Kellogg Sample
Kellogg Sample
Kellogg Sample
5-315-501
Reinventing E-Commerce:
Amazons Bet on Unmanned Vehicle Delivery
I would define Amazon by our big ideas, which are customer centricity, putting the
customer at the center of everything we do, [and] invention. We like to pioneer; we like to
explore; we like to go down dark alleys and see whats on the other side . . . I know
[drone technology] look[s] like science fictionits not. It will work and it will happen.
Its going to be a lot of fun.1
Jeff Bezos, CEO and founder, Amazon.com
Videotex, developed in the mid- to late 1970s, was a technology that incorporated a television with a computer interface, allowing
users to send messages and content to each other.
2015 by the Kellogg School of Management at Northwestern University. This case was prepared by Professor Russell Walker and
Rafique Jiwani 14. Cases are developed solely as the basis for class discussion. Cases are not intended to serve as endorsements,
sources of primary data, or illustrations of effective or ineffective management. To order copies or request permission to reproduce
materials, call 847.491.5400 or e-mail cases@kellogg.northwestern.edu. No part of this publication may be reproduced, stored in a
retrieval system, used in a spreadsheet, or transmitted in any form or by any meanselectronic, mechanical, photocopying, recording,
or otherwisewithout the permission of Kellogg Case Publishing.
5-315-501
additionally, he felt that firms could gain competitive advantages by externalizing labor costs and
serving customers more efficiently.2
While finding short-lived success with B2B-focused customers (e.g., General Motors tested
the system to sell spare truck parts), Aldrich was unable to make headway in the minds of
consumers. They had just woken up to the idea of the VCR in their homes and the costs were
extremely high, and there was no real online marketplace in use or laws regulating how firms
could conduct business online.* Further, consumers were wary that these systems had no way of
transacting orders securely. Aldrich later commented on his release of the Teleputer:
It is also clear that moving a company [or consumer], for strategic reasons, from a low
technology profile to a higher technology profile is not an overnight activity. Before
using information technology strategically the goals, capabilities, positioning, and
constraints on the enterprise must be established . . . If the first step in our guide was the
management of change, the second step must be the business implications of change.3
Aldrichs insight, though ahead of its time, gave way to evolving technologies such as the
ATM, electronic payment through credit cards, and telephone banking during the 1980s.
However, it was nearly fifteen years before his work would be remodeled for a consumer willing
to accept the risk of shopping online.
1994 Resurgence
As the 1980s continued, little notable advancement occurred in the online shopping industry,
though there were signs of potential. The first mass-market online services, Prodigy and AOL,
began advertising flowers on their welcome pages in the late 1980s, but the attempt was more of
an advertisement than a platform to complete a transaction. Security was still seen as the major
impediment to consumer adoption. It wasnt until 1994, four years after the invention of the
World Wide Web, that security protocols (SSL) and high-speed connections (DSL) were
established, allowing users to confidently purchase goods online securely and quickly. The
withdrawal of entry barriers enabled the emergence of hundreds of online retailers by 1995,
including Amazon, eBay, and Dell. Offering goods at prices 10 to 20 percent below those at
brick-and-mortar retail locations allowed online retailers to grow substantially through the latter
part of the decade. By 1999, the U.S. online retail market reached over $15 billion in annual
sales.4
Along with the exponential rise of Internet retail companies came a substantial amount of
venture capital (VC) funding. VCs believed that the success of an online retailer would come
after realizing net losses in order to gain market share, and were willing to back startup dot-coms
based simply on an idea. Within a year, the NASDAQ fell 78 percent from its high in March
2000, and 52 percent of dot-com companies established between 1995 and 2000 disappeared.
This left only a few pure-play online retailers, including Amazon, eBay, and Priceline.
*
By 1984, California was the only U.S. state that had passed an electronics commerce act that defined corporate and consumer rights
online.
