The classic two-step hypothesis was rejected by its creators. Subsequent studies have demonstrated shortcomings in the data on which the hypothesis was based. For one thing, the dichotomy of opinion leaders and followers was shown to be unfortunate because the 'followers' proved to comprise two very different categories - information seekers and initiatives. The Marknadssociologen study abandoned this dichotomous definition and exploited the seeking and advising activities instead. The finding of a strong positive correlation between these two activities demolished one of the corner-stones of the two-step hypothesis. One and the same person was in fact frequently an opinion leader and a follower. It was also found that topically active persons were also high on both topical exposure to advertising and topical personal communication. The dynamic multilateral hypothesis indicated that "shared interest seems to be a channel through which communications flow". Interest proved to be a suitable variable for defining the target group for a message, because interest is strongly correlated with both seeking and advising activity. Interest has been used in practice to define the persons who are likely to disseminate the message most effectively, the persons who fill the same function as the classic opinion leaders.