With shrinking budgets, changing consumer expectations in a hyper-connected age, as well as constant pressure to extract maximum business value from big data and deliver strong return on investments, the Central Marketing Officer (CMO) is fast becoming the catalyst to help the enterprise’s data science evolve to be business science.
Data science in isolation, without business value in sight, can reveal very interesting patterns, trends and observations. However, the best data science is purposeless unless the data science and the data scientist can recommend actions for the business owner with the express purpose of creating value.
What does it take to evolve your data science into business science?
Lack of action is problem #1. The most severe problem facing several enterprises is not that they have insignificant IT investment in big data technologies or that they lack the resources to build dashboards and reports. The real problem facing enterprises is the inability to traverse the last mile of data science and to convert interesting observations and insights into value and generating action.
Arm your data scientists with the business context. Ensure that your data scientists do not work in isolation and that they interface and work very closely with the business owner and the product managers. Data scientists need to understand the business drivers, business critical issues, and the enterprise and product strategy.
Capture business state as KPIs. Push your data scientists to implement business-focused key performance indicators (KPIs) using the data that is being generated through the use of your products and services. Encourage your data scientists to fill the gaps in your instrumentation required to implement the defined KPIs.
Encourage insights that predict a result from a recommended action. Encourage your data scientists to deliver insights that take the following shape:
Predictions of enhanced business value
Demonstrated through desired movement in business KPIs . . . when
Recommended Actions are implemented.
Over time, following a framework like this and taking advantage of newly-arrived data in the feedback loop builds an ever-improving knowledge base of “actions” and “effects”, which in turn enables and informs more accurate and efficient data analysis in the future.
Working to a model like this allows all the insights that are created to be actionable and further allows your organization to evolve to practicing business science as opposed to data science.
Check out Apigee Insights to learn how Apigee approaches data science and providing end-to-end visibility into your digital business.