5-315-501
20012013 Boom
Continuous revitalization of goals, identification of white space,* and diversification of risk
became a major theme for dot-com companies that survived the bubble and the subsequent
decade. Amazon explored interests in B2B, reseller, and hardware operations; Priceline shifted its
product focus from air travel to hotel and rental car travel; and eBay drastically expanded its
product lines. By 2013, Amazon was the clear and dominant leader in online retail. With over $74
billion in revenue, it outpaced eBay, Priceline, and all other online retailers many times over.5
Amazon.com
History
The story of Amazon is widely known. Founded in Jeff Bezoss garage in 1995 as a reseller
of books, the company took advantage of the rapidly growing e-commerce space that followed
the increased security measures enacted by Netscape founders. Attempting to carry every
product from A to Z, Bezos enacted an unusual business plan (though not uncommon to Internet
startups at the time) that focused on customer and revenue growth rather than profit. He also
emphasized the need for Amazon to pursue areas within and outside online shopping from the
beginning of its existence as a public company. In a 1997 letter to shareholders, Bezos contended:
Our goal is to move quickly to solidify and extend our current position while we begin to
pursue opportunities in other areas. We see substantial opportunity in the markets we are
targeting. This strategy is not without risk: it requires serious investment and crisp
execution.6
By 1997, the companynow publichad greatly expanded its reach. Employee headcount
grew from 185 to 614; distribution center capacity grew from 50,000 to 285,000 square feet; and
book inventories surpassed 200,000 titles. Amazon soon found itself competing less on price with
incumbent retailers and more on convenience and time to delivery.7 Throughout the next fifteen
years, Amazons focus on winning with time and price in every aspect of its business was crucial.
Retail competitors had a difficult time transitioning from brick-and-mortar operations to ecommerce platforms. Even the ones who experienced some online success still had trouble
competing with Amazon because of its sheer range of product offerings and logistical
capabilities. By the end of 2013, Amazon had $74 billion in sales and had turned a profit for five
consecutive years. (See Exhibits 1A, 1B, and 1C for Amazon financials.) It occupied nearly 50
million square feet of distribution centers around the country and employed more workers than
Google and Microsoft (Exhibit 2).
White space is a management term coined in 1991 by Geary Rummier and Alan Brache to identify areas of an organization where no
one is in charge. Mark Johnson redefined the term in 2010 as an area in which businesses can create new business models.
5-315-501
As Amazon grew over the next decade, it leveraged the Amazon brand as it introduced cloud
computing services (Amazon Web Services) and its first hardware product, the Amazon Kindle.
In an article in the Harvard Business Review, Accenture partners Larry Downes and Paul Nunes
described the Amazon Kindle:
Amazons real innovation was waiting just until the right combination of technologies
was ready for mainstream use and then leveraging its powerful brand and customer
network to launch Kindle with easy access to a huge catalog of books on day one . . .
along the way, they scrambled every link in the supply chain.9
While the Kindle grew, Amazon invested heavily again in serving customers as quickly as
possible with Amazon Prime, a service guaranteeing two-day shipping for certain products.
Initially, 1 million products were targeted for Prime at its launch in 2004. By 2014, 120 million
products in Amazons catalog qualified for Prime. The $79-per-year subscription service became
a huge profit machine for Amazon, and the company soon began adding digital services for Prime
members, including Prime Instant Video. Many Prime members claimed they would pay over
$100 for annual services, and they spent nearly double at Amazon than regular customers ($1,200
per year versus $600 per year). Instant access became a major competitive advantage for
Amazon. Customers saved time on everything from watching movies to downloading books to
ordering groceries, and viewed Amazon as their premier online shopping destination. Reports
indicated that the number of Amazon Prime members could reach 25 million by 2017.10
5-315-501
arrangements with these companies were beginning to become a major concern for Bezos and
Amazon. In a 2013 statement to shareholders, Bezos discussed the necessity to reduce fulfillment
and delivery service costs in order to maintain the health of Amazons business:
We rely on a limited number of shipping companies to deliver inventory to us and
completed orders to our customers. If we are not able to negotiate acceptable terms with
these companies or they experience performance problems or other difficulties, it could
negatively impact our operating results and customer experience.12
Amazon heavily relied on delivery services such as UPS and FedEx. Both UPS and FedEx
announced in 2014 average price increases of 4.9 percent and 3.9 percent, respectively, on
delivery costs due to increased fuel prices. It was estimated that fuel for the long-haul fleets that
UPS and FedEx employed accounted for nearly 40 percent of their operating expenses.13 (See
Exhibit 6 for financial highlights for UPS and FedEx.)
