Who are the stakeholders for Analytics within a business?

I would define the stakeholders for Analytics as those people who will be making business decisions using the results from analytics, as well as more broadly those people who will be impacted by the decisions being made. As an example I do a large amount of government consulting. In such a context there will be a wide diversity of government employees who are making different types of decisions about government programs. There will also be members of the general public who won’t be making any business decisions, but who will be affected by the decisions that government makes. As such as a business analyst or data scientist if I present analytics results in such a way that it biases the decisions made by government then I should ask myself whether I am treating the general public with respect and fairness.

 

Given this role of stakeholders I should always put effort into understanding who are the people within any organization that will be contributing data to an analytics team or interpreting subsequent analytics results. This means holding discussions during the early stages of an analytics project to produce an accurate stakeholder mapping.

 

At the same time it is crucial to involve stakeholders throughout all stages of an analytics project including:

  • Project conception – to understand the research questions where analytics might be of value (knowing that sometimes stakeholders might have a good understanding of these questions, and other times they might only have a vague sense of a decision that they are making and have not realized that analytics might be of use to them)
  • Project design – this involves describing the data that might be of benefit to an Analytics Engine as well as efforts to report analytics results back to stakeholders. Stakeholders will often be able to provide valuable insights into additional data that could be used. I will also often produce a prototype of analytics findings that I ask different stakeholders to comment on how useful they are for them (eg. if I propose building a website showing analytics results then I will produce a prototype website showing the layout of the different graphs and tables but without using real data to begin with – the prototype graphs are for demonstration purposes only. I then ask stakeholders to comment on features such as whether the website shows the results of most interest to them, and how easy it for stakeholders to interpret the graphs and tables using the format that I have shown).
  • Project implementation – just like any software development there is a recommended Agile approach. Often stakeholders need some time viewing real data and formats of analytics findings before they can provide insights into what they find useful, and what new features they would like to add. Indeed perhaps this task never ends as businesses change, different stakeholders will ask different types of questions, and new types of data sources are incorporated into the analytics.