Applying Traditional Data Mining Techniques to Social Media Strategy

There is a small but growing number of bloggers, Tweeps, and other social media professionals who are beginning to discuss traditional database marketing and mining techniques as an important part of social media strategy. I’m one of them, and I think this group is right on.

Data mining – the extraction of information, often unseen information residing in large databases, for the purposes of predicting behavior, is a powerful technology. Data mining tools and the data mining process allows businesses to make informed, data-driven decisions, and possibly predict future customer behavior.

An important component in mining data is the concept of modeling. Modeling is fairly straightforward – you build a model out of positive information. Put another way, you plot all of the points or characteristics of something successful each time it is successful, and use this data model to predict the likelihood of future successes. If I’m a fisherman and every time I catch a fish I make note of the time, temperature, bait, wind, water conditions, season, and which pocket my lucky coin was in, chances are that I will build a “profile” of success. Now by using this profile and understanding that if I fish in a manner consistent with my model, chances are I’m going to catch fish.

This brings us to the importance of “flagging” your social media customers in your database (I’m making a critical assumption here – that you have a customer database). Understanding which customers are fans on Facebook , which customers have downloaded a mobile application, which customers have possibly written reviews or provided ratings, and which customers have responded to a blog post are all indicators of specific customer behavior. A profile begins to emerge – a cluster of customers who prefer to interact with your brand in a way that is different from other clusters. From this interaction and using the techniques discussed above, you can build a predictive model. This model can then be used to make data driven decisions on marketing strategies to address the huge migration of customers to social media channels that we will all experience in the coming months and years.

To us, social media marketing is no longer an “early adopter” business model. However in mainstream businesses today, the social space is still largely “uncharted waters”. Implementing tried and true database marketing techniques is one way to help our businesses predict the future based on a small but growing group of “early adopters” to our social media marketing efforts.

This is going to be a continued area of focus for me in the coming months. I’ve challenged myself and my team to think of ways to implement these ideas in order to continue to develop the Social Media marketing model for our partners. I would appreciate your thoughts and challenges.

*Photo credit: NIOSH (flickr)

Related articles by Zemanta

• This week’s #Socialmedia Tweetchat Topic: Social and the New Model For Market Segmentation #sm48 (directmarketingobservations.com)
• Interpreting your analytics with eyes wide open (solutionset.com)

Tagged as: Business model, Data mining, Database marketing, Databases, marketing, social media

This article is from Greg Chapman

Greg is known for his ability to think creatively about new solutions to conventional problems. He’s intensely collaborative in his approach to business dealings both internally and externally and is thought of as both an idea person and a team player.

Greg has written 8 posts.