Classifying users as ‘givers’ and ‘takers’

It might seem we have all sorts of options for tackling inappropriate content on Chatroulette. But the larger the scale, the harder it is to solve apparently small problems.

With millions of users, any changes we make must be resilient enough that there will be no adverse effects. That’s why it pays to think through an overarching model that’s sufficiently theoretically solid and doesn’t rest on chances.

Psychologist Adam Grant’s notion of ‘givers’ and ‘takers’ is hugely helpful here. We’re finding it useful to think about our users in terms of those who bring more value to the site than they take (givers) and those who take more value than they bring (takers).

Whenever we let a user onto our platform and they provide negative value – by making others no longer wish to use the site – Chatroulette is threatened. So our basic long term goal is to have a community solely populated by givers or, at least, ‘matchers’ (those who give and take in equal measure).

This framework allows us to solve a more general problem, rather than fretting about particular behaviours. The challenge is not to solve the issue of pornographic content, for example, but to answer the wider question of giving or taking.

With millions of users, we can’t enumerate all the ways people might have a negative influence on others. But by creating this concept of internal currency (a karma/reputational system) then we can identify and get rid of the worst users, and identify and reward the best.

We don’t want to play the role of censor or get involved in moral questions. Instead, we can turn to our users, allowing them to inform us about their preferences through direct and indirect means, like by reducing or increasing their activity when they meet a certain user.

While inappropriate content is a complicated area, this simple ‘givers and takers’ model is how we’re currently assessing most of our ideas.

AT