1. Fundamentals of storytelling
Welcome! My name is Hadrien and I will be your guide for this course.
2. Challenger
Before we jump in, let me tell you a story. In 1986, the Challenger shuttle's fuel tank ignited and the shuttle crashed, killing the seven astronauts on board.
3. Good warning, bad delivery
The culprits were two redundant seals in a joint that broke. NASA employees warned that flying was unsafe in the following slide (in the circled part), but the message didn't go through.
This is an extreme example, but it shows how crucial results can fail to have an impact (or have a bad one) because of bad communication.
4. About the course
Throughout the course,
we will learn how to communicate results to different stakeholders using storytelling.
We will also cover how to structure a report and
build an oral presentation.
5. Chapter 1
We'll translate technical results for non-technical stakeholders,
and we'll see how to structure stories to impact the decision making process.
And let's be clear right away: data storytelling is not about spinning results,
it's about making them stick.
For this to work, they should be simple (pruning the message to its core), concrete (can be described or detected by human senses), and credible (can be put to test).
6. Data storytelling road
In any communication strategy, there are several pieces we have to put together to create an effective story.
7. Why are stories needed?
But why are stories needed in the first place?
People working with data spend a lot of time collecting and processing data, and modeling and evaluating results.
8. Why are stories needed?
But what about communicating insights?
Even the best results could fail to have an impact if the insights are not presented appropriately.
One of the main challenges is to convince change-adverse stakeholders that some things should be done differently,
especially if these stakeholders are non-technical, and do not feel comfortable with math or statistics.
9. What is data storytelling?
That's why telling stories with data is paramount. So what's data storytelling?
It's a powerful mechanism for sharing insights supported by a compelling narrative and efficient visualizations. Why is it so powerful?
Anecdotes drive people's imagination
and stories are more memorable than metrics,
and they add value by giving context to the data.
Thanks to that, the audience can understand the insights. So their attention is captured.
Those insights facilitate the decision-making process and in turn,
drive actions and changes.
10. Data storytelling
Let's start with building an effective data story. Two concepts help us build its core.
The first concept, the three-minute story, makes us think: if we only had three minutes to tell a story, what would we focus on?
The second concept, the big idea, pushes us to state our story's unique point of view in one sentence.
These concepts help us articulate our story clearly and concisely.
11. Data storytelling
There are three central elements for any data story:
the data, the narrative and the visuals.
These elements make the story
insightful, as it derives clear learnings from our analysis,
explanatory, because we help our audience understand the insights,
and concise, giving only concrete and specific facts.
12. Data
To build an effective story,
we should include only findings from our models or analysis
that apply to the situation.
Our story should be based on accurate and reliable data, leaving out all untrustworthy results.
More importantly, our story should always include actionable insights: data that drives action.
13. Narrative
A good data story
includes a compelling and easy to understand narrative
that includes only the key points needed
to drive change.
14. Narrative
Even though it is tempting to include many disconnected facts, a compelling narrative revolves around one central insight,
that takes into account the background and our intended audience so it can clarify the facts for them. Who are they? What do they need to know?
Lastly, every data story follows a linear sequence. So every data point builds on each other until the conclusion is reached.
15. Visuals
However, we should be careful not to include misleading graphs.
16. Fictional company
We are going to look at how all the concepts are implemented at a fictional data science consulting company named communicatb.
17. Let's practice!
Let's put these concepts into practice!