BI Meets Design: 4 Reasons Why You Should Visualize Your Data


Knowledge is power, we all agree on that. But what power are we talking about? The power of understanding the world that surrounds us; the power of taking conscious decisions; the power of being in charge of our destiny.

How can we achieve a better understanding of the information that we have within our company? But we’re not talking about a Google doc with a list of the people participating in a meeting; we’re talking about huge amounts of data that your company collects everyday.

If you think your company doesn’t have this kind of data, just drop by your Business Intelligence team and ask for it. It doesn’t matter if your company is selling orange juice or designing the latest model of electric car: you are collecting everything everyday.

So, now the question: what am I doing with all this data? Well, of course, we’re talking about something that goes beyond the aspect of business and finance: the keyword now is ‘data visualization’. If you’re asking yourself why you should work on it, sit back and let us explain: Stylight’s BI and Interaction Design experts know the top 4 reasons why you should visualize your data.

1. Simplify complex information in order to present understandable data

Some businesses might think that a Google sheet is enough to present your data. If needed, some fancy pie charts can be added and everything is clear, right? But tremendous amounts of data require new forms of data presentation. On the one hand we need a high level overview in order to suit top management needs, on the other hand we need the possibility to offer playful interaction with the depicted data to gather deeper insights. The possibility of directly interacting with our data encourages the curiosity of users and best takes advantage of human perception and cognition.
What are we talking about? Think about the 2016 caucus elections that are running right now in the US, think about a state like New York, divide it into every single county and fill it with every piece of data about the voter you can have. Of course we’re talking about which candidate got more votes, but also about the ethnic representation of the people voting, the time, the age and many other pieces of data.

New York Times caucus data visual

The New York Times designed a data visualization of the caucus 2016 results in New York.

A single doc would look messy for everyone involved. The New York Times provided a data visualisation that drilled down all the data in order to present understandable information to their readers.

Now think that all that data is your users, and that the candidates are your products, you would get the exact same amount of information. So, do you still think you don’t need data visualisation?

2. Data analysis: what you need to discover trends & patterns

Let’s get back to that Excel file with a lot of sheets and a lot of numbers. The real problem in that document is that the focus is on the single number and not on the big picture. Data visualisation is not just about putting a fancy chart over numbers, the most interesting part is about finding correlations between numbers.

If in your Excel file you have just the number of sold products per day, you can’t get much out of it, maybe a simple cartesian chart. But let’s pretend you have another sheet with the weather forecast per day in your most valuable market. With data visualisation you can easily correlate these two data sets and discover if there’s a trend. And once you discover it, you can plan your assortment in order to have more or less products in a certain period of the year. You can do the same if on the other sheet you have other information about your business that you can correlate.

The more data sets you have the more complex the data viz will be, but you’ll have a higher chance of discovering a path/pattern or a trend that was hidden in your numbers.

Polygraph Disney movies female dialogues

Polygraph took all the Disney movies and counted the words that male and female characters said to discover trends within the dialogues.

3. Use your viz to persuade stakeholders

Imagine you found interesting trends or patterns in your underlying data. You built some nice graphs and are ready to tackle the next steps. How should you deal with the graphs to persuade stakeholders within the company or clients and get the highest business value out of it? For hundreds of years humans have loved listening to storytellers, so don’t let your data visualization stand alone, instead use it to tell a compelling story.
With data visualization you are in a great place to start, but keep in mind a few basic didactic principles about how to tell the story:

a) From easy to complex

You already spent a lot of time analysing and visualizing the data and therefore you know the data very well, but your audience doesn’t. So make sure to start with the easy stuff and then go to the more complex parts. This helps people to follow along and doesn’t overwhelm them at the beginning.

b) From general to particular

Make sure to present the bigger picture first and keep the details for later. Imagine you’re showing the worldwide sales of your company’s products, you might start with the worldwide sales and then go to sales per country or region, right? This creates a tension curve and maintains the curiosity of the audience.

c) From abstract to concrete

Show the meaning of the data first and then dig deeper to the concrete implications. For example, if you share insights about how to increase customer lifetime value, make sure everyone who listens understands why in fact this is important for your company’s success. After you make your point and make sure everybody follows along, you can come up with your data driven solutions to improve.

And finally do not forget that the attention span of human beings is no longer than 20 minutes, so better keep it short and simple!

4. Working closely with data specialists and design enthusiasts is so much fun (and effective)!

There are two main challenges you are facing if you want to build an outstanding data visualization. The first one is getting the data in a proper, exploitable format, and the second is to make it visually appealing. In order to get the most out of it, ensure you team up with data scientists and design enthusiasts!
As a design enthusiast you know everything about visualization, but you most likely lack the skills of data manipulation and analysis. Hence teaming with a data specialists who knows everything about the numbers, but is probably missing the understanding of which visualization will work best for displaying the data.

An interdisciplinary, cross-functional approach will help you a lot to first understand your data correctly and then to treat and process it in the best possible way.
If you keep this in mind you will see how much fun data visualization can be and you will be surprised how beautiful your data can be!

Google Hot Search

Google Hot search is a cool experiment where you can see all the trending topics on Google.

Data visualization at Stylight

At Stylight we tested what can be achieved if a Designer and a Business Analyst play together with their data. As a topic we depicted the clicks on different fashion categories over the different seasons within a whole year. This is how we wanted to show that seasonality influences interest in various categories, like evening dresses or knitted sweaters.

Data visualization Stylight - reasons why you should visualize your data


We managed to build the view above with several steps, from a data exploration phase with Tableau on the Business Intelligence side, to visualizing it with NodeBox and Adobe Illustrator on the design side. One of the key success factors for us was to have valuable input from Data Scientists and design enthusiasts throughout the company.

If you want to have a deeper look at what we did, save the date, we will be speaking at the “Data Visualization in Munich Meetup” on June, 22nd.

|By Alberto Andreetto – Interaction Designer & Stefan Tippelt – Business Analyst|

Header: Estimates of migration flows –



Leave A Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.