Translation of technical results into business insights that support strategic and operational decisions.
The concept of a data lens is used to translate data into a visual representation. This is also one of the functions of a data scientist, preferably the same person or team who originally worked on the business questions. The lens has to not only translate the mathematical results of the analysis, but also make sure that it takes the form of the answer to the question. If the results are meaningful, it is also the function of the lens to provide further insight in relation to the question.
The lens is also responsible for the translation of more abstract concepts such as risk, probability and confidence factors. These concepts presented correctly to a business leader go beyond being insightful, and lead to actionable insights.
Selecting the most appropriate visual is both an art and science. There is also a healthy slice of psychology involved since it has to appeal to people, but also focus attention on specific elements of importance. Specific consideration has to be given to the target audience, both in terms of their personal preferences, but also the role.
If a decision maker has personal visual preferences, a lens can accommodate the selection of visuals and colours. A similar principle also applies based on the role of the consumer. If the role is technical then the graphs and visuals can reflect that, but if their are more strategic decisions required the views can be more abstract in nature and cover topics such and risk and predictions.
The other role of visualisations is to provide the consumer with a consistent way to measure the quality of the insights in an easy visual way. The point is to create a level of comfort with the results, so that they are used consistently over time and confidence is created to rely on them.
Decision Support Services
Ultimately most companies will invest significant amounts of money and effort into data and analytics. The industry also refers to this investment by the same name: 'data and analytics'. Without picking on the naming convention, the return on the investments made is only measured by the insights gathered from data and analytics. Going one step further, if the insights only informed and did not lead to actual actions then the investment is potentially wasted.
Actionable insights that can and should be used as decision support services is the value proposition of the entire Data Science Framework.
The key role of people involved in the interpretation in insights is not to be enslaved by the technology. The opposite outcome should be the aim - it should free people up to focus at a higher level. The confidence that is built up by repeatable good insights should lead to a focus on automation.
Any workflow that leads to consistently good insights, followed by consistent decisions, can eventually also be automated using the actual decisions to train the next model to prescribe the next best action. If the training data is good enough, the model should perform at least as good as their human counterpart. The cost benefits of such automation can justify the investment and also empower staff to do greater things.