Lobbying
Amazon began heavily lobbying the U.S. government in 2010 on issues such as taxes on
online sales, transportation safety, data protection, and intellectual property. In addition, the
company fought hard at the state level to ensure that it could retain multiple competitive
advantages, such as tax breaks on large distribution centers and of course the ability to sell
without charging customers sales tax. (Amazon threatened to leave the state of Texas when it
asked to collect taxes from the companys shipments.14) Amazon spent $3.4 million on lobbying
in 2013, its highest spending year since its inception, and over $200,000 on political donations. It
partnered with the CIA in a $600 million cloud-computing contract in 2012. In his 60 Minutes
interview, Bezos claimed that Amazon conducted a significant amount of lobbying in 2013 for
the legalization of unmanned autonomous drones. (See Exhibit 7 for Amazon lobbying
spending.)
Customer Data
Amazon considered the data it captured from its customers one of its most important assets.
The data allowed the company to foster a relationship with customers that it claimed is superior to
that of a traditional retailer. Aside from promising its customers that their data was kept secret,
however, Amazon rarely discussed how much it did or did not use it, though customers highly
valued the product recommendations that Amazon provided based on the customer data. A
special report in Time magazine investigated the data that Amazon collected:
While brick-and-mortar stores are black boxescustomer behavior inside the store is
effectively invisible to managersAmazon is able to collect endlessly useful information
about shoppers and use it to sell more stuff by targeting customers through e-mail and
the website itself. Whenever a customer buys something from Amazon or logs in without
buying something, Amazon is collecting all kinds of information about that person.
Theres a lot of data that can be mined about how they peruse the website, what they put
in the cart, what they abandon, and how the customer actually goes about searching for a
product.15
5-315-501
By 2014, Amazon enjoyed tremendous customer trust in the management and use of
customer data, providing it a clear advantage over other leading digital firms.16 Shoppers trusted
Amazon more than Google, Facebook, Twitter, and Apple when it came to consumer privacy. In
fact, only 7 percent of American consumers viewed Amazon as a threat.17
Autonomous Vehicles
Background
By 2014, the development of autonomous vehicles (AVs) had been underway for quite some
time. In his 60 Minutes interview, Bezos suggested that Amazon would deliver goods by drones,
or flying AVs. At the same time, however, advances in autonomous cars and trucks were being
made, allowing for the possibility that retailers could use a fleet of both drones and autonomous
cars and trucks. By that time, the National Highway Traffic Safety Administration (NHTSA) had
classified autonomy into five different levels when associated with a land- or air-based vehicle:
Level 0:
Level 1:
Level 2:
Level 3:
Near autonomy; a driver is present only for critical control of the vehicle.
Level 4:
Full autonomy; a driver is not needed to be present for control of the vehicle
and it runs solely on data it collects from surrounding vehicles. (Amazon
focused on using this level of autonomy.)
*
In 2014, Peloton Technology, an autonomous trucking company, estimated that AVs could save $6 billion per year for the trucking
industry.
The cost for a vehicle crash on the job can cost an employer up to $74,000. See OSHA, Guidelines for Employers to Reduce Motor
Vehicle Crashes, https://www.osha.gov/Publications/motor_vehicle_guide.pdf (accessed September 14, 2015).
5-315-501
efficiently and without the error introduced by drivers. Without drivers, AVs could use less fuel
and become more efficient in operation and ownership.
But there were concerns with investing in AV technology as well. First, a major impediment
for the mainstream introduction of AVs would eventually be consumer adoption for the emerging
technology. Without adoption, there is no convergence, and without convergence the limitations
for scale are real.18 Scale was important, particularly because of the reliance of AVs on big data.
