Machine Nudging

On the train from London to Brussels, the passenger sitting next to me suddenly gets agitated. I look at him puzzled, as this is a wagon entirely populated by people traveling for business and sudden noises or exuberant behaviour are noticeable like being dressed in a pink suit. My travel companion apologises and adds:”Damn, I forgot my phone in the station lounge. It’s a tragedy, it contains all the reminders for my pills”. After such statement, I can only be sympathetic.

Lately, I have been working a great deal with computer scientists and when I end up chatting about behavioural research and nudges, they usually look unimpressed. They often reply that they have been nudging for a long time in the design of user interfaces. Of course, it is not as simple as that, but I have been amazed by how much of the scientific literature on usability and human machine interaction contains overlaps with decision making research, in particular about the use of nudges.

As the online world becomes the most common and frequent context of information choice architecture in developed countries, the concepts and experience developed in decades of research related to improve usability of interfaces are a rich resource. In a way, traditional usability research used crude models of human decision making processes, but at the same time made large use of small scale empirical tests that informed them about valid solutions without the need of a sophisticated theory of human decision making.

But things are changing. This is what a recent book for web designers states:

The days of purely aesthetic design are long gone. Today’s web designers are driven by pertinent questions like these: How will I win the battle of the short attention span? How do I put visitors at ease and provide the information they’re consciously (and unconsciously) expecting? How will the design of my site encourage users to engage, browse, or buy? There’s a body of tested psychological principles that can transform your digital designs by anticipating and benefiting from how human beings react to stimuli.

As other recent publications, a book by Chris Nodder is even more explicit in its book cover ‘Evil by Design. Interaction design to lead us into temptation’ [1].

The world of ‘online nudging’ is already a reality. The necessary dialogue between psychologist or behavioural scientists on one side and computer scientists and designers on the other has just started and does not receive the coverage that it would require.

It is a grey area in which big online corporate players appear to be very interested. To recognise the potential, think about the recent study conducted on Facebook about emotional contagion (with all its limits and ethical issues) by Kramer and Guillory [2] to witness a massive nudging (as an experimental manipulation) online.

The pioneering work by B.J. Fogg introduced some years ago the idea of computers as persuasive agents and it should not surprise if the next step were to consider how machines (such as algorithms) are nudging machines. Of course, people design machines that nudge other people. Except that more and more machines design other machines…that can dynamically respond to users. Well, I don’t want to end with a Terminator scenario so I’ll stop here.

Online nudging is happening and my suggestion is that this is a domain in which we should learn from what human machine interaction research has to teach us before trying to reinvent the wheel and to be aware of the level of sophistication that already exists.

You will never surf the Net again with innocent eyes (if you ever did) now.


[1] Nodder, C. (2013). Evil by design: Interaction design to lead us into temptation. Indianapolis, IN: Wiley.

[2] Kramer, A., & Guillory, J. E. (2014). Experimental evidence of massive-scale emotional contagion through social networks. Presented at the Proceedings of the ….

Originally published at

Show your support

Clapping shows how much you appreciated Giuseppe A. Veltri’s story.