Taking Analytics Up A Notch

Amazon.com, Inc.

Founded in 1994 and headquartered in Seattle, Washington, Amazon.com, Inc. (Amazon) has grown to be a Fortune 500 company and one of the largest online e-commerce retailers. While Amazon was launched as an online bookstore, it has since expanded to offer millions of unique new, refurbished, and used items in categories such as books; movies; music & games; digital downloads; electronics & computers; home & garden; toys; kids & baby; grocery; apparel; shoes & jewelry; health & beauty; sports & outdoor; and tools, auto & industrial. Additionally, in 2007 Amazon entered the e-book industry and produced its own e-reader, the Kindle, as well as compatible e-books. Since then, Amazon has expanded the Kindle family to include the Kindle Fire HD 4G LTE Wireless, with HD display, Dolby Digital Plus, and 4G connectivity; and Kindle Paperwhite, the world’s most advanced e-reader. Lastly, Amazon more recently launched its Amazon Web Services (AWS) to provide customers with access to in-the-cloud infrastructure services, which can be used to enable virtually any type of business. (Amazon, n.d.).

Data Collection

As Amazon works to offer customers more types of products, more conveniently and at lower prices, it continues to keep its mission in mind - “to be Earth’s most customer-centric company, where customers can find and discover anything they might want to buy online, and endeavors to offer its customers the lowest possible prices,” (Amazon, n.d.). In order to remain a consumer-centric company, Amazon has put forth a great deal of effort into developing a ‘Culture of Metrics’. Even since its beginning, every decision and strategy outlined from Amazon is based heavily on big data. (Chaffey, 2014).

According to Amazonia: Five Years at the Epicenter of the Dot.Com Juggernaut, a book that charts Amazon's early growth from an employee’s perspective, James Marcus (2004) described an occasion at a corporate 'boot-camp' in January 1997 when Amazon CEO Jeff Bezos 'saw the light' and stated that the company will have a ‘Culture of Metrics'. Bezos believed that it would provide the company with an ‘amazing window into human behavior’. In the book, Marcus provided a great recap of one ‘boot-camp’ discussion about the beginning ideas of incorporating big data into their business strategy: (Chaffey, 2014).

Gone were the fuzzy approximations of focus groups, the anecdotal fudging and smoke blowing from the marketing department. A company like Amazon could (and did) record every move a visitor made, every last click and twitch of the mouse.

As the data piled up into virtual heaps, hummocks and mountain ranges, you could draw all sorts of conclusions about their chimerical nature, the consumer. In this sense, Amazon was not merely a store, but an immense repository of facts. All we needed were the right equations to plug into them.

The book also included insight into breakout group discussions of how Amazon could better use measures to improve its performance, especially customer-centric metrics. Since then, Amazon has developed internal tools to support its ‘Culture of Metrics’. One analytics tool that has provided Amazon with great success is its ability to run A/B testing experiments. Instead of having arguments about what should go on the home page or category pages, etc. Amazon utilized real-time experimentation tests to answer those questions, since actual customer behavior was the best way to decide upon tactics. However, it became apparent that consumers’ online experience evolves over time, which means that Amazon has to consistently test and tweak its features to match. ‘Automation replaces intuition’ became the philosophy around business decisions.  (Chaffey, 2014).

User Recommendations

With a glimpse into how powerful data can be, it became known as “King” around Amazon. The company collects nearly every piece of data available on consumers visiting the website. Every purchase, every page viewed and every search is recorded. Amazon relies on acquiring and then crunching a massive amount of data to make strategic business decisions including providing customer channel preferences; managing the way content is displayed to different user types such as new releases and top-sellers, merchandising and recommendation (showing related products and promotions). (Chaffey, 2014).  

User recommendations have become a signature feature for Amazon. In order to provide meaningful recommendations, Amazon uses its “item-to-item collaborative filtering” algorithm to customize the browsing experience for returning customers. The algorithm takes what a user has bought in the past, which items they have in their virtual shopping cart, items they’ve rated and liked, and what other customers have viewed and purchased into consideration when suggesting recommendations. When looking at Amazon.com, it’s clear that the company has integrated recommendations into the majority of the purchasing process because multiple panes of product suggestions are visible including “Frequently Bought Together”, “New for You”, “More Items to Consider”, and “Related to Items You’ve Viewed”. (Mangalindan, 2012).

David A. Steinberg, the founder of XL Marketing, a digital marketing company, states, “Amazon's algorithm is a recommendation engine, and the company runs in such a way as to help the consumer. Amazon is employing behavioral targeting and recommendation engines to help create a transaction with the consumer. The megaretailer wants to drive the most meaningful offers to its users, so the more information it compiles, the more accurate Amazon's recommendations based on psychographics, demographics, or spending habits are,” (Carlozo, 2013).

Moving Forward

Driven by technological innovation, Amazon continues to amaze by taking steps forward in creating the Earth’s most customer-centric company. With the success of the collaborative filtering algorithm, Amazon is now taking its analytics development up a notch and focusing on predictive analytics. In this latest business move, Amazon has acquired a patent for its new venture - Anticipatory Shipping, “a shipping system designed to cut delivery times by predicting what buyers are going to buy before they buy it — and shipping products in their general direction, or even right to their door, before the sales click even (or ever) falls,” (Lomas, 2014).

This system is purely based on Amazon’s predictive big data analytics, which the company trusts enough to accurately predict what a consumer will order next, and when. Even though Amazon offers two-day shipping, next day air, and Sunday deliveries, cost is a huge concern and having to wait even a day can cause some consumers to visit a physical store to purchase the item. Amazon hopes that Anticipatory Shipping will encourage them to shop online and avoid brick-and-mortar retailers. (Marr, 2014).

So how will Anticipatory Shipping really work? According to the patent, this algorithm uses data from a consumer’s prior Amazon activity, including time on the site, duration of product views, what links were clicked and hovered over, shopping cart activity, and what items were added to a wish list. When available, the predictive algorithm will also include data from real-world information collected from customer telephone inquiries, responses to marketing messaging, as well as other factors. (Ulanoff, 2014).  In its patent, Amazon included a basic flowchart to help demonstrate how the process might work.

In addition to reducing shipping time, this system could also help to increase sales and potentially reduce shipping, inventory and supply chain costs. “Supply chain and logistics optimization is neither easy nor cheap, but it is the biggest opportunity for most companies to significantly reduce their cost and improve their performance,” wrote H. Donald Ratliff, Ph.D., executive director of the Supply Chain and Logistics Institute (Ulanoff, 2014).  Ratliff continued by saying, “For most…operations, there is an opportunity to reduce cost by 10% to 40% by making better decisions.” For a company the size of Amazon, a 10% to 40% in annual savings would be huge, but it would require the company to get it right each time. If the big data algorithms get it wrong, the company could end up losing the money spent on shipping the product out and also returning it. “The way Amazon proposes to deal with cheaper unwanted items is to either heavily discount them or give them away as a free gift to build customer ‘good will’,” (Marr, 2014).

Additional Tools and Strategies

While it seems like Amazon already has everything under control, there’s always room to grow and improve. As a Prime member, I enjoy all of the benefits that come with the membership including Amazon Instant Video, which is an Internet video-on-demand service. The service offers television series and films for rent or purchase, however the service is free to customers with an Amazon Prime subscription. I don’t think its recommendations are quite on par with Netflix, but I enjoy not having to pay extra to use the service. In addition to Prime, I’m also an Audible member, which is an audiobook-on-demand service and owned by Amazon.  This service provides access to over 150,000 titles and is the world’s largest provider of audiobooks and spoken word content.

Amazon Instant Video is a great service, but it does not receive as much attention as other Amazon services when it comes to cross-selling. One way that analytics can help improve its awareness is for Audible to suggest videos to purchase or rent to users based on what audiobooks they are looking at or buying – especially when the book was made into a movie. When searching audiobooks on Amazon.com, Audible shows up as a purchase option, which gives awareness to that service. I think Amazon Instant Video would benefit significantly from the same situation on Audible. Additionally, Amazon.com only offers Instant Video suggestions when searching for a book from the homepage. When searching for a particular book under the “book” category, Instant Video suggestions are removed. This would be another great place to cross-sell and bring awareness to Instant Video based on suggestions related to the books that are being searched.  

The lucky one searched on Audible could include a link to the video on Instant Video and suggestions for other Nicholas Sparks videos:

  

The Lucky One searched on Amazon in the “book” category included a link to Audible as a purchase option. It could also include a link to Instant Video and suggestions for other Nicholas Sparks videos:

The Lucky One was only suggested on Instant video when searching for The Lucky One from the home page:

References:

Amazon. (n.d.). About Amazon. Retrieved from: http://www.amazon.com/Careers-Homepage/b?ie=UTF8&node=239364011

Chaffey, D. (2014, June 30). Amazon's business strategy and revenue model: A history and 2014 update. Retrieved from: http://www.smartinsights.com/digital-marketing-strategy/online-business-revenue-models/amazon-case-study/

Lomas, N. (2014, January 18). Amazon Patents “Anticipatory” Shipping — To Start Sending Stuff Before You’ve Bought It. Retrieved from: http://techcrunch.com/2014/01/18/amazon-pre-ships/

Mangalindan, J. (2012, July 30). Amazon's recommendation secret. Retrieved from: http://fortune.com/2012/07/30/amazons-recommendation-secret/

Marr, B. (2014, February 5).  Amazon: Using Big Data Analytics to Read Your Mind. Retrieved from: http://smartdatacollective.com/bernardmarr/182796/amazon-using-big-data-analytics-read-your-mind/

Ulanoff, L. (2014, January 27). Amazon Knows What You Want Before You Buy It. Retrieved from: http://www.predictiveanalyticsworld.com/patimes/amazon-knows-what-you-want-before-you-buy-it/