In fact, it was the large amounts of consumer data that vehicles could collect and use that would
make them truly valuable. Once customers could hail a driverless car to go where they needed,
the data collected about the trip would not only provide operational efficiencies, such as which
route to take, but would also provide a view on what individuals elected to do and where they
went. Wide-scale use of AVs in society would provide data that could be used to determine not
only how customers travel and where they shop, but also when and where they work and other
elements of their lives. For critical adoption, consumers would need assurance that data would be
secure in the hands of whichever company controlled it.
Second, although consumers were excited about the prospect of autonomous features in their
vehicles, only 12 percent claimed they would actually feel safe purchasing an AV or being on the
road with one (though this number was highly skewed toward urban dwellers). Predictions about
when consumers would embrace the new technology widely varied. (See Exhibit 10 for
predictions of full Level 4 AV adoption.) Political support for Level 4 autonomy would also be
crucial for consumer adoption. Following on the heels of advancement in AV technology was a
rise in legislation associated with it. (See Exhibit 11 for a timeline of recent U.S. state legislation
involving AVs.)
Competitive Landscape
The 2010s saw many different parties investing meaningful capital in the production of AVs.
(See Exhibit 12 for renderings of AVs for some of these firms.)
GOOGLES SELF-DRIVING PROJECT
Google did not plan to sell the car but the technology behind it; its commercialization plan
involved partnering with retailers, who could use the autonomous technology to serve customers
5-315-501
quickly, and car service companies, such as UberRush, in densely populated areas. Additionally,
the vehicles could be given routes via mobile devices, which would allow for greater fuel
efficiencies.
CATERPILLAR AND THE ROBOTICS INSTITUTE
In 2010 Carnegie Mellon Universitys Robotics Institute conducted a study with Caterpillar
to develop an autonomous truck to be tested in a barren terrain of Central Australia. By 2013, the
self-driving truck had been developed and tested, demonstrating the ability to carry 240 tons of
iron ore in a single trip. By 2014, Caterpillar had six fully functional mining trucks. Each truck
had 2,650 horsepower from combustible engines and was driven by over 25 million lines of
software code. These trucks operated without drivers and could work day and night. The
automation replaced four drivers per truck per 24-hour period.20 Ted Scott, director of
engineering and safety policy for the American Trucking Association, which sponsored the study,
said:
Ubiquitous, autonomous trucks are close to inevitable. Eventually we are going to have
wireless, self-driving trucks because there will be money in it. Safety features like
automatic braking will be commercially available over the next decade, but they will be
quickly superseded by autonomous trucks after that.21
Caterpillars trucks were considered connected vehicles (at Level 3 autonomy), as they still
needed technicians to monitor and sometimes guide several trucks at once, similar to
advancements developed by Peloton Technology.* Most experts believed that even at its nearterm technological peak, full Level 4 autonomy might not be a realistic goal.
UPS AND FEDEX
UPS and FedEx had kept fairly quiet about their potential use of AVs until Bezoss
appearance on 60 Minutes, though both had expressed a desire to significantly decrease labor
costs. After the interview aired, a UPS spokesman mentioned that the company might have many
uses for drones. Particularly, UPS trucks could bring packages quickly from an airport to a major
distribution center in more remote locations, speeding up the delivery process from business to
customer. The spokesman said, UPS invests more in technology than any other company in the
delivery business, and were always planning for the future.22
FedEx founder Fred Smith spoke more about his eager desire to get into the unmanned aerial
vehicle game than UPS. FedExs goal was to have an aircraft that had no one on board and that
could carry a significant amount of cargo. Smith continued:
A modern 777 is already capable of being an unmanned vehicle. They let the pilots touch
the controls for about 20 seconds, to advance throttles, and then the plane takes over.
Today, pilots drive the planes on the ground, but theres no reason a computer cant do
that. Its just a matter of getting the laws into place so companies can begin building to
those specifications and doing some real field testing.23
*
Peloton Technology was at the forefront of Level 3 autonomous innovation. Its driver-assisted truck platooning (DATP) model was
working with the Federal Aviation Administration for a 2015 launch. Exhibit 13 shows an illustration of the technology.
5-315-501
Although Amazon did not release figures for the portion of R&D spent on Prime Air, in 2013
total R&D expenses reached $1.73 billion, an all-time high. Estimates put the cost of owning and
operating air drones at $20,000$30,000.
