CW500: Forget the fancy graphics: how to make big data work for ordinary people  

If there is one thing big data specialist Chris Osborne has learned after five years of working with big data, it is that the public don’t like graphs

If there is one thing big data specialist Chris Osborne has learned after five years of working with big data, it is that the public don’t like graphs.

A lot of people have problems just setting the thermostat at home so expecting them to make sense of pie chart is a step too far, he told CIOs at Computer Weekly’s 500 Club for IT leaders.

Osborne is the big data product manager at, a company that is using technology to help people monitor the energy consumption in their homes.

He has learned the hard way just how difficult it is to turn big data into products and services that the public can understand.

“An important thing to remember is that the average reading age of a UK online consumer is 12 years old,” he says. That means that whenever you design anything on the internet, it has to be simple enough for a 12 year old to understand.

Pie charts may be common currency among data professionals but user testing has shown that time and again ordinary people find them difficult to interpret.

Far better, he says to present information in the form of a simple ordered list, than to make people jump through the mental hoops trying to work out whether the blue segment of a chart is bigger than the green.

Osborne’s first foray into big data offered some valuable lessons in the difficulties of working with the public.

At first sight, an application to help people find properties in London within commuting distance of their work seemed like a good idea.

The application allowed users to specify how much time they wanted to spend commuting, how much they wanted to spend and what sort of property they were looking for.

By combining the users preferences with data on London’s transport network, the app was able recommend suitable properties.

“In some ways, it was very successful," says Osborne. "I was very proud of the style. It was very advanced for its time and it had some quite complex data processing. 

“But it completely failed.”

With hindsight, Osborne realised that the application was unsuccessful because he had not taken the time to understand how people actually buy and talk about property.

“Most people who had bought a house, or were looking for a house, would nearly always say, I want to live here. I am looking for a house there,” he says.

Only then do they decide what type of property they want, how many bedrooms, whether they are looking for a garden or whether the garden is south-facing.

“I made the fundamental mistake in not understanding the context and environment of how people make a decision,” he says.

The Four Rules of Insight

  • Personalise – The application should tell the user something relevant  to  them.
  • Accessible – Data should be simple and accessible. Hide data that is not relevant to the user. Don’t represent data in the form of graphs or pie charts.
  • Actionable – Don’t give consumers information unless they can do something with it.
  • Instinctive – The application should be based on a understanding of people’s behaviour, and the context in which they make choices and decisions.

Source: Chris Osborne

This, and other experiences have helped Osborne to put together what he calls his Four Rules of Insight (see box, right); four rules of thumb that help to make big data accessible to the public.

Today, Osborne is putting those lessons into practice at, a company of 100 people that is working on ways to provide the public with useful data by connecting intelligent devices in their home to the internet.

One of the firms goals is to take data from the 45 million smart electricity meters that will be rolled out across the UK by 2019 and convert that into information that will mean something to consumers.

Osborne’s starting point was to find out how people talk about their electricity consumption.

They don’t talk about how many kilowatt hours their fridge uses a week or how much they spend per hour on electricity. But they do talk about their actions, such as charging a mobile phone or cooking a chicken. So’s challenge is to find ways to present electricity consumption to consumers in a way they can understand.

“We have sensors in our homes and in our researchers’ homes. We can monitor what energy people use and break it down and show users exactly where all the energy is going in their home, which is a great thing for consumers,” he says.

If there is one thing Osborne is clear about, it is that whatever the answer is, it is definitely does not include a graph or a pie chart.

And certainly not the main stay of big data visualisation used in businesses – the dashboard.

“Consumers are not interested in them and we should not expect them to be. They are interested in convenience and they are interested in having information presented to them in a way they don’t have to think about it,” he says.

The challenge is to take the pain away, and only share information that is relevant. Hiring someone from a non-IT background to train as a coder can help.

They help to bring a consumer ‘s perspective,  says Osborne, who moved into big data after studying Geography at University.

“If people need to make a decision, guide them in how they could make it, but at the end  goal is a lot of automation and removing of complexity rather than generating lots of data,” he says.

Read more on Technology startups