What is data storytelling?

Please go to my personal website if you want to know more about me or get in touch.

Datainsightout helps crunching your numbers... discover how

Datainsightout is for you... for a fair price!

Discover our product

It's all about data, right? Getting the data right, cleaning the dataset, making sure data perimeter is valid, removing the outliers... that must sound pretty familiar to anyone who had to manipulate data in it's day to day job. However, this is just part one; and it's actually the easiest part of the data work. Storytelling is what distinguishes between good analysts and great analysts. 

Before knowing everything on data storytelling, you need data and good dashboards. You should read our post on business intelligence.

It's all about data, right?

Nope. Actually, it's all about data storytelling

Numbers are just numbers. Randomly set, they are not helping. Put one by one, it's just a collection and intepretation can goes in multiple directions. Data are secondary and are only there to back-up a message. The messaging and getting to the point should be the priority number 1 in any presentation. 

It's sometimes hard to detach yourself from the data. However, the audience will want to hear a story. Not re-do hard calculations live with you. Getting attention is precious but getting cognitive reasoning from an audience is almost impossible. People will listen more what you say than on what's written on the deck. 

You could actually separate the deck from the story. The deck is for exec to read later. The story is what you will tell during the presentation. A story could almost be written on a Word document in parallel. Data don't drive actions but stories speak to the heart. It should be like in book. You should have a plot and a logic.

Focus on undertstanding general business context

Numbers need to be digested in the context of the company current strategy if you want to be understood and your audience's approval. Numbers make a difference to business people when they fit the company's overall strategy and projects. To make an impact, business elements such as latest deals signed, new projects being kicked off or product feature launch, need to be part of the overall company story.

Datainsightout helps small and medium companies crunching their data. What about yours?

Discover our product

It implies to get a deep understanding about the context: latest discussions at an executive level regarding the topic you're in charge of presenting the numbers. A strong business acumen is required to engage your audience, especially when it comes to numbers interpretation followed by business recommendations.

One idea to actually create a good data storytelling is to go through the 3 following points.

First, get alignement around the common question

First step is to gather the audience and get agreement around what you want to achieve with the analysis. Explaining the purpose, the frame of the analysis, what's in and what's out. This helps building a common ground for the analysis and clarify the context of the presentation you are going to do.

The objective needs to be clear. Audience must understand what you want to solve.

Getting the slide title straightforward

A good PowerPoint is a deck where the reader could skim accross the deck and only read the big title at the top. You should almost be able to append each slide title and form a paragraph you could read. This implies to focus on a very straightforward story. As such, all which is great but not relevant to the story you want to tell needs to be pushed to the back-up section. Avoid being not to the point and complex sentences.

You can also use subtitle in powerpoint. The title could be the main data-driven fact you observed. The subtitle could be more like a comment. Here is an example:

  • Slide title: "We noticed a +12% increase in customer sign-ups in Feb on email channel. This is +5pts more than all other channels"
  • Slide subtitle: "Performance was driven by our February marketing campaign"

Relate numbers to business actions

You should be able to list on each deck a concrete business action. Impactful analysis are analysis where you point the direction to solutions. During the deck preparation phase, you could investigate each point with business owners, to make sure your final read-out will be actionable and will actually makes sense.

Sometimes, pre-sharing to exec before the read-out is a good idea. Executives have time to go through it. They can raise relevant questions during the read-out. No surprise effects for them. They can also help you correct your titles and give you more business context.

Getting feedback can change the storyline of your presentation. Remember: executives already know 90% of what you will present. They are not more clever than you: they just have access to more information than you. The remaining 10% will be new to them. It somewhat needs to fit what they have in mind.

Reduce the amount of cognitive aptitude required to understand the deck.

Focus on the message and reduce cognitive amount required.

Data visualization (all the graphics and visualization used) needs to be as meaningful as possible and the amount of cognitivie ability to understand what's going on needs on the contrary to be reduced at its minimum. Complex slides, hard to read graphics... sometimes, less is more. You need to put yourself into your audience's shoes and get actions implemented out of your big data crunching piece of work.

As a data analyst, you might know in depth all about the chart you are going to present. However, people in the room might come from another meeting, have another agenda and other priorities. They also might don't know much about the topic you are going to present. Data storytelling is about reducing the complexity of the data and metrics presented and to focus more on simple but direct messages. Most of the time, one graphic should be featured at maximum per slide. Readers can't digest too many information at the same time. 

About that precisely, you'll find below a great conference by Isaas Reyes. Not too short, not too long, but it gives definitely a good understanding of what's required to build impactful data storytelling.

Good data but bad storytelling

That being said, you probably have by now plenty of experiences of bad data analysis read-out popping-up to your mind: analysts presenting in front of an audience a great piece of work, hard-time data crunching... but nobody in the room get the "so, what?". Confusing slides, too long presentation. Hence why Jeff Bezos decided to ban PowerPoint from Amazon's meetings. Doing a presentation is far from being the easiet thing.

Sometines, presentations are not required and simple email memo can trigger more actions and drives engagement from people who are going to read you. Thinking about the way you are about to deliver your message can have an importance.

Remember: it's all about the story. Numbers are just back-ups. 

This post was rated: 5 / 5 by 2 readers
You rated: Not rated yet... but please be kind!

Datainsightout helps crunching your numbers... discover how

Datainsightout is for you

Discover our product
Get the latest posts directly in your mailbox for free

Enter your email address: