Select Page

In today’s era, you can’t escape Amazon. If you’re out shopping, you check to see if Amazon has a better price. If you’re considering purchasing a certain item, you read the Amazon reviews to determine the quality. For that person you have no idea what to get for their birthday, you browse Amazon for the things that person likes and see what other shoppers also purchased.

It’s incredible how Amazon has seeped into every imaginable part of our shopping experience, but they did it in a very calculated, personalized fashion utilizing data science to make it the number one site for online commerce. By the same token, however, Amazon has been criticized for the amount of data it “takes” from its users, and some privacy advocates have argued that the benefits don’t outweigh the costs. As producers of data, we need to know how our data is being used and make sure we feel good about what we receive in return for our data.

In his book Data for the People, Andreas Weigand discussed exactly how Amazon devised its online algorithm for user satisfaction by harnessing buyer data and linking and distributing it in ways that other buyers find helpful. Weigand himself was one of Amazon’s chief scientists and now spends his time advising other data refineries and teaching higher-ed courses on big data management, algorithm development, and The Social Data Revolution.

In talking about how Amazon collects data and puts it to use, Weigand wants users to understand the “currency” of their social data, that is, the data about who they are, what they like, how they move, and how they interact. As users become more versed in the “worth” of their social data, they can call for more transparency from the likes of Amazon and Google and ensure that everyone benefits from the use of the data.

Take one of the most ubiquitous examples: When you’re examining an item on Amazon, the site will offer items usually bought along with this item as well as other items that “shoppers who bought this item” tend to purchase. In general, Amazon shoppers like this feature. Especially if they’re unsure of what they’re getting and if they should complement the item, these suggestions prove helpful. Amazon gets this data by aggregating lots of buyer data and determining how its use will drive sales.

Amazon shoppers also love to read reviews, even if they have no intention of purchasing the item from Amazon. Amazon had to endure some trial-and-error before it perfected the art of collecting honest reviews that customers find beneficial. For a while, Amazon did not require that a person’s username be posted along with a review. It considered requiring a person to list their full name, but all it requires today is a public username. Why? Users may want to investigate the reviewing history of another user to “consider the source,” as the old saying goes.

A recent article on medium also tapped into that key aspect of the internet that Amazon also capitalized on: connectivity. Mogul-turned-president Donald Trump’s app Whisper and the popular campus app Yik-Yak have both realized the hard way that statements and reviews need some way to self-validate and connect people to each other — and anonymity is inconducive to this end. Amazon respects that not everyone will want their full legal name out there, but users’ ability to track a person’s reviewing history is important to building a community.

Another level of concern arises in Amazon’s use of your vocal data in two particular areas of Amazon’s services. First, consider what happens whenever you call a customer service hotline: right before you are about to be connected with a real person, you hear an automated voice declaring “This call may be recorded for quality assurance purposes.” After the call, some combination of algorithms and people may code your call for your needs, your mood, how you reacted to certain statements from the representative, and so forth to better train and tailor employees for how to deal with callers like you.

Harvard Business Review recently ran an article about just how seriously the call center sector needs to improve its customer service and training, so on the one hand, the collection and analysis of your information will help  not only you but every other caller have a better experience on the phone with Amazon. However, consider all the data they now have about you that you have no control to edit, comment on, and analyze for yourself. Amazon has the power to annotate, code, determine the outcome of the call, but you have no access to all this data. Andreas Weigand calls for recorded phone conversations to be available to both the caller and to the recipient so that everyone utilize it however they see fit. Amazon users want their data to be used to improve their experience, but they also want to see exactly how that’s done. The exchange of data for convenience is not bad so long as everyone is fully aware of the terms of the trade.

Amazon Echo is another example. This “Smart Home” device is programmed to be on standby mode all the time and jump to attention when it hears the hot word, Alexa, it’s name. In standby mode, Echo “listens” to the background noise and filters out anything that’s not it’s hot word. Privacy concerns instantly emerged regarding how much of this background noise data Amazon keeps and analyzes for its own opaque purposes. Here again we have an exchange of data for convenience, and data producers need to ensure that the exchange is fair and transparent. Amazon has already clarified that it will not release noise data to police, but users also need to understand the give and take of turning off features regarding voice data.

Amazon’s means of harnessing data to produce a good user experience is not a sum-zero game, but users should be aware of the “rate of exchange” and understand how their data is being used. Its tactic have made it an unstoppable force in E-commerce and users need understand how it came to be and ask for more transparency in the process.