CW500: Why data analysts need a soft edge to make the most of hard data

The ability to present data in ways that make sense to business leaders is at the heart of analytics

Businesses may have convincing data and compelling statistics, but when it comes to convincing executives to act, numbers alone are not enough.

Data analytics is transforming business decision-making across every industry, from music, to retail, to the media.

But it's not the numbers themselves that are driving the data revolution, it is the ability to present data in ways that make sense to business leaders that is at the heart of analytics.

Alexandra Mitchell, senior manager for lead generation analytics at Getty Images, learned that lesson early on in her career. Working for a startup company, she turned to Google Analytics to look for ways to improve the performance of the company’s website. 

The figures showed that people were spending a lot of time clicking on one particular area of the site and then abruptly left. Mitchell discussed the results with a user experience (UX) manager and they both presented their findings to the CEO of the company, she told CIOs at Computer Weekly’s CW500 club for IT leaders.

The CEO’s reaction surprised her.

“Thank you very much. Very interesting,” he said. “Your wire frames, I am not sure about them. And we know that Google Analytics data is not completely correct. I think this page is fine. In fact, I designed it myself.”

So the analytics team tried a different tack, using a video camera to film a panel of people as they used the website. 

Mitchell felt she had made her case when one frustrated women turned around and spoke directly to the camera, saying: “I know this is not my problem. I have two degrees and I can’t work out how to use your website. It's your problem.”

Her CEO agreed. The lesson? People understand people better than they understand data.

Putting data in the picture

Getty Images, one of the world’s leading creators and distributors of still imagery, video, music and multimedia products, uses data analytics to drive product and strategy decisions.

“We can demonstrate that a product can add incremental value to a customer. That can change the attitude of the sales team. Suddenly the product becomes the fastest-growing product,” said Mitchell.

But it’s the way the numbers are presented that makes the difference. Graphs and bar charts are a poor way of communicating, she said.

Alexandra Mitchell, Getty Images

It’s the way statistics are presented that makes the difference. Graphs and bar charts are a poor way of communicating

Alexandra Mitchell, Getty Images

If you want to show a trend, tell the story of an individual customer and their experiences, rather than talking about data.

It’s a principle that Davide Cervellin, an analytics specialist at eBay, understands well. Cervellin uses analytics to persuade retailers of the value of selling their products on the web-based auction site.

Merchants are often cautious when it comes to selling their products on eBay, he said. “Sometimes they just want to explore the channel. Sometimes they want to explore an opportunity because it’s a simple way to get into a market. Sometimes they don’t want to be on the market but they want to test demand.”

Cervellin uses data analytics to work with the merchants to demonstrate the value they gain from eBay. He projects sales trends for eBay’s merchants six months into the future, then he sits down with them and goes through the figures to check that they match the actual sales.

“That builds confidence. And it's fine if it does not work out as expected, as long as you understand why it has not worked out,” he said.

Because eBay is a public company, it needs to keep a tight grip on its financial performance, so Cervellin’s projections also feed directly into eBay’s own financial projections.

“The numbers help drive the chain of decisions in eBay. For example, what discount will the retailer get, if any? Does the merchant need more visibility?” he said.

Tyred and emotional

Cervellin has seen some powerful examples of analytics affecting business. In his previous role, at tyre company Pirelli, he helped set up a loyalty programme for people who bought tyres – with spectacular results.

With numbers you need to be able to make the point in a soft way, which is a skillset that is hard to find

Davide Cervellin, eBay

Cervellin and his team realised that drivers have no inherent love of tyres, so they devised a loyalty programme centred around, not which tyres they bought, but the type of car they owned.

“People are much more passionate about their cars. They want to make a point – 'I own a BMW',” he said. “But knowing which car they had meant we knew which tyres they needed for winter and summer.”

The team presented the idea to the marketing manager. He though it was a waste of money, Cervellin told the IT leaders at Computer Weekly’s 500 club event.

But the analytics told a different story. The team pulled some favours with tyre dealers to build up a database of the cars their customers were driving.

