An analysis of Twitter messages indicates that users' mood swings correspond to time of day and season. Could this knowledge help companies to target their advertising messages more effectively ?

The micro-blogging service has now proved useful in identifying the state of mind of a human population. And it appears that, irrespective of country or culture, we can observe great similarities in individual moods. Scott A. Golder and Michael W. Macy, from the Sociology Department at Cornell University in New York State, analysed over 500 million ‘tweets’ posted online by some 2.4 million users from all over the world.  They then categorised the tone of each message – happy, sad, angry, gloomy, etc - and logged them by the date and time they were posted.

Mood follows a precise rhythm

The results of the Cornell study show that whatever the country or type of population group you look at, the mood characteristics are identical depending on the moment they tweeted the message. The first conclusion is that users are happier and more receptive just after waking, and again in the evening after 6pm. Contrary to what one might have expected, this trend holds good during the weekend as well and the phenomenon is not due to work start and finish times either. Nevertheless it is worth noting that, on average, people are in a better mood at weekends.  However, the season – especially the length of the day – has a definite impact on mood.

An opportunity for companies to target customers more effectively

Unsurprisingly, the winter season, when the days are shortest, proved to be the period least conducive to expressions of happiness.  The study may be useful to companies in that it points to the fact that people are happier – and therefore more receptive – at certain times. A happy person will tend to be more receptive to an online advertisement than a person in low spirits.  So, it’s up to commercial enterprises to adjust their advertising campaigns so as to align them as closely as possible to their customers’ biological rhythms.