Data Mining
Finding meaning in your data
Data should always be managed is such a way that you can extract an understanding of overall patterns - to see the wood for the trees. At an individual level, data can be overwhelming but when it is used to form a larger picture, stories can be extracted - stories that put meaning into your customer behaviour to give you a guide to your future plans. Statistics are at the heart of these data science techniques and whether based on a complete data set or a sampled data set, existing data sets are used to drive
Analytics
Tools such as Google Analytics has made tracking and visualisation of patterns of behaviour across websites for over a decade and in more recent times, specialist software has been developed to help bring data to life across all parts of the business.
Visualisation
Over this time, traditional management information reporting tools have increased in sophistication and can enable linking and virtualisation of data sets outside of their own internal storage, increasing the power and potential impact of the understanding they can drive.
Sampling & Significance
Testing for significance on a result (i.e. limiting the chance that it's just down to chance) or testing on a statistically significant sample to predict the likely future result of a campaign or action, limits business risk.
Segmentation & Clustering
Segmentation is a mathematical exercise to group "like-minded" individuals together based on the data you have. This could be their demographics, their actions with your organisation or even their wider interests.
Recommenders
Whether it's "people who bought this, also bought this" recommendations or "we have detected unusual activity on your account" anomaly detection at a checkout, prompting the "next best action" for a customer extends the relevancy of your business to that individual.