Computer flattery

Silicon sycophants: The art of computer flattery

n. flat·tery

a : the act or practice of flattering
b (1) : something that flatters (2) : insincere or excessive praise [1]



We all like a bit of praise now and then.

Whether it’s sincere or just flattery, the human ego is deliciously susceptible to a bit of linguistic petting. Indeed it is this ability, to flatter with seeming authenticity, that sits at the heart of many successful persuasion techniques.

But since many of our daily interactions are increasingly taking place online, mediated by computers of all shapes and sizes, could flattery as a persuasion tactic be falling by the wayside?

Silicon sycophants

Apparently not.

It turns out that, for all our intellectual sophistication, humans are just as susceptible to flattery from computers as we are from our fellow warm-blooded companions.

In an intriguing study carried out by B.J. Fogg and Clifford Nass [2], 41 undergrad students volunteered to take part in the ‘Animal Game’ – a computer guessing game similar to ’20 Questions’, consisting of 12 rounds. In each round, the computer would prompt the student to think of an animal, any animal, which the computer would then have to guess the identity of through a series of yes/no questions.

If the computer guessed wrong (which it did most of the time), it would ask the student to input a yes/no question (such as “Does the animal have hooves?”) for use in later rounds, so as to ‘refine the game’s algorithm’.

Here’s where it gets interesting

At this point, the computer would ‘provide feedback’ to the player by either giving him/her:

1. Sincere praise

These students were told that the computer’s feedback would be contingent on their input, evaluated against the work of hundreds of previous gaming participants; e.g. “You seem to have an uncommon ability to structure data logically.” 

2. Flattery

This group was told that their would have nothing to do with their actual performance.

3. Generic feedback

At the end of each round, these students would only see the message ‘Begin the next round’.

Another round would then begin, and once all 12 rounds were completed (taking around 15 mins total), the students could finally fill in a questionnaire about their experience.

Flattery gets you everywhere

We may like to think that if we received ‘praise’ from a computer we’d be immune to its charms by virtue of the fact that we know it’s just a machine.

Well, this may be so, but we’re not rational creatures, and it turns out that the effect of flattery received from our silicon friends is even more pronounced than we might have predicted. Whether the students had received sincere praise OR flattery, the results showed that both groups actually responded in very similar ways:

– they reported feeling more positive
– performed better in the task
– evaluated the interaction as more positive
– had greater positive regard for the computer

A final word

Asides from the fact that they actually seemed to like the computer more, the strangest thing of all is that all this points towards one, crucial thing: we respond to flattery from a computer in the same way that we respond to it from other humans.

Which means that if you want to create a persuasive game, website, or programme, you could do worse than to throw a bit of charm into the mix.





[2] B.J. Fogg & C. Nass (1997) Silicon Sycophants: The Effects of Computers that Flatter. International Journal of Human-Computer Studies, 46(5), pp 551–561.


The following two tabs change content below.
Best-selling author of "Webs of Influence" (Pearson). International speaker, consultant, columnist at Marketing Week.

Latest posts by NN (see all)