When the team sent out an email to 10,000 people which identified the car they actually owned in the subject line, the open rate was 10 times that of Pirelli’s standard emails.

“The head of marketing thought it was amazing,” said Cervellin. “Then we showed him the redemption figures [which were just as impressive].”

That experiment, put together with no budget, turned into a successful national campaign.

Numbers don’t do the talking, people do

But when it comes to getting your message across, Cervellin agrees with Getty Images’ Mitchell – that soft skills are more important than hard numbers.

Starting with a blank sheet of paper and coming up with something breathtaking is an art. Being presented with a spreadsheet of data isn’t usually going to bring the best out of someone's creativity

Chris Carey, consultant

For many data analysts, the mindset needed to convince and cajole a skeptical audience of the veracity of their work may not come naturally. They may be more comfortable discussing the nuances of database commands than making a public presentation.

“With numbers you need to be able to make the point in a soft way, which is a skillset that is hard to find,” said Cervellin.

But soft skills can be developed. Cervellin encourages his data analysts to try different roles within the company, to help them develop their communication skills and push them out of their comfort zones.

“You really need to work on the soft skills first. You don’t want to expose them too early if they are not ready to behave with the merchant or customers,” he said.

Cervellin has encouraged analysts to take courses in English, public speaking, presentation and self-awareness. These courses can teach specialists to be more flexible in their work life, and that can help people progress their careers.

“If you are too experienced in something, the opportunities for you are closed, and in the future [your company] may not require you. You need to have the agility that comes from relating to people,” he said.

Not everyone has what it takes, but a little coaching can go a long way. Mitchell, for example, said she has seen business analysts who operate in an introverted style transform to behaving in a highly extroverted way in four to six months.

Beyond the numbers

While companies such as Getty Images and eBay have been analysing data for some time, the music industry is a late-comer to analytics.

Until digital took off, record labels had little direct contact with their customers, and little understanding of what motivated them, said Chris Carey, independent consultant and former global insight director of Universal Music.

Record companies sent CDs to distributors who delivered them to retailers, and the retailers had the customer relationship. And if your products were in the right place at the right time, a customer might buy one, he said.

Today, with the advent digital downloads, the more savvy music companies are getting to know the behaviour of their customers in ways that were unimaginable even a few years ago.

For example, it is possible to compare the people who bought a band’s first album with those who bought the second album to understand if you are engaging existing fans effectively. 

If there is a big difference, record companies can use that information to direct their marketing strategies and to target fans who may have missed out on one of the recordings.

Add to that data from streaming music services, which shows the popularity of music tracks among listeners, and record companies are able to get closer to their customers than ever before.

But there are some things that data cannot tell you, said Carey. People always ask whether data analytics can help artists write a hit song. The answer is a firm, "No".

“Writing a hit song takes incredible skill. Starting with a blank sheet of paper and coming up with something breathtaking is an art. Being presented with a spreadsheet of data isn’t usually going to bring the best out of someone's creativity," said Carey.

And sometimes it makes business sense to lead the consumer and the market, rather than simply following the statistics. Universal Music, for example, nurture bands for years before the public is ready for them.

“Sometimes you talk to consumers, and only the tastemakers - about 3% to 4% of people- 'get it', but no one else. That's not a failure, it's still a great record, we just have to wait for the market to be ready for it,” said Carey.

"As Henry Ford put it, 'If I'd asked people what they wanted, they would have asked for faster horses'.  There is a role for labels in funding creativity and leading consumer trends."

And while analytics can tell you how many times someone has listened to a track on a music streaming  service, it cannot tell you how they got there, why they started, why they stopped or what they will do next - but it is incredibly powerful for understanding who artists' fans are and engaging them like never before. 

Decisions directed by data 

Ultimately, it's not how much data you have, but the business decisions you make based on the data that are important, the meeting of CIOs heard.

Companies may have all the data, but managers still make irrational decisions.

As Cervellin told the meeting: “You can have two people betting on the results of tonight’s game: one who has studied everything, and one who has studied nothing. And the one who has studied nothing wins.”

Read more on Business intelligence and analytics