TRC

The popularity of the TV sitcom "Seinfeld" and the imminent end of its run has generated much news and discussion. From a marketing perspective a key question is how the show's current ratings can be leveraged in the future (outside of syndication).

A popular option is spin -off shows based on one or more of the characters from the original. Anecdotal evidence indicates that a show based on the Kramer character is the most-preferred choice.

This may indicate, at first glance, that a show based on Kramer is likely to be the most successful. To investigate this question and to quantify some of the dynamics underlying the show, we conducted a short study. The results show that using analytical methods, we can arrive at conclusions not possible when we just look at percentages.

The telephone survey was conducted with a national random sample of 667 people, of whom 368 (56%) indicated they watched "Seinfeld" at least sometimes. Among these "Seinfeld" watchers, Kramer was by far the most popular character, being chosen as the favorite by 56% of the respondents. Jerry(18%), Elaine (13%) and George (12%) were bunched together and considerably behind.

Further, when asked how funny each of the characters was, Kramer came out on top again with 80% considering him "extremely" or "very funny." Jerry (56%), George (51%) and Elaine (43%) were well behind Kramer. Moreover, unlike the other three characters, Kramer's attraction remains unchallenged regardless of age, income or gender group. This would seem to indicate that Kramer has the most impact on the show and that a spin-off show with Kramer would be the most likely to be successful.

However, analyzing the data using key driver analysis (i.e., regression analysis) gives us pause.

While Kramer is notably popular, the impact that each of the four have on the show (as measured by the importance weights attached to each character from the key driver analysis) is almost equal.

In fact Kramer, George and Jerry have equal impact on the show's rating with Elaine only slightly behind, indicating that it is indeed an ensemble cast with no one character's impact dominating the ratings.

Another way of looking at this is that Kramer's huge popularity does not mean that character has a large impact on the show's ratings. Further, among frequent viewers of "Seinfeld" (the most likely the target audience for a spin-off), Kramer's impact on the show is the same as Elaine's and George's. What does this imply about the success of spin-oft shows featuring each of these three characters?

Presumably a spin-off starring Kramer would try to capitalize on his popularity and would therefore not change the essence of the character. The question then is, would the spin-off show be dominated by the Kramer character or would it be a strong ensemble show like "Seinfeld?"

If the former is the case then there is a chance that Kramer's popularity could t ranslate to the new show, making it a success. However there also is considerable risk in this approach for two main reasons: a show based on one character loses the synergy that can be gained from an ensemble cast, and the Kramer character's inherent quirkiness may make him unpalatable as a strong main character.

Assuming the show does have a strong ensemble cast, then—as indicated by the regression—there is no reason to believe that a similar show starring the George character would not be equally successful. After all, if the Kramer and George characters have equal impact on the show's ratings, it means that the individual ratings or popularity of each character taken in isolation does not matter much. It is their combined strength that matters.

If we are considering only frequent "Seinfeld" viewers, one could even extend the argument to include Elaine. It is quite possible that the show featuring Kramer would be more successful initially because of Kramer's huge popularity, but over the long run (such as a season), the show starring George should be equally successful.

This study illustrates the usefulness of conducting multivariate analyses such as regression to obtain insights from data that may not be apparent looking just at percentages or cross tabulations.

There are many situations in which managers believe that just asking a customer what is important about a product or service is enough to provide information on the drivers of satisfaction. While this may be true in some cases, relating the performance of the company on specific attributes with the customers' overall satisfaction often provides more useful information about which attributes really drive satisfaction, and therefore are more important.