Even the most sophisticated analysis is no match for an apathetic audience. If the human at the other end of the table doesn't care about your results, the hard work you put into it won't make an impact.
As analysts and data scientists, we have to make our work compelling and persuasive to change the way our organizations think. It can be easy to assume that our well-constructed graphs and tables are plenty convincing on their own. But this ignores a fundamental fact of analytics: it's still a human making the decision behind those numbers, and that person is rarely as quantitatively inclined as we are. Persuading this audience takes a skill that isn't discussed enough in the data science community: empathy.
So how do you deliver your analysis with empathy? You have to meet people where they are. This is a conversation we have with our customers all the time, and something we strive to make a pillar of our own analytical work here at Mode. Over years of practice, a framework for the empathetic presentation of analysis has emerged:
1. Start with their problem.
Analysis has to start by understanding the problems of the people involved. It's almost always going to be related to the core function of their department. They may need to increase sales, ship products faster, or expand the marketing funnel. Your analysis will only have an impact if it clearly helps people meet their specific goal.
2. Eliminate caveats and nuances.
Yes, you want to present truth, and the truth can be messy. But your audience won't always understand everything you tell them, and the parts they do understand won’t always be the headlines. The muddier your point, the less likely someone will walk away with the most important information. The more perspectives and angles you pack into an individual dashboard, the less likely your audience will find the one they need. Clarify the point to make sure it sticks.
3. Tell the story.
Dashboards packed with data can make anyone's eyes glaze over. Without a clear narrative, a dashboard will feel like an overwhelming wall of numbers with no clear start or end. When presenting results, provide context for your analysis in the form of a narrative. This will help translate abstract numbers into tangible ideas and make those results memorable.
4. Address the assumptions.
Every person who consumes your analysis will bring their own assumptions and beliefs. These priors can lead them to a different conclusion than the one you intended. Identify what your audience's assumptions are and make them part of the analysis itself. Without acknowledging where your audience is coming from, you won't be able to show them where to go.
5. Make it quick.
The more effort it takes to get to an answer - including digging through a complex dashboard - the less likely someone is to absorb that information. Don't make someone work any harder than they have to to get the point.
How does this compare to your own experience? If you're looking for help with your analytics initiative, building out your data team, or just looking for someone to commiserate with who understands your problems, hit us up: firstname.lastname@example.org