SPSS on what your customers tell you #pawcon

Eric Brethenoux of SPSS was up next talking about analytics using what your customers tell you. Customer segments are inherently blurred because they are a composite. Identifying the people behind a segment is hard and increasingly more difficult as the number of brands, devices, channels, media and social networks are exploding. And people have multiple online personalities depending on their hobbies, their work, their online communities. At the same time people are using TripAdvisor or Yelp and their friends and families to see what other people have said before committing to a hotel, a service, a product. Yet companies are not mining this data very much.

In a customer relationships there are often "moments of truth" that have a disproportionate impact on a relationship. Eric gave an example of AA using a bad moment (a delayed plane) to make an offer (3 months of free lounge membership) that resulted in long term benefit (he renewed his lounge membership when the offer expired). Positive moments of truth result in growth, new business, more than 85% of the time while negative ones result in a loss of business more than 65% of the time. Analytics, especially text analytics and social network analysis, can help companies find these moments of truth and the kinds of things that result in positive or negative moments.

In some research IBM is looking at the sentiment, trends or buzz and key authorities in social networks. This tool derives sentiment and relevance of blogs or twitter streams and allows you to identify authority as a result. IBM has also been analyzing its own, internal social networks. In this case they focused on seeing if more developed networks generate revenue and found that, for instance, Systems Engineers show a positive impact for strong ties within the internal network while consultants do not.

Eric ended by talking about IBM’s smarter planet initiative where they talk about an instrumented, interconnected and intelligent world. Devices are connected and we are increasingly talking about an internet of things that know where they are and can be connected. Analytics, the ability to turn all the data being exchanged into predictions, is a critical element of these smarter planet initiatives.