You are being programmed. Programmed to spend more time on social media, delivering more data points, which platforms like Google, Facebook and Twitter use to serve you hyper-targetted ads. Every day you are feeding information, consciously or unconsciously to firms, whose financial prosperity depends on their ability to convert your attention data into advertising spend. It is easy to portray the tech firms as the evil overlords and remind them of their self-proclaimed matra’s to organize the world information (Google) and bring the world closer together (Facebook). Pointing blame, however, obscures the common thread that ties these problems together. The root cause of evil is the attention-based business model itself. It permeates all facets of the business as the survival of tech firms and increasingly their advertisers depend on it. To paraphrase Thoreau: “There are a thousand hacking at the branches (tech companies) of evil to one who is striking at the root (business model).”

Incremental adjustments to reconfigure the attention-based model are not sufficient. A more radical change is required which puts consumers back in control of their data and replace the attention-based business models of tech companies. We need a new business model where consumers and not tech firms control their data and can make independent choices while allowing brands to turn a profit.

The problem

Google and Facebook are set to account for 84 percent of global spending on digital advertising, estimated to be at 100 billion dollars [link]. Brands have been shifting budget towards digital ads in a big way. For brands, Facebook is an attractive proposition, where else can you select your target audience in such great detail and obtain such detailed metrics on all kinds of conversions. Developments in AI coupled with personal customisation provide firms with the tools to optimise their business model. These developments hint on a view of the adjacent future in which consumer coercion is evermore effective and evermore undetectable.

Coercion is becoming effective

Tech firms have been building AI driven capabilities and are learning very quickly how to optimise the effectiveness of their business model. Effectiveness is increasing because AI systems can learn very quickly due to closed-loop learning. Tech firms to simultaneously measure everything about us (the input) and use this data to present the information we consume (treatment) and measure our behavior (the output). Continuously repeating this process at scale allows for rapid learning of what drives conversion and what doesn’t. Recent reports of Google buying large amounts of transaction data from Mastercard make sense in this respect [link]. Google is trying to close the loop between online behavior and offline purchase. Understanding, better than brands do, how online search and ad clicks lead to final purchases is, an ultimate sales argument to spend more on these platforms.

Compared to AI learning, human tendencies and biases are typically static and evolve in at a very slow evolutionary pace. With enough data AI systems can learn our biases and use these to get the optimal outcome. In a 2012 paper, researchers manipulated the newsfeed of 20K users on Facebook. Some users were fed more positive news, some more negative. The paper showed, how by altering valence of the user’s news diet, Facebook can influence users happiness up to the third degree(= the friend of a friend of a friend). This example illustrates the ability (already six years ago) to alter the state of mind of large groups of people without them knowing about it. To achieve the desired outcome, tech firms can exploit human tendencies which they can observe in the data. A recent study in Nature, for example, uses credit card transaction data to successfully predict a consumers lifestyles [link]. Tech firms thus have a huge advantage because they control what information we consume and can evaluate how this impacts our behavior. Over time, a few tech firms will be able to coerce preferences of large groups of people and sell that to the highest bidder.

Coercion is undetectable

While advertising has always been about influencing consumer behavior, the manipulation used to take place in plain sight. You know an advertisement when you see one and can weigh the message since motives of the advertiser are transparent, to sell you products. Advancement in machine learning coupled with mass customisation, however, enable more covert ways to program you into making choices which might not be in your own best interest. Research shows psychological micro-targeting can be used covertly to attract up to 50% more purchases (Match et al 2017). How do you know, why some search result or item was presented in the newsfeed? Imagine for a moment what the world would look like when not brands but you are in control. What would be in your newsfeed if the AI optimised for your well-being or self development? The reality, however, is that we are being programmed for the sake of buying products. The problem with this is not that AI is bad, but rather consumers do not control the objective which AI is optimising against, brands do. In the long run, increasingly effective and undetectable coercion of consumer preferences at scale becomes incoherent with a fundamental principle of free markets; the freedom to form a personal preference. How can we be certain that the choices we make are our own?

A solution: demand based business model

Tech firms are so dominant because few real alternatives for brands exist, the balance of power lies with the centralised. While fragmented, in the end, brands are the main source of revenue for these tech companies. A change will therefore only come when brands have a better way of connecting to consumers which is more effective while being less invasive.

The attention based business model works so well because brands knowledge of their customers and how to reach them bleaks compared to the ad-based tech firms. Where else can you select your target audience in such great detail and obtain such detailed metrics on all kinds of conversions (behavior)? As long as brands have no better alternative to shape perceptions and behavior, the accumulation and exploitation of consumer data will persist. We believe recent technologies allow consumer data to become decentralized and offer a better, less invasive and more effective business model for both consumers and brands.

Reverse demand information flow

The single most important piece of information about our purchase behavior is very hard to distill from our profiles. It answers the simple question: are you in the market? In the market means that a consumer has a need for a product (What) at a specific point in time (When). Matching products and preferences is difficult enough, but figuring out when there is a need is even more vexing.

What however if we could reverse the information flow, and have the consumer indicate that he or she is in the market, and open to hearing arguments from brands. The reversal would give consumers back control, would reduce the need for an abundance of data to profile consumers, have no need for an intermediary and provide brands with highly relevant information on prospective (and receptive) customers.

The process would involve a small slice of information exchange, but it would be the most relevant information. And it provides opportunities for consumers to clearly indicate the need or problem they have. Instead of trying to distill what consumers want from large amounts of (often irrelevant) data, they will tell you directly. In exchange for this information, consumers receive a stream of competing relevant offers to fulfil their needs. If the need is fulfilled the information flow will cease and no more irrelevant information will be shown to consumers.

The benefit of such a system is twofold, first consumers would have control over when they share data and will not have to experience the pain of annoying and irrelevant ads. Brands will have relevant (but anonymous) information on consumers without relying on third parties such as Facebook or Google, and can improve the effectiveness of their marketing spend. In essence, it would make the advertising business model obsolete and replace it with a business model which is more information efficient and is exempted from undetectable coercion. Rather a consumer indicates he/she is in the market and a brand can make it’s best effort to convince the consumer of their solution.

For such a system to function a number of problems such as anonymity, trust and scale would have to be solved. Using the blockchain technology, however, a decentralized solution for demand aggregation has become feasible solving these problems.

The ad-based business model we allowed to emerge leads to mass undetectable coercion of consumers. The business model is becoming increasingly effective but is highly inefficient and centralizes control in a few companies. We intend to replace this business model by a simpler, more elegant solution which gives consumers control and brands a highly relevant information on the (timing) on the needs of their clients.