Start of Main Content


Second-year student Rohan Rajiv is blogging once a week about important lessons he is learning at Kellogg. Read more of his posts here.

Amazon opened its first bookstore in Seattle earlier this week. This led to a many interesting questions in the media: Has Amazon taken a step backward by jumping back into traditional retail? Didn’t Amazon start an online store to improve on the traditional bookstore model?

To understand this, let’s begin by taking a walk down memory lane and look at Jeff Bezos’ initial rationale for starting an online bookstore.

Bezos understood a fundamental benefit of having an online store — “unlimited” access to inventory while also eliminating the fixed cost of owning a physical location. Why did this matter? For a category such as books, there are millions of published works. However, even the largest of bookstores can possibly only stock tens of thousands of books. So, within bookstores, you now need to forecast/guess demand for books. And that, inevitably means high inventory costs because estimates are rarely right – especially for niche category books.

So it is now easy to understand why Bezos narrowed in on the following five categories as possible areas for Amazon to focus on for its initial product:

  • compact discs
  • computer hardware
  • computer software
  • videos
  • books

All of these categories have a “long tail” of niche products that make selling them via traditional retail very challenging and expensive.

The principle we’re getting at is that the characteristics of a product drive the ideal supply chain/distribution strategy.

Let’s imagine two kinds of products:

  1. Low demand uncertainty, low value products.
    Examples of such products are daily groceries or toilet paper. These have consistent demand and low value. So it makes sense to make these available near customers as the cost of shipping these products from a centralized warehouse is probably going to exceed the cost of these products. Besides, we’re not going to lose money on wasted stock since it is fairly straightforward to predict their near-constant demand.
  2. High demand uncertainty, high value products.
    A great example of these are diamonds. It is very expensive to carry diamond inventory. So shipping them from a centralized warehouse makes a lot of sense since the shipping costs are small relative to the value of the diamond.


This, in turn, leads to the next natural step in the logic:

  1. It is very expensive for online retailers like Amazon to ship low uncertainty, low value products like diapers and toilet paper. So they should only do so if customers are willing to pay a premium for the convenience. While basic items like diapers and toilet paper are cheaper at Costco, one could make the argument that Amazon is still subsidizing shipping costs far too much as the prices are still comparable. And Amazon’s financials in the past few years have reflected higher shipping costs.
  2. Similarly, it is very expensive for physical stores to carry expensive inventory. This is why Tiffany sells most of its diamonds online and sells cheaper products via its retail stores. Still, keeping even some of its diamonds in physical locations is expensive, and that means Tiffany should only do so for customers willing to pay a premium for that. And they do. Tiffany’s margins are much larger than Blue Nile. This isn’t a luxury for Tiffany — it is a necessity.

We’ve only discussed the two extremes in this graph. What about everything in the middle? The reality for most large retailers is that they carry products that are scattered all over the graph. This, in turn, leads us to the final natural conclusion: It is in the interest of larger retailers to develop hybrid/”omni-channel” distribution strategies. This is why retail models such as “click-and-collect” have become popular in Europe.

Essentially, it is in the interest of Amazon to have physical locations to complement its online offerings. There are three massive advantages to doing so:

  1. It can leverage its incredible scale to truly be the retailer with the lowest prices – across its physical and online stores. Amazon store diapers will be the cheapest in the market. If you want to buy them online, however, you should be prepared to pay a premium for the convenience.
  2. It can use its physical locations as warehouses for “Prime Now” and “Fresh” offerings.
  3. Amazon has the data and analytics capabilities to be smarter about its inventory in physical retail locations than any of its competitors. This means it can have a real cost advantage — a big advantage in a traditionally low margin business.

As is the case with many things in life, the answer in picking the right distribution strategy lies in replacing “or” with “and.”

And at the rate at which Amazon has added businesses to its portfolio in the past decade, one could make the argument that few understand that idea better than Amazon and Jeff Bezos.

(HT: Prof Chopra’s work on Omni-Channel retailing @ Kellogg)

Rohan Rajiv is a second-year student in Kellogg’s Full-Time Two-Year Program. Prior to Kellogg he worked at a-connect serving clients on consulting projects across 14 countries in Europe, Asia, Australia and South America. He blogs a learning every day, including his MBA Learnings series, on