Stakeholder Engagement For Data Science Projects

Adequate and effective engagement between data professionals (Data Scientist & Analysts) and business stakeholders is crucial for generating the right or best data solutions. However, this engagement is often overlooked by companies that are new to data science and also by data professionals in the early stages of their career; this is primarily because there is the notion that data solutions can be successfully developed as long the data professional has the required data and knows at a surface level what the project objective is. In this article, I explain why stakeholder engagement is important and share tips on how data professionals can engage with business stakeholders throughout the life cycle of a project.??

Starting a project?

Data Science projects fail or are not as successful as they can be when data professionals and business stakeholders do not spend enough time at the start of the project working together to clearly understand and frame the business problem to be solved.??

For example, a data professional working in a health insurance company might receive a project request to develop a model to predict who is likely to get diabetes. The request sounds straightforward enough so the data professional jumps right into building the model only to realise after completing it that there are different types of diabetes - Type 1, Type 2 and gestational which occurs only in pregnant women - and the focus for this model should have been Type 2 since it is the most common among the company's customer base. If the data professional had had further engagement with the business stakeholders before building the model to learn more about the context for the project and how the results will be used, this detail would have surfaced thus allowing for a more tailored and appropriate model to be built.

This is why adequate engagement between both parties to start off a project is critical. It provides clarity and alignment on the project context and objectives which

• ensure that the right or best solution is developed
• cut down on the need for rework at the end of the project

Questions data professionals can ask to get a clear understanding of project request include:

• What problem are you trying to solve??
• Why are you trying to solve it?
• How will the results be used?
• Are there any specific metrics / data points of interest?

Executing a project?

As a project is being executed, it is important to keep the stakeholders up to speed on any progress made, communicate any barriers as soon as possible, and share intermediate results, where appropriate, without overdoing it. This is beneficial because:

• Stakeholders are more likely to understand, buy into, and leverage a data solution when they are involved throughout the life cycle of the project.

Finishing a project

The final stages of a data science project involve sharing results with the business stakeholders. In my article, Growing In Your Data Science Career, I share tips on how to effectively communicate your findings under the “Data Storytelling / Presentation” section. In addition to this:

• Share actionable recommendations based on your findings whenever possible. For example, a finding might be that on average a hospital runs out of a particular medical supply within two weeks. Additionally, on average it takes 1 week to get new stock of that supply from the time it is ordered which is too long. Possible recommendations could include a) placing the order for new stock earlier or b) switching to a supplier that can deliver faster. Sharing recommendations demonstrates your commitment to helping your business stakeholders make better decisions and builds trust.?

Soft Skills in stakeholder engagement

Due to the nature of the job of data professionals, there is a strong emphasis on the hard skills required to execute the job. However, given the needed collaboration with business stakeholders, soft skills are just as important. For example, if you have a different perspective from a stakeholder on how to approach solving a problem, rather than saying “I disagree with your approach” and then sharing your perspective, you could instead propose your thought in the form of a question: “What do you think about this?” which gives the stakeholder an opportunity to reflect and respond; the latter is more likely to be well received than the former.?

Data professionals should take steps to work on their soft skills as it impacts their ability to collaborate with stakeholders to successfully execute a project.

Conclusion

While a majority of the work involved in developing data solutions falls on data professionals, the reality is that these solutions are being developed to meet business needs within different contexts so effective and adequate collaboration with business stakeholders is needed for the development of successful solutions.?

To all data professionals who have experience collaborating with stakeholders to execute projects feel free to share:

• an experience which highlights the importance of having adequate discussions with stakeholders before executing a project?
• an experience in which continued collaboration with stakeholders prevented a project from failing?
• important lessons you have learnt in doing so

Thank you for reading! Don’t forget to share this article with your network if you found it helpful! ??