Conclusion
Though Prime Air would not be ready for deployment for at least another year, Bezos had
many questions before it could launch. He knew that Amazon needed to continuously innovate to
provide the level of service its customers were accustomed to having. But what risks would be
associated with venturing into autonomous delivery before consumers fully adopted the
technology? Moreover, what risks would Amazon take on if it waited for another competitor to
enter the market first? How would the retail landscape change and could Amazon change with it?
The problem was, the information he needed to answer these questions would not be available for
monthsor even years.
5-315-501
10
5-315-501
11
5-315-501
12
5-315-501
Source: Blair Hanley Frank, Amazon Soars to Nearly 110,000 Employees, Surpasses Microsoft for First Time, GeekWire, October 24,
2013.
13
5-315-501
Source: Nicholas Carlson, Visualizing Amazons Acquisition History, Business Insider, July 27, 2009.
14
5-315-501
Source: Amazon Fulfillment and Distribution Center Locations Map, E-Commerce and Auction Site News, January 26, 2014,
http://auctionsitenews.com/amazon-fulfillment-center-locations.
15
5-315-501
16
5-315-501
2012
Revenues
55,438
54,127
Operating expenses
48,404
52,784
4,372
807
Net income
FEDEX ($ IN MILLIONS):
2013
2012
Revenues
44,287
42,680
Operating expenses
41,736
39,494
1,561
2,032
Net income
17
5-315-501
No Automation (Level 0): The driver is in complete and sole control of the primary
vehicle controlsbrake, steering, throttle, and motive powerat all times.
Function-Specific Automation (Level 1): Automation at this level involves one or more
specific control functions. Examples include electronic stability control or pre-charged
brakes, where the vehicle automatically assists with braking to enable the driver to regain
control of the vehicle or stop faster than possible by acting alone.
Combined Function Automation (Level 2): This level involves automation of at least two
primary control functions designed to work in unison to relieve the driver of control of
those functions. An example of combined functions enabling a Level 2 system is adaptive
cruise control in combination with lane centering.
Limited Self-Driving Automation (Level 3): Vehicles at this level of automation enable
the driver to cede full control of all safety-critical functions under certain traffic or
environmental conditions and in those conditions to rely heavily on the vehicle to
monitor for changes in those conditions requiring transition back to driver control. The
driver is expected to be available for occasional control, but with sufficiently comfortable
transition time. The Google car is an example of limited self-driving automation.
Full Self-Driving Automation (Level 4): The vehicle is designed to perform all safetycritical driving functions and monitor roadway conditions for an entire trip. Such a design
anticipates that the driver will provide destination or navigation input, but is not expected
to be available for control at any time during the trip. This includes both occupied and
unoccupied vehicles.
Source: U.S. Department of Transportation Releases Policy on Automated Vehicle Development, NHTSA 14-13, press release, May 30,
2013.
18
5-315-501
Percentage of
Total (%)
Fuel
0.540
39
70,200
20
0.240
17
30,600
25
90
Component
AV Savings
(%)
Drivers salary
0.360
26
46,800
0.120
10
15,000
Insurance
0.050
6,500
80
Tires
0.030
4,000
Permit, licenses
0.020
3,600
Coffee
0.004
Total
1.380
600
180,000
39
Truck,
KPMG
Truckers
Report,
Year
20162018
Summary
Plans to release Google-car technology
Nissan
2020
Available in showrooms
GM
2020
BMW
2025
Ford
2025
McKinsey
2025
IDC
2040
IEEE
2040
19
5-315-501
May 2012:
The Nevada Department of Motor Vehicles issues the first license to a selfdriven car to a Toyota Prius using Googles technology.
July 2012:
September 2012:
January 2014:
Michigan allows the testing of AVs with a human being inside the vehicle.
Source: Bryant Walker Smith, Automated Vehicles Are Probably Legal in the United States, Texas A&M Law Review 1, no. 411 (2014).
20
5-315-501
CATERPILLAR
AMAZON
21
5-315-501
22
5-315-501
Endnotes
23