Targeting retail sales
Smart use of location data could be a vital asset in the war against spam by allowing genuinely targeted advertising. If a firm has spotted that every Thursday I go to Paris from my office after work and that heavy showers are forecast in Montmartre, it might be useful to offer me a deal at the umbrella shop on the way to the station.
Getting smart about where we are could provide a much-needed lifeline to the struggling daily deals business, whose talisman Groupon has lost more than 80 per cent of its market value since going public a year ago, as merchants find they gain little repeat business in return for heavy discounts.
“There is a saturating effect of this type of marketing,” says Carl Tsukahara, chief marketing officer at Monitise, which advises on mobile banking, payments and commerce solutions. “If I have the proximity, preferencing and demographics information, then I can give you something that’s specific to you.”
With their bird’s eye view of individual’s spending patterns, the information kept by banks will be gold dust in helping retailers work out when and what to sell us.
Clearly the privacy hurdles are considerable. But if the result is more targeted advertising, consumers may warm to the idea.
Data-swapping between Google and his bank, for example, would have spared a shopper adverts for swimming trunks on every second webpage he visits, three weeks after he bought his new Speedos from John Lewis.
Combined with real-time position data, bank records could help alert us to buy stuff we want when we have the money to pay for it, say a nearby golfing sale for someone who has a regular habit of spending on golfing products the week after pay day.
Mining the data from our spending patterns should improve banks’ services to their customers too. Combining information about our cash flow, disposable income and daily balance will facilitate advice about financial planning, as well as issuing timely alerts, such as at the point of a purchase that will take us over our limit. Mobile data should soon help us find an ATM abroad that doesn’t charge a transaction fee or one at home that doesn’t have a long line on a Friday night after work