In the regional Chinese city of Rongcheng, the ruling one-party state is perfecting its soon to be nationwide “Social Credit System”. Starting with 1,000 points, every citizen has points deducted – or added – for their historical (and current) behaviour. This includes: applying criminal convictions; contributions (or not) to charity; promptness of invoice payments; credit history; and any other recorded incident which indicates how committed the individual is to the “well being” of the state overall.
Those with low Social Credit scores are potentially denied access to various state and local government services; travel – both internal and external; certain types of job; and the elite universities. Underpinning this system is a digitally-based network also involving facial recognition technology and AI, with national roll-out planned for 2020.
We can observe this through our GDPR-tinted glasses, comforted by the knowledge that we have significant privacy barriers in place to prevent this from developing in our own western democracies. Except, of course, we are not safe. The Chinese government has formally announced that it is collating every piece of information that it holds on its citizens into one database. In the West various bodies – both private and public – hold this information, but not yet in any coherent framework.
However, announced or not, we are still moving to the outcome being pioneered in Rongcheng. There has been no formal acknowledgement, but we are moving rapidly to what Shoshana Zuboff – a Harvard professor specialising in social psychology – calls “surveillance capitalism”. She defines this as “ongoing commercial surveillance…utilising big data…that threatens individual autonomy and democracy”. This is partly happening because the global advertising industry, blinded in the headlights by initially Google, then Facebook, and latterly Amazon, has actively supported the growth of these companies, with tens of billions of advertising dollars being switched from “legacy” media.
As recently as last week, Aaron Greenspan, a university contemporary of Mark Zuckerberg and tech investor, expressed the view to the DCMS Parliamentary Committee that Facebook had perpetrated “the largest fraud in corporate history” on the basis that it has billions of fake user accounts, so advertisers are not actually reaching the audiences that they believe that they have paid for. (Greenspan has shorted Facebook stock, so he does have a clear interest in driving the share price down.) In May, Facebook said that fake accounts only represent five per cent of their global monthly users.
Whether it is the circa 50 per cent plus of monthly unique user accounts that Greenspan claims are fake, or the five per cent or so that Facebook admits, it is clear that a significant slice of advertising money is being wasted. Google and Facebook in particular, and now potentially Amazon, have prospered because in the chase for consumers the advertising sector has been insufficiently robust in its focus on the audience metrics. It has suited both buyers and sellers to operate with an approximate audience short-hand that emphasises broad numbers over accuracy.
The effect of this huge – and by historic standards very quick – shift of advertising revenue has been to allow these three companies to aggressively invest in developing four key areas as defined by Google’s chief economist Hal Varian. These are: “enhanced data extraction and analysis; new “contractual” forms using computer monitoring and automation; hyper-personalisation; and using their world-leading tech infrastructure to carry out continual experiments on users and consumers – with constant refinement to generate more revenue.”
The key aim here is described as “more revenue” rather than “control” of course. However, state-driven or not, we are seeing a new age of a type of data-industrial complex that monetises all of our actions. Our commercial activities are monitored, recorded, and then used against us to sell us things. That feels very strongly like control to me.
Last week Sir Martin Sorrell asserted that the rise of client procurement departments had accelerated the decline of the holding company media agencies. This is of course true, but only partly explains their decline. Facebook’s global ad revenue in the last quarter of 2018 was $16.6bn. This is greater, for example, than the market capitalisation of WPP. In any battle for data, and therefore for revenue, between the marketing holding and the tech companies, it is Google, Facebook and Amazon that will dominate.
The larger holding company media agencies are simply becoming irrelevant. They lack both the capital and ultimately the depth of intellectual firepower to win this data battle. In this context, even IPG’s 2018 $2.3bn purchase of Acxiom, and the $4.4bn Publicis purchase this year of Epsilon, feel more like a circling of the wagons than any meaningful attempt to seize back the data high ground.
By shifting billions of ad dollars into Big Tech from non-digital media channels, the marketing holding companies have become architects of their longer-term media planning and buying irrelevance. This torrent of diverted money has funded levels of R&D investment that is beyond most global companies, let alone those in the marketing services industry.
With such long-term (and growing) ad-funded support, Big Tech has moved into exploring other data areas that impinge on our lives. For example, Google’s DeepMind has enjoyed ongoing access to certain NHS patient records. It has subsequently attempted to shift these records into a sister company – Streams – where governmental oversight would be greatly reduced. Their argument here is that less oversight would enable Streams to develop a global “digital health app that would be an AI-powered assistant to nurses and doctors everywhere”. Perhaps in all of this altruism there might also be an opportunity for Streams to sell our data back to us?
Elsewhere, in a development that carries echoes of the events in Rongcheng, another division of Google is in dispute with the politicians of Toronto. Google‘s Smart City project plans to regenerate a 12-acre quayside area which will involve over-arching “data harvesting” from its Sidewalk Labs division. The project is in the balance as concerns over data privacy are galvanising understandable local resistance. However, the project is still likely to go ahead.
The only difference between Rongcheng and Toronto would then be in how the surveillance data would be used. State-funded coercion leading to control on the one hand, and public/private coercion leading to sales on the other. (Incidentally, what conclusion do we draw when discussing this particular private/public project about how government interacts with the private sector, when we know that Theresa May or members from her senior team met representatives from Huawei on average once every two months during her premiership?)
98.2 per cent of Facebook’s 2018 fourth quarter revenue was advertising-based, largely funded by the marketing services industry, which has become in thrall to Big Tech. These companies have been able to invest massive sums in R&D which has allowed them to loosen their dependence on agencies. They do not have – and never have had – our interests at heart. By shifting billions of ad dollars from legacy media channels, the holding companies have weakened these channels as well as themselves. As a result, WPP etc are scrambling to redefine themselves strategically in the face of client bemusement and Big Tech indifference. (The recent announcement of the merger between Publicis London, Poke and Arc probably won’t take up much of Silicon Valley’s thinking time.)
China’s Social Credit system may feel remote and, to us in the West, implausible. But available here, amongst other data strands, are: our Experian credit records; phone usage data; company directorships; rent/mortgage payments records; DBS records; finger-print records for those who have been to the US; Oyster card travel data; DVLA records; our spending habits (and we are quickly replicating China’s cashless society); how we rate Uber (for example) and how they rate us; our unique online and ad serving history and our GPS tracking history.
We are much closer to the Social Credit System than we would admit, and this next opportunity for the Big Tech companies has largely been funded by our own industry.
Roy Jeans is CEO of Grey Scorpion.