Stakeholder Analysis
Week 3 | Lesson 5.1
LEARNING OBJECTIVES
After this lesson, you will be able to:
- How to identify different stakeholders
- What are their incentives/ what do they care about
- How to frame your analysis and results in context of what matters to them
STUDENT PRE-WORK
Before this lesson, you should already be able to:
- Compute basic statistics
- Fit linear models to data
- Compute regression metrics and model fits
INSTRUCTOR PREP
Before this lesson, instructors will need to:
- Brainstorm 1-2 good examples of audiences you have presented to in the past and be ready to discuss your experience in terms of stakeholder identification and analysis.
- Read in / Review any dataset(s) & starter/solution code
- Generate a brief slide deck
- Prepare any specific materials
- Provide students with additional resources
LESSON GUIDE
| TIMING | TYPE | TOPIC |
|---|---|---|
| 5 min | Opening | Discussion |
| 15 min | Introduction | Identifying Stakeholders |
| 10 min | Demo | Sample Stakeholder Identification |
| 15 min | Guided Practice | Stakeholder Incentives |
| 10 min | Demo | Framing Your Analysis and Results |
| 10 min | Guided Practice | Good and Bad Visualizations |
| 15 min | Independent Practice | Project 3 Stakeholders |
| 5 min | Conclusion | Review / Recap |
Opening (5 mins)
- Review prior labs/homework, upcoming projects, or exit tickets, when applicable
- Review lesson objectives
- Discuss real world relevance of these topics
- Relate topics to the Data Science Workflow - i.e. are these concepts typically used to acquire, parse, clean, mine, refine, model, present, or deploy?
Check: Review students' progress on Project 3. Which students chose each scenario? How are they thinking about framing their results at this point?
This lesson will be much less technical than most of the other lessons this week.
Introduction: Identifying Stakeholders (15 mins)
One of the most important aspects of data science is communication, particularly when stakeholders are involved. Results must be communicated effectively to clients, customers, and managers, taking their needs and technical backgrounds into account.
When communicating with fellow data scientists and other technical folks we naturally use a lot of technical jargon. This is usually not appropriate for other audiences.
Primary stakeholders include anyone that your analysis will be presented to and whom will act on your conclusions and recommendations as well as secondary stakeholders, anyone significantly affected by your efforts. Stakeholders will sometimes have technical backgrounds but typically this is not the case.
Identifying stakeholders is the first step in determining the right tone and level of technical depth to use when communicating analysis and results. Often the primary stakeholder is the party that hired or contracted you, and there may be many stakeholders in any given scenario.
A good first approximation at identifying primary stakeholders is by their influence. Executives and other leaders, large shareholders, government officials, and people in positions of influence such as college presidents, department heads, project managers, and community leaders are all good examples. Secondary stakeholders tend to be those that are influenced -- rank and file employees, community members, students and others that are more often affected by the decisions of those with influence.
Exercise
Identify the stakeholders
- An ad-tech company has hired you to evaluate the accuracy of their click-rate metrics and analyze their traffic for fraudulent clicks. Specifically you were hired by an engineering manager that believes that the metrics are inaccurate.
- You are hired by an engineering consulting firm to analyze data from a recent industrial accident. The consulting firm was hired by a factory owner to determine the root cause of an explosion, and the owner is considering litigation over a faulty component.
- You are hired by a local librarian to model the likelihood of borrowed books being returned based on various demographics and historical data.
Some Stakeholders:
- The company and the manager that hired you; the project manager of the ad-network's analytics, the companies that purchase the ads, other employees
- The engineering consulting firm, the factory owner, the component manufacturer
- The local library and its patrons (eventually policy changes will have to be communicated and justified to the borrowers), possibly the city manager or other local government official that directs library funding
Demo: Sample Stakeholder Identification (5-10 mins)
Present a demo stakeholder identification (or more than one) relevant to the background of your students. If possible, identify their motivations as a hook for the next section.
Guided Practice: Stakeholder Incentives (20-30 mins)
Once you've identified stakeholders, the next step in stakeholder analysis is to understand their interests, motivations, and incentives.
This is a good group-discussion activity.
Let's discuss the stakeholders and their incentives in the above scenarios. In each case you've been hired to discover something from a set of data. Often a stakeholder will have an expectation of the results. It's crucial to
- Manage expectations throughout
- Present data-based analysis with evidence-backed explanations
- Communicate results with an authoritative and non-confrontational tone
For each of the scenarios above, try to identify the stakeholders' motivations and expectations. If you are not sure, try to think of some plausible motivations.
You can have students theorize potential motivations or present some ideas first.
To do these effectively you need to understand the stakeholder's motivations and incentives. There are many possibilities. The stakeholder may simply:
- Need help analyzing data
- Have a point to prove or disprove
- Have a desire for technical justification for a hypothesis
These high-level incentives often have a more direct incentive:
- Increasing profit -- cutting costs, optimizing pricing, identifying production bottlenecks
- Social change -- the stakeholders may be trying to justify an inequality or quantify its effects
- Many others -- can you think of any?
Demo: Framing Your Analysis and Results (5-10 mins)
Back to slides or lecturing
As a data scientist you will have significant influence over the decisions of stakeholders and institutions. It is crucial that you remain objective and unbiased with your analyses and conclusions. Often we are tempted to please or agree with those that we work with or are contracted to. It is a disservice to all the other stakeholders to slant your conclusions to suit the desires of one stakeholder over another.
Even more crucially, you must take care to communicate your results in direct and understandable language, including enough technical information to justify your assertions. It's fine to explain your results in the context of your stakeholders' interests so long as you can maintain objectivity.
Good visualizations are critical to the communication of results. Avoid excessive details. Plots are better than tables. That said, avoid unnecessary visualizations -- each plot should make a concrete point clearly. Keep your plots simple -- two simple plots are better than one complex plot. Just as we must balance bias and variance in our models, we must also balance complexity and information in our figures.
It's tempting to make complex figures -- when they land it can be quite dramatic:
But for every good complex visualization there are at least ten bad attempts.
Guided Practice: Good and Bad Visualizations (20 mins)
Take a look at examples of good and bad visualizations with the class or in small groups
Examples | There are entire websites based on showcasing bad visualizations
Have students practice different communication techniques to different stakeholders. For example, in the library scenario above the goal is to, sketch out a few visualizations of results that you might communicate to the librarian and government employees, and another to the library patrons (perhaps a flyer informing them of new policies and why).
Independent Practice: Project 3 Stakeholders (15 minutes)
If you are short on time, students will pick this up in the next lab. Go with the class dynamic -- if the group discussions have been going well, continue with them.
Spend some time practicing what we've learned today for your Project 3 scenario. Your tasks are to:
- Identify stakeholders, primary and secondary
- Identify stakeholder interests, motivations, and incentives
- Sketch (on paper) some potential visualizations of your anticipated findings to various stakeholders
Conclusion (5 mins)
- Recap topic(s) covered in short takeaway bullet points, such as: (a) importance of identifying stakeholders (b) importance of proper framing of analysis and results (c) what makes a good visualization.
- Describe homework or any upcoming tasks