@mappingtools
By plotting locations of objects or events, it is possible to use Microsoft MapPoint to perform a visual cluster analysis. To do this, all you need is a list of your objects, and their coordinates or addresses. Then use the Data Import Wizard in Microsoft MapPoint to read the data. By plotting the locations, you can see possible relationships in their distribution.
Our MPCluster add-in can be used for a more automated and sophisticated approach. MPCluster will identify multiple clusters within various constraints as to cluster size and distribution.
A good application of cluster analysis, would be to analyze your customer locations. If you discover that your customers are grouped in a particular location, you can investigate why. This information is then invaluable in your future marketing efforts. You can increase your marketing efforts, or identify future markets that have similar characteristics.
You do not need to use the default black pushpin symbols. Instead, you can use different pushpins sets with different symbols and/or colors for different attributes. For example, you could plot different delivery times or different levels of customer satisfaction as different pins. You can then quickly identify problems, fix them, and find out the root causes.
The lower map is John Snow’s famous map of the 1854 Broad Street Cholera Outbreak in London. By plotting data on a map like this, John Snow was able to identify that the outbreak was centered on a specific water pump. Interviews confirmed that outlying cases were from people who would regularly walk past the pump and take a drink. He removed the handle off the water pump and the outbreak ended almost overnight. The cause of cholera (bacteria Vibria cholerae) was unknown at the time, and Snow’s important work with cholera in London during the 1850s is considered the beginning of modern epidemiology. Some have even gone so far as to describe Snow’s Broad Street Map as the world’s first GIS. Although this is clearly an exaggeration, it does demonstrate the potential power that maps have when analyzing complex data.
John Snow only had one cluster to identify and diagnose. Modern sales data usually consists of dozens, if not hundreds, of clusters. This is where tools such as MPCluster can be used to identify clusters in your MapPoint data, today.
