response. But it's fair to say owner earnings are much higher than reported earnings.
This is why Amazon trades at such a high multiple to its earnings...a good chunk of its expenses are investments in growth. These value of these investments will compound with time, and those earnings will grow as a result.