No matter if you want to use #AI (#Artificialintelligence), #machinelearning, or #businessanalytics, you need to find and define the right problem. A well-established problem that clearly defines why and how any or couple of these big data management tools are to be used is the first and most critical step in the whole process of improving the decision making, by digesting big amount of data.
Considering above mentioned criticality Henke, Levin and McInerney in an article on #HBR (Hravard Business Review) summarize the main steps of translating a business problem into analytical use case as follow:
1- Problem identification and prioritization
Ask from business. They are always occupied with numerus decision making questions which keep them awake during the night. First step of any analytics is to understand problems, scope them and prioritize them from the business impact/case. Within this step, it is also important to stay focused on problems which can be addressed by analytics
2- Data collection and data preparation
Access to data and understanding the complete scope of data that is needed to address the analytics problems are often the most difficult things to do in an organization. Second step is to identify what kind of data, for the problem at hand, is needed to be gathered and from where in order to create most useful insights for business
3- Create analytics engine
The data analytics team is responsible to create analytical solution that can address business problem in the most efficient way, so that business users can have proper and timely interpretation of the results of the analytics
4- Validation
Data analytics team needs to make sure that the results are well translated back to business as actionable business implications. Collaboration of the translator in the data analytics team with the business is highly critical for further proper execution of business implications and creating the buy-in from the business
5- Solution implementation
Last step is to execute decisions based on the derived business insights from the analytics case. Learnings and change management is of crucial considerations in this step to improve the success rate and efficiency of similar practices in the future.
All in all, for a successful data analytics practice in solving business cases a well functionating team of at least an analytics translator, data scientist, and data engineer is essential. While some of the skills of these professionals are overlapping, each are mastering a specific crucial aspect of successful implementation of an analytics business case.
Whether hiring external consultant or using internal resources, your organization need specialized resources in each area to master the data analytics.
What skills you have in your data analytics organization, and how you obtain the skill sets your organization requires?
This article has major inputs from the following articles on HBR:
Comments