What do you do if your data science project stakeholders are not understanding your progress updates?

As data scientists, we communicate across a vast array of audiences with a high variation in skillsets and goals. Therefore knowing your audience is crucial before providing them with progress updates. Just as we don't like noise in our data, our updates can be noise for stakeholders if they're not targeted to goals and missions that are important to them. Ensure that stakeholder presentations are specific to a set of stakeholders, with the appropriate level of detail and business impact. Senior stakeholders typically are not interested in the inner workings, and would prefer high level updates on business impact, while project owners and end users would prefer knowing the inner workings and challenges faced at present.

It is essential for Data Scientist to convey the results to stakeholders in bite-size manner, as developing insights in silos will have less impact. Even though, it is natural to engage ourselves in developing state-of-the art algorithms, if these are not adopted by stakeholders, it will just be an addition to the attic Avoid Jargon: Use clear, simple language instead of technical jargon. If you must use technical terms, ensure you provide concise explanations Use Analogies and Examples: Analogies and real-world examples can make complex concepts more relatable and easier to understand Link to Business Value: Always tie your updates to the broader business objectives. This helps stakeholders see the relevance and value of your work

Understanding your audience is crucial when preparing progress updates for a data science project. Tailoring your message to meet the diverse needs and expectations of stakeholders ensures effective communication. For instance, sponsors may seek high-level objectives and risks, while end-users desire specific features and benefits. Adapting the message format and avoiding technical jargon ensures clarity and relevance, engaging stakeholders and fostering meaningful collaboration throughout the project lifecycle.

Visualisations when done right, can be exactly what is required to convert complex technicalities into easily interpretable information, especially when stakeholders have limited time. As with any presentation, focus on building a data story, focusing on key points being presented through visuals. Use key metrics graphs to quantify progress towards goals, represent improvement from the original baseline as well as projected improvements. Ensure visualisations follow visual theory concepts, avoiding complex visualisations that defeat the point by being too hard to digest.

Utilizing visuals and stories enhances the effectiveness of communicating data science progress. Visual aids like charts, graphs, and prototypes elucidate data, methods, and findings compellingly. Meanwhile, narratives, scenarios, and testimonials bring relatability and clarity to problem-solving and value propositions. Ensuring relevance, accuracy, and conciseness in visuals and stories addresses audience questions and concerns effectively. This approach captivates attention, facilitates understanding, and fosters meaningful engagement, enabling stakeholders to grasp project intricacies and outcomes with clarity and enthusiasm.

High order of transparency is key to communicate results and impacts such that stakeholders can fully understand how utilise what you've built. Therefore explaining assumptions and limitations to the capabilities of the methodology is crucial for insights and actions to be reflective of the reality. Communicating this especially to a non technical audience can be challenging, but achievable through clever use of analogies, visuals and data and research by other companies as well. Through explaining the why of the assumptions and the how of the impact of limitations, you can enable a more nuanced discussion of the what, deciding the best course of action based on findings, while explaining all the mitigating factors and validations.

Transparently explaining assumptions and limitations is crucial in communicating data science progress and results. This includes addressing data quality, methodological choices, and outcome validity. Being honest about potential impacts on project scope and performance is key. Additionally, outlining validation and mitigation strategies demonstrates accountability. Providing contingencies for changes or challenges ensures readiness and fosters trust with stakeholders. This approach promotes clarity and confidence in project outcomes, facilitating informed decision-making and effective collaboration.

With insight driven and predictive projects like data science, getting feedback and thoughts from stakeholders with experience, and industry knowhow is highly important for fine tuning project quality, and enhance business impact. Ensuring feedback can be tricky, but building a feedback first and a continual improvement mentality between your stakeholders and teams is possible through continual progress check ins, project demos and workshops as well as knowledge sharing sessions wherein the floor is open to discuss the project direction and impact. Frank discussions in these sessions have really helped me and my team to refine the scope of our projects and deliver something that in the end is more useful for the business

Soliciting feedback and collaboration enriches data science projects, enhancing quality and relevance. Inviting stakeholders to share input and perspectives ensures alignment with their needs and expectations. Acknowledging and incorporating their contributions fosters a collaborative environment. Demonstrating how their input contributes to project success strengthens engagement and fosters a sense of ownership. This iterative process not only improves outcomes but also builds trust and fosters meaningful partnerships, ultimately driving project success and impact.

Simplify progress updates for stakeholders by emphasizing key insights, outcomes, and tangible impacts. Use visualizations and non-technical language to enhance clarity. Solicit feedback and adjust communication based on their understanding, ensuring alignment between Data Science progress and stakeholder expectations.

Taking a top down approach always helped to communicate effectively with technical and non technical stakeholders. Start with impact of your work in terms of time-cost-quality translated into dollars. Have few points on your findings/recommendations/results. Be open and humble to accept feedback and collaborate with their ideas/approaches. Depending on audience and time, dig into technical details. Prior to initiating project, define success criteria for each phase/milestone of the project.