
Foursquare just passed the million user mark with 40 million check-ins. This explosion of location-based social networking is generating a wealth of data that can be used for smarter marketing. With the availability of data, arises the need to integrate, synthesize and ultimately visualize this data to derive patterns and relationships and most importantly, insight.
Integrate & Synthesize
Often referred to as spatial intelligence, the ability to discover and analyze relationships using location as a dimension and integrating it with other business intelligence elements can lead to data nirvana. However, the biggest potential of location-based data is the insight that it can offer to what might happen in the future. (Feel like Nostradamus yet?)
Predicting patterns of behavior based on ‘check-in’ behavior can help marketers achieve what they have always strived for: being in front of the customer at the exact time the need exists. Think contextual marketing (but defining context became a challenge, think behavioral targeting (‘uniqueness’ in online behavior is debatable). How about trip-based segmentation and targeting based on Foursquare/Gowalla streams? Trip-based predictive segmentation is not new especially in the traditional DM world built on the RFM model. Imagine adding a content layer to that model, with attitudinal intent. I wouldn’t be surprised if we start talking about persona-based segmentation and then clustering various personas into communities based on the user’s paths and check-in trajectories.
Mary, the social butterfly
John, the ‘living-out-of-a-suitcase’
- Mayor of home airport and checked into 12 other airports
Jack, the responsible husband
- Bi-weekly trips to the grocery store, bank
Visualize
Powered with the ability to segment location-based data and its predictive possibilities, how do you manifest it into a compelling visual? Let’s look at of one of the most famous information design artifacts: Charles Minard’s representation of Napoleons retreat from Moscow in 1812 and understand how we can apply some of Tufte’s fundamental principles to location-based data visualization

More than 2 dimensions (multivariate)
Current location-based visualizations are map-based and some have attempted the geo-temporal dimensions. Austin Vicarious.ly‘s time-lapse visualization of check-ins during SXSW had two dimensions, time (temporal) and location (geo). On the other hand, Minard’s representation has multiple dimensions (army size, location, direction, time, temperature).
A multi-dimensional approach to location-based data visualization can be to integrate time, location, size (based on check-ins), pre-defined content clusters or segments, non-spatial attributes (financial, economic, competitive), predictive variables like propensity to check-in etc. The possibilities are endless.
Mix words/images/numbers
If an info-graphic can survive without the help of a key, it has done its job. Often times ‘content’ and ‘data’ are not uttered in the same sentence, because of the perception that data is ‘structured’ and content is ‘unstructured’. If the visualization can marry the two, the result of the insights can be far more than cursory.
The ultimate potential of location-based data and visualization is the ability to identify patterns, causality and relationships to formulate smarter strategies. It will be interesting to watch how this plays out for business intelligences providers, marketers who want their hands on this data, marketing services companies that will emerge and of course for Mary, the social butterfly who will be leaving much more than a digital footprint the next time she checks into her favorite club.
Knowing where something happens can often lead to knowing why it happens.