April 5, 2017    5 minute read

Why Facebook Should Introduce Bespoke News Feed Algorithms

DIY    April 5, 2017    5 minute read

Why Facebook Should Introduce Bespoke News Feed Algorithms

Although Facebook is the world’s largest online social network, the tech giant has increasingly run into various problems and controversies. It would be fruitful to examine some of these problems and their impact on peoples’ wellbeing as well as their inclination to be continuing, loyal users of the service.

To combat this, the author suggests that Facebook could open up the platform to developers and individuals to implement bespoke, ‘DIY’ (do it yourself) algorithms to determine the layout, content displayed, news feed, etc. in order to harness a variety of benefits.

Facebook has several problems: quitting, unhappiness, lack of control

A 2015 study conducted in Denmark by the Happiness Research Institute found that, overall, “Facebook users are 39% more likely to feel less happy than non-users”.

Another article published by psychologists from the University of Michigan, Ann Arbor and the University of Leuven, Belgium found that “[on] the surface, Facebook provides an invaluable resource for fulfilling the basic human need for social connection. Rather than enhancing well-being, however, these findings suggest that Facebook may undermine it.”

It was reported on ZeroHedge that “it appears Facebook’s new algorithmic censorship practices are causing some of its employees to quit… [because Facebook has] a new tool developed specifically to restrict certain kinds of posts from appearing in users’ new feeds in certain geographic areas”.

Even more broadly, although Facebook has seen year-on-year increases in the number of active users, individuals are seeing many other reasons to quit Facebook such as its impact on productivity. Finally, Facebook and its users should rightly be concerned with the reported ‘lack of control’ experienced by the community.

Optimisation Would Spawn New Industries

Instead of the users simply telling Facebook that they do not like particular advertisements, posts, reacting, liking and so on, the users themselves could program their own algorithms (or hire others to do so) instead of relying on Facebook’s algorithms to translate their preferences indirectly.

For example, many developers could work on Facebook’s platform to customise the way it is used on Android, iOS, Macs, Windows, etc. so that their clients have more control, they can decide exactly what sorts of information they want and thereby increase their productivity and overall satisfaction with the social network.

This would then spawn many employment opportunities related to making Facebook optimised for each individual’s purposes. Essentially, Facebook can become a tool for people to accomplish their objectives.

Of course, Facebook could still specify that a certain number of advertisements will have to be allocated in a certain amount of space but this can be customised by the user and there the tech giant would certainly not lose any money.

The Power of ‘Revealed Preferences’

If the purpose is to gather as much useful data as possible, enabling bespoke algorithms on Facebook would accomplish this better rather than through highly generalised algorithms that learn from user inputs and behaviour that are highly impersonal and can even be injurious. Essentially, people’s preferences are hard to discern, but if one gives them a channel by which to directly act on their preferences, more can be learnt from that.

Where people have greater choice, a lot more can be learnt from those choices. For example, if an individual wants to arrange their News Feed into different streams, change the layout and adjust the way in which content is received from various channels, groups, pages, etc. depending on the times of day it is accessed and so on, Facebook would learn a lot more about the objectives, preferences and motivations – this makes for a far richer and deeper dataset (albeit one that is more difficult to analyse).

Revenue and Productivity

Given the fact that the ‘revealed preferences’ would then produce more data from which Facebook could learn about its users, the various entities using Facebook would benefit from an intermediary that could anonymise and handle their data and, therefore, improve their productivity.

For advertisers and businesses, Facebook could tailor and target their advertisements further (thereby improving their accuracy and efficiency as well as simultaneously increasing the advertisers’, Facebook’s and the business’ revenues).

For users more broadly, this would mean that they would have greater control over their Facebook interface, content, etc. and to decide how to use the information exchanged through Facebook to enhance their own productivity. This would be a win-win situation for everyone involved in this transaction of information and services.

Concluding Remarks

Facebook (and, indeed, other social media giants) must realise that most people are aware that they are its products (or, more strictly speaking, their data is). In the same way that Facebook disrupted traditional media and social networks, it too can be disrupted through the ongoing torrent of technical progress and innovation.

Although it currently benefits from network effects, networks can self-implode if one is not careful and one of the key factors that might discourage Facebook introducing DIY algorithms is that it could further reveal the extent of the data collected by Facebook on individuals (thereby exacerbating privacy concerns).

Sooner or later, Facebook needs to realise that its users deserve control. It is in Facebook’s interest, the economy’s interest and society’s interest that it gives users more direct control.

Get articles like this straight to your inbox each morning with our Breakfast Briefing. Sign up by clicking here!

Log in with your details

or    

Forgot your details?

Create Account

Send this to friend