The Importance of Data Visualisation
Data visualisation is, arguably, the most important stage of the data analytics process. A good visualisation will present your data in such a way that makes it far easier to uncover the patterns that lead to insights. It will also enable you to demonstrate to decision makers how you arrived at these insights and help persuade them to act.
Too many companies still see data collection as an end in and of itself. Without insight and action, all of your efforts collecting, cleaning, and storing the data have ultimately been for nothing. As Koen de Brauw, CEO of Cartha, says: “Even if you have a massive mountain of data, it’s still useless if you can’t actually use it.”
Cartha has developed automated software to find actionable insights from internal and external datasets and deliver them directly to the end user by visualising them on interactive maps. Their goal is to help organisations get more value from their data and become more data-driven, without all employees needing a data science degree.
We sat down with Koen to find out more about the importance of data visualisation, and how and why they use maps to tell stories through data.
How Much Data Is Too Much Data?
The amount of data collected today would have been unimaginable just a decade ago. There are 2.5 quintillion bytes of data created each day, and this is only increasing as connected devices bring more of our world online.
For organisations both private and public, it is more important than ever to exploit the insights that this data holds. It helps them to understand the market for their product, potential risks that could occur, new efficiencies that could be created in their processes… the list of benefits is endless. Many organisations are, however, still ill-equipped to make proper use of data and are subsequently missing out on all of these.
“The last 15 years or so have been about gathering more and more data,” says Koen. “We’re at the point where everyone has all the data they need. The question is, how can we actually use the data?
“What companies need to understand is how they can get actionable insights from their data that can be used on a continuous and consistent basis. They need to deliver it to end-users, so they can use it in their day-to-day operations — not just to business intelligence teams somewhere in the building. Ultimately, they need to understand how they can profit from it.”
The tools and talent to make sense of all this data are out there, as you would expect after more than a decade of ‘big data’. However, both are still prohibitive to most organisations, with the exception of those who have deep pockets. Koen explains that “Software tools are usually very expensive, large, and complicated. They also need to be run by a dedicated business intelligence team, as your average user still lacks the skill-set and training to use data. Besides, the average user doesn’t really want to use Excel or look at graphs the whole day.”
In effect, by parking data with a business intelligence team, a company is stove-piping its data. This is inefficient. It prevents data from reaching the areas of the company where it’s needed in a timely fashion. It disrupts the flow of data within a company, which could cause insights to be missed or potentially important data to not be passed along. It prevents a company taking a holistic approach to data that could see someone spot, say, a new supply chain efficiency that could save a company millions.
If it’s not enough that data is being siloed to one team, the range of data streams that come in through different collection tools also mean data being siloed. “Data is often spread across many different places,” Koen says, “whether this be your CRM system, Google Analytics, your marketing tools, or some public database. The spread of data over different sources creates a barrier to most users. Even once you know where it all is, it is hard to pull together and compare, and therefore quite hard to make actionable. This leads to potential insights being missed. Even something as simple as combining data from your webshop and google analytics can result in valuable actionable information.”
Cartha’s goal is to solve these problems. First, their machine learning software automatically uncovers actionable insights in the datasets by identifying patterns. Second, they present it in such a way that ensures it can be used effectively by everyone across the organisation.
What Makes a Good Data Visualisation?
Data visualisation tools have advanced tremendously over the last decade. There is now highly sophisticated software capable of processing and presenting incredibly complex datasets in seconds.
Equally important as the technology, however, has been the realisation that it is how we tell a story with data that really matters. The human brain has remarkable pattern recognition capabilities, but the majority of people struggle to comprehend anything that they cannot visualise in some way. As Koen notes, “The key problem is that people have a hard time fathoming endless lists of numbers and graphs.”
To present data to people, it is no longer enough to chuck out a few bars charts and hope for the best. Today, we are presenting data through increasingly inventive ways — highly-engaging, often interactive designs that immerse your audience in the story you’re telling. As one of the great data storytellers Hans Rosling said, “Having the data is not enough, I have to show it in ways people both enjoy and understand.”
This is no easy task though. When you have large amounts of data coming from multiple sources, representing these streams in a logical, concise way is always going to be a challenge. An effective data visualisation will draw an impactful response from users, taking them on a journey that enables them to understand the insights being presented or discover insights for themselves. An effective visualisation helps users to get to the right data easily and answer a specific question visually.
A good visualisation does not overload the audience with extraneous information. You need to think about what data is actually needed. You need to think about the person using the data, not just the data itself. “If you want to make your data actionable on a daily basis, it shouldn’t be data first,” says Koen. “Instead, it should be human first. How can the data help humans do their jobs better? I think that one of the most important elements of this is how we communicate data to humans. This is a very important aspect of our app.”
Using Maps to Visualise Data
Koen’s idea to present data through maps came from his experience as a military historian. He was often asked to make sense of and explain certain recent global armed conflicts to friends and acquaintances.
“I completed my military history studies as the war in Syria was taking place,” says Koen. “Being involved in the field, people were coming up to me asking, ‘What the hell is going on with the war in Syria?’ because no one really understood.
“I would tell them to read certain books, watch videos etc, but people told me that no one had the time to read anything because we’re always being bombarded with information. As an alternative, I started making geographical maps of Syria for people. These explained which group was fighting which group, which group was allied to that group, and so forth. Taken together, this built a picture of what was going on.
“The map helped people to understand an incredibly complex war. These simple maps were able to convey the complexity of the war much easier than text or video could. That was an important realisation: visualising issues on maps helps people understand complex issues much faster. How could we use that?”
If such maps can make something as complex as the war in Syria easier to understand, it is only logical that it could be applied in other scenarios. Koen shares the example of an organisation that rents out rooms in Southern California.
“Southern California has a lot of wildfires which dramatically impact the local economy,” he explains. “Say, you’re an organisation in the hospitality industry. You get a message saying that there’s a wildfire near one of your hotel rooms. To determine what actions to take, you need to gather external data such as information from fire departments, current and future weather data, Twitter data… whatever data you can imagine. You also need to gather internal data like where the hotel roms are, which ones are booked, which ones will be booked in a week, how close they are to the fire etc.
“Bringing together all those data streams into a single spot and making an actionable decision, like closing certain hotel rooms because they’re too close to the fire, takes too much time. What a company will do instead is cancel all the rooms in that area. This is extremely costly. There are refunds to pay and lost revenue.
“By combining all of the various data streams and using our back-end engine, we can work out where the fire is, where it will be going and determine what rooms should be canceled and when. The map then makes all this information directly available to the end-user. The Smart Map helps the customer make better decisions in less time. Resulting in less unnecessary cancelations, saving them a lot of money.”
The key to the effectiveness of the map above is that it demonstrates the consequences of the insights discovered. “By generating actionable insights and visualising them on a map, people have a much better grasp on what is going on and what to do next,” explains Koen.
It’s the immediacy of the map that is really interesting here. If you tell a smoker they’re going to die early, they’ll rarely listen because it’s an abstract that the brain will not really register. Tell a smoker that they’ll die tomorrow and show them a picture of lung deterioration and they’ll sit up and take notice. A map is convincing because it shows you a story unfolding from beginning to end — the entire journey and the consequences that will arise.
A map is also powerful because it is so familiar. Humans have been using maps for millennia, and for good reason. Familiarity is the key to an effective visualisation as it allows the human mind to process the information displayed that much faster. “Maps are powerful because humans instinctively understand them,” says Koen. “They therefore allow for a much better understanding of complex issues. Put something on a map and it becomes true. Much more so than data in a graph or table.”
Cartha’s goal is to make it easier for organisations to get actionable insights from their data and enable more data-driven work in all layers of their organisation. With such a powerful tool, they’re definitely one to keep an eye on. If you’re keen to find out more about how Cartha works and what they can do for you, say hi at Primalbase AMS or reach out to Koen at team@cartha.io.
What do you think makes an effective data visualisation? Let us know in the comments.
Be sure to keep up to date with all the latest developments from Cartha by checking out their website and blog.
Join Cartha and other tech startups in spacious offices across Europe and the US today. Get in touch and discover a workspace solution that really works for you.
