A couple of days ago I came across this article - NBC Turns 'Fresh Eyes' to Its Ad Sales (subscription required). It talked about Michael Pilot joining NBC from GE to manage Ad sales without any prior media experience. Michael was hired in part it seems because he has been "focused on using statistical analysis of potential customers to figure out which are the best prospects". The article went on to say
More than many other businesses, TV ad sales are heavily reliant on telephone negotiations, paper purchase orders and personal relationships. The industry is so old-fashioned that orders for commercials are often processed manually by a TV network
Amazing. I knew TV stations were notorious for being technology luddites but even so... This range of problems seems ideally suited to the technologies and approaches of Enterprise Decision Management or EDM. Business rules are already in use at several media outlets for handling ad pricing. The rules allow even complex sets of conditions to be applied to come up with the right price, freeing employees to spend time with clients rather than grinding through the mechanics of calculating the ad price. Business rules also deliver agility, allowing the rules to be changed and managed more aggressively and consistently. Clearly predictive analytics could be applied also to create some dynamic, demand-based pricing. Meanwhile actually slotting the ads purchased into ad breaks is another great rules-based problem, and one where inferencing really matters - check out this story about slotting. Optimization, another EDM technology, could also be applied to manage slotting options. Meanwhile analytics are getting used in keyword bidding and online ad issues are creating some interesting opportunities.
EDM would be perfect for NBC, it even has the same number of letters...
