×

Language analysis: Using crimefighting tech to help brands

Communication has always been a core part of a company's strategy, as they strive to tempt consumers into making that crucial purchase. Now – thanks to big data - brands are using language analysis to get closer to their customers.

A growing number of businesses are working with specialist firms to conduct in-depth analysis of how their customers talk. One such firm is Relative Insight, which was founded by a team of researchers from the U.K.'s Lancaster University.

The company started by using its language analysis technology to help the police identify criminals – for example, to determine whether a chat room user was a 12-year old girl or a 55-year-old man imitating a 12-year-old girl.


Peter Dazeley | Getty Images

Now it has identified a new – and more lucrative - market: big business.

"It started dawning on us that we had something very powerful from a marketing perspective," Ben Hookway, Relative Insight's CEO, told CNBC.

Relative Insight analyses the way a brand's target market speaks online – for example on Facebook, Twitter, in comments and on forums – and advises the company how best to appeal to this audience.

Read MoreGet ready for video ads in your Instagram feed

For instance, analysis for a skincare brand in the U.S. revealed that younger American women (18-30) were much more likely to say they "wear makeup" than older women (35 plus), who say they "apply makeup".

"That might sound trivial, but if you're trying to appeal to a younger demographic, you might not clicking with their language style," Hookway explained.

Posh or accessible?

Kirstin Sinclair | Getty Images

Marius Luedicke, senior lecturer in marketing at Cass Business School in London, said it was crucial for companies to focus on the way companies and their brands communicate with consumers.

"Marketers have to define a personality for the brand, and this comes with a way of speaking," he told CNBC. "They have to decide how posh it is; how accessible. It gives personality."

He highlighted that some brands – such as Nike – are good at this, whereas others – such as Puma – are not so successful. "Nike has tapped into the women's fitness community, and uses specific phrases and rhetorical styles to appeal to it," he said.

Competitive advantage

But there is a stumbling block, according to Luedicke, even when complex language analysis has taken place: the pitfalls of imitating customers.

"That doesn't work," he said. "Imagine a classic marketer who is new to social media - they think they can fool their customers. They can't… The audience won't take it seriously."

Read MoreHow to avoid a corporate Twitter disaster

Relative Insight's clients include Microsoft and advertising agencies such as Saatchi & Saatchi and Havas.

Ben Silcox, head of data and digital for Havas, said the analysis provided a crucial layer of context for clients.

Read MoreInternet ad spend up32% as old media takes a hit

Benetton | Fabrica, 72andSunny

"People are complex, and our analysis needs to match that complexity," he told CNBC. "We do a better job… if we can differentiate from other agencies by providing faster, richer, more insightful analysis."

And it works, according to Relative Insight, which provided a case study where its analysis boosted transactions by 218 percent and revenue by 256 percent.

'In its infancy'

Havas' Silcox did admit that this type of analysis was "in its infancy", and even Relative Insight's Hookway said it felt like the company was "almost building a new discipline here".

"Conventional marketing techniques and big data techniques already exist. What we're focusing on is that last section of saying exactly the right thing, which really emotionally resonates with the consumer," Hookway added.

"You can do the best marketing work in the world, but if the brand then says the wrong thing, it's lost."

- By CNBC's Katrina Bishop