The
speech begins with the speaker, Sebastian Wernicke, explaining to the audience
that Roy Price was the senior executive with amazon studios. Roy is in charge
of finding TV shows for amazon. Sebastian then goes on to explain that Roy
needs to find the right TV show for viewers, a well written, addicting TV show
if he wants amazon to be successful. The first thing that Roy has to do is
select eight shows that are deemed worthy candidates. Then he puts the first
episode of each one of these shows online for free. Because the episodes are
free, millions of people will see them. While the people watch the episodes,
Roy and his crew record the viewers’ actions: when they press play, when they
press pause, what parts they skip, or what parts are re-watched. They then
record the data that they receive and decide which show would best succeed.
Sebastian
explains that one of the first shows that Roy had decided on based off of the
data they collected, actually turned out to be an average show, instead of a
great big hit like they were hoping for. Sebastian then tells us that at around
the same time another man that worked for a different company, Ted, was
collecting data in a similar way. Instead of picking candidates, Tom and his
team looked at data that had already been collected from Netflix viewers to
decide on a show. Then, Tom and his team use that data to pick out information
about their audience. What kind of actors, the audience liked, what kind of
genre, etc. They then use that data to make a VT show that was very similar to
Roy’s. Toms show was a much bigger success.
Them
Sebastian asks what happened? The data was similar, why did it not work. If
data analysis doesn’t work, what will? Data analysis does not always work.
Although it often does, it’s not a perfect system. There is a pattern that
shows up in data analysis that works and data analysis that doesn’t. Data
analysis is only really good for taking a problem apart, but not so good with putting
together an answer. He then says that he thinks that things begin to go wrong
when people let data completely drive their decision making. If we really want
to be successful, we need to be able to make decisions for ourselves and take
risks.
Comments
Post a Comment