How Ethical is Amazon's Flywheel?

For over twenty years, Amazon has radically evolved its business model through great lengths and at rapid speeds from selling books to becoming a trillion-dollar global marketplace, cloud-data provider as well as a world-leading advertising behemoth. Almost half the UK households, including mine, are frequent users of the platform mainly due to its extensive infrastructure, speed, selection of goods, convenience, and low pricing enabled by Prime subscription and now Prime everywhere. Over the years I have wondered whether moral, ethical, and governance choices come into question through Amazon's advertising business regarding consumer data and privacy and its use to stimulate its Flywheel. Below I reflect on some of these and ponder the what ifs.

As we know digital advertising and big data operate in a symbiotic partnership and have fast become the core driver of Amazon’s top and bottom-line growth. Digital Advertising is big business and is the workhorse for Amazon. Targeted online advertising drives growth for all the tech giants, serving targeted, relevant paid search and display ads based on interactions.

Whilst Facebook and Google collect, analyse, and monetise personal information by anonymising and creating segments based on individuals’ preferences which in itself is an ethical consideration, Amazon’s approach is unique. It presents more significant business and moral considerations through its algorithm across both the retail supply and demand platform (1P / 3P). Given Amazon’s ability to micro-target its own customers and the in-market user through devices like Alexa through browsing history, they can predict what an individual will buy before that shopper has even considered it. Through pervasive tracking and data mining of users from their retail business, they can manipulate all aspects of the eCommerce value chain from assortment to availability. The Flywheel Effect helps nudge browsers through a combination of disclosed and non-disclosed (native) advertising tactics, ensuring the user moves from browsing to buying, which can be argued to be an ethical issue given the way that it is deployed.

Data misuse has been a source of ethical and moral concern in retail since the beginning of modern trade. The issue of data privacy and the need to shield buyers from organisations collecting data has grown in importance over the past two decades. An article by the New York Times showed that many large companies misuse their customer’s data and it also showed that micro-level data collection is standard practice in all businesses and life. The reality is that most consumers outside the advertising industry are largely unaware of what unsolicited personal data is being harvested. This issue continues to be exacerbated, with the disclosed and non-disclosed data extraction methods becoming more nuanced and intricate, which makes it difficult for customers to understand how their personal data is collected, manipulated and used.

It could be argued that Amazon has sacrificed its ethical principles to improve repeat purchases, improved transactional value, loyalty and frequency. Given the importance of big data, privacy, and data management that is used to predict customer behaviours and market changes, Amazon has become an expert in using customers' data from its retail business across all its operations, including entertainment, advertising, and retail, cloud solutions without making shoppers fully aware of how their information will be used. Amazon has been criticised for this approach toward customer data collection and the lack of transparency across how this will be delayed and cross-fertilised with its own value chain. Within Amazon’s current code of ethics, there could be a need to enhance consumer trust through transparency around data management and principles of use.

Here we explore the ethical decision-making framework that Amazon uses and the environment it operates its advertising business as a means of subconscious manipulation. It has been argued that this can be seen to deny individuals the right to freedom of thought and given the digital shelf has increasingly become AI-powered, and auction-driven, it has been argued that Amazon’s practice is unethical, and its approach infringes human rights.

But is that really the case? Let's see.

Over the years, Amazon has faced many ethical challenges surrounding its business strategies. However, the issue of how it collects customer data and uses it across other parts of its business, particularly within data collection from unaware users who have not directly consented is often seen as a challenge. Given the sophistication of its algorithm (A10), which optimises the digital shelf rankings based on personalised marketing insights could be seen to be a complicated ethical dilemma. Amazon's extensive data collection and evaluation process are some of the most comprehensive available in the world. The algorithm profiles every user, can manipulate, and personalise almost every aspect based on its understanding of the individual. An argument often made is that this can go against an individual's own free will to browse, and select products and services they do not need nor can afford. Amazon’s strategy for data collection and through its monetisation mechanisms often uses tactics that bypass direct consent, and it is seen to be deliberately ambiguous about how it's customers' data will be used.

Amazon's actions operate within the law and comply with global data standards, that said the terms are well hidden in the small print when an individual signs up to shop through Amazon’s retail business. Amazon states that its approach is only used as part of its strategic imperative to improve customer experience/customer obsession by providing users with adverts explicitly targeted to them based on their browsing and behavioural insights. However, the customers' lack of awareness creates a problematic situation and can be seen as misleading consumers with adverts that violate' an individual right to freedom of information and expression.

The consumer data that Amazon collects is harvested in a subtle and nuanced manner. Amazon enriches its insights through telecoms data assets as the means of eliciting customer-related information where customers are not fully informed or in control of how their data is processed. Amazon then monetises its deep shopper insights to create highly effective advertising strategies for brands and sellers across its AMZ business, which can be argued to be a questionable ethical practice. Given that Amazon mines information that is then used for subconscious influence and control through its algorithm. The Flywheel Effect denies individuals the right to freedom of thought and steers shoppers down a dedicated path that has been set out for them by a machine. It manipulates insights around preference, choice, and selection to ensure users stay engaged on the platform repurchasing. However it can be argued that this tactic is as old as time and has been used by magicians, and traditional advertising the whole world over, so what's the difference. The difference is that Amazon controls the whole value chain, and is increasingly becoming an echo chamber and a one-stop-shop for everything from entertainment to impulsive shopping.

Approaching the problem from the ethical perspective requires viewing the issue as a dilemma between choosing customer-oriented services or obtaining customer consent. The consequentialist ethical theory explores that the rightness of an action is determined by its consequences. Therefore, consequentialism is a theory that says whether something is good or bad depends on its products and outcomes.

Bentham’s Utilitarianism can help justify the shareholder centric view for Amazon when collecting, harvesting, manipulating, and enriching consumer data that helps benefit the overall retail experience. The argument is that this is the right thing to do as it increases the level of happiness and creates more utility for the consumer. Utilitarianism suggests that any action ultimately leading to buyers’ benefit can be justified and, therefore, should be approved. Thus, through the lens of Amazon's customer obsession, the harvesting of disclosed and undisclosed customer data to enhance the retail experience is warranted. Consequently, it can be considered an ethically neutral action.

However, utilising Kant's Deontological ethics focuses on judging the actions themselves with a focus on means, and not ends with a focus on following the rules when decisions are made rather than focusing on outcomes and conclusions. It suggests that any action misaligning with the current ethical standards is, in fact, morally unjustifiable. Consequently, Amazon should be blamed for failing to inform its users and collecting their data without their true knowledge of how it plans to use this. Applying Deontological and Utilitarian ethical theories to Amazon’s current approach to personal data processing and its use in advertising proves that the company exploits customers and its third-party sellers.

Following Kant's Categorical and Hypothetical imperatives on moral law on how Amazon collects personal data and their approach to consent. We see that the three categorical imperatives are all compromised—the first, outlines the universal rules that we should not steal or lie. Amazon gains consumer consent to consumer data use through its retail business by employing confusing and subtle opt-in policies. The access and permission follow dubious procedures that can be argued as Amazon stealing its users’ data. Amazon can also be viewed as compromising the second categorical imperative by using consumer data for subconscious manipulation of all its retail operations to nudge shoppers to buy items they do not want, need, or can afford. This can be seen as infringing individuals' right to freedom of thought and, by default, not treating individuals as being who have the liberty of rational thought. Finally, Amazon should act by the laws imposed on itself, so that it doesn’t cross its hundreds of businesses and subsidiary trading entities. An example of this in practice is that Amazon leverages its first-party and third-party selling utilities to inflate the auction to increase competition and price. This is based on consumer data-led insights around selection, choice, preference, and demand. This can be argued to be unethical and against the categorical imperatives outlined by Kant. Amazon’s decision to use consumer data for more enhanced and effective advertising and the monetisation does not align with the current standards for data ethics around customer information. There is a moral responsibility to handle personal data in a way that does not harm them or infringe upon their rights. It can be argued that Amazon's Flywheel puts consumers in a dangerous position by exposing them to products and services they do not need nor can afford. Customer-led data transferred to a third for monetisation create better-targeted advertisements that are morally and ethically flawed. Amazon’s current behaviour appears to be misaligned with the principles of ethics, particularly customer autonomy. Therefore, Amazon holds the moral obligation to alter its approach to personal information management and introduce transparency into its data collection policies.

Consumers must be aware of how Amazon handles their data for legal and ethical reasons. Based on the Deontological principles of moral obligations and ethical values as the essential basis for decision-making, an organisation like Amazon should act upon the ethical standards for customer autonomy. However, the challenge is that Amazon has many different business-facing priorities where each aspect of their business is critical for the Flywheel Effect to work successfully. If one part is compromised, it harms all the others, negatively impacting its share price. The outcome is that for the Flywheel to function at peak performance for all aspects of its business, including for both buyers and sellers, there is a need for an auction where consumers become the by-product and commodity.

It can be argued that Amazon’s detachment from ethical and moral obligations can be attributed because it is a business entity that does not represent a specific individual. Thurlow shared that there is no point in expecting a corporation to demonstrate the presence of consciousness since "it has no soul to be damned, and nobody to be kicked." If corporations like Amazon will align with essential ethical behaviour standards without an external impetus would be na?ve whether their products, services or approach is morally or ethically sound.

At the heart of this challenge is thinking and acting responsibly that respects human rights. When reviewing Human rights within this case, Kant’s distinction between perfect and imperfect obligations helps frame discussion points. Perfect obligations are obligations of justice who owes what to whom can be fully respected. Imperfect obligations are obligations of beneficence. They are neither owned nor can they be claimed. It can be argued that data protection laws are underpinned by respect for fundamental human rights; in the case of Amazon, they use dubious tactics to gain consent for users’ data, and they are exploiting this access across many of its businesses for corporate and shareholder value. Amazon is harvesting users’ data without transparent consent, which can be harmful and restrictive to individuals’ freedom of thought through the augmentation and enrichment of their personal data, resulting in unconscious consumer influence.

This issue is exacerbated by the lack of relevant data understanding and Digital IQ by Amazon’s customers. Studies show that most people are unaware of the implications of consenting to Amazon harvesting and using their personal and behavioural data across all parts of its business and how that data would be applied. Whilst Amazon could counter-argue that the monetisation of personal information is all in service of its customer-centricity and uses this to tailor the digital shelf to improve customer experience.

Given Amazon’s positioning and strategy focused on consumer obsession, there is a failure to recognise the threat, leading to most customers being rather careless with reviewing what data they are giving away and how it will be used. Amazon tends to represent the specified process as the means of using advertisements and recommendations to meet and exceed customers' needs: "Amazon.com offers customised information to each customer through his or her Web page and recommendations based on the customer's interests and buying pattern." However, Amazon has not helped efforts to improve consumer awareness and understanding of how data is being used to drive unconscious influence through filter bubbles that drive up the advertising in the auction. Therefore, Amazon can be seen as irresponsible though subtle and ambiguous data usage that breaches ethical standards, human rights, and the overall safety of buyers.

Corporations like Amazon are morally responsible entities. It is natural to blame corporations for morally reprehensible behaviour, yet it is not apparent why we can blame them for having moral agency. Given Amazon's structure and setup, it makes it challenging to pinpoint blame for the ethical and moral issues that have been discussed through Amazon's current approach to data management. While one could claim that the company's information managers are at fault, their role in Amazon’s ethical breach boils down primarily to following the set framework for information management which is imperially connected to the organisation’s corporate strategy and long-term vision for the Flywheel and customer obsession. You could question the legitimacy of the strategy process established as the default, but the fault lies with the corporate standards leading to the development of the system. Therefore, you could argue that Amazon’s CEO Jeff Bezos and company leaders setting the ethical standards could be seen as the responsible parties.

The use of customers' data could be seen as minor at first glance as Amazon has not been selling customer data for malicious purposes and, instead, has been seeking to improve customer experience by making targeted advertising more precise. However, embracing the larger picture of the observed phenomenon, one will realise that the breach of corporate ethics and selling buyers' personal information for advertising purposes has created a rift between Amazon and its purpose and values, undermining their trust in the company. More importantly, Amazon’s actions have contributed to making the line between harmless shopping and possible unlawful sharing of customer data very blurred, which will, in turn, erode trust and loyalty. The duty falls onto Amazon to embrace the rights of all stakeholders. This attributes moral agency, and the people impacted by the organisation are responsible for setting the correct values around privacy and data use for profit.

To address the questions around the current ethical, moral, and human rights failings of how Amazon harvests user data for monetisation in advertising, it can be argued that Amazon will need to revisit its current framework for data management completely and reconsider the requirement for consumer data to power all aspects of its business. The reality is that changing the entire system is unlikely and near impossible, as policy changes will impact all of Amazon's businesses and operations, which will create shockwaves and a sharp reduction in share price. Instead, Amazon should start making its data policy and management process more transparent. Promoting customer awareness concerning data processing within Amazon must be seen as a crucial step.

To address the current situation and improve consumer confidence regarding data collection and use, it can be argued that Amazon will need to change its ethical practice toward all its customers, both on the seller and consumer side. It will be important for Amazon to take a fresh new approach to the transparency of its data value chain by providing precise and concise information about its data management policies. Amazon must also revisit its framework for Flywheel and its monetising of data through its advertising business.

First, they must consider the idea of introducing transparency across Amazon's retail business so that customers can be explicitly aware of how their data is used and how it will be used across all of Amazon's related services and companies. It can be argued that the echo chamber is becoming harmful and that the ethical code in the industry should be tightened up to establish rigour in how data is harvested, management and sold to avoid conflicts of interest. Amazon could also consider an awareness program to inform consumers about the industry’s information. Amazon has consistently argued that customer obsession gives them the right to use pervasive data to improve retail and consumer experience. We see this with consumers are choosing accessibility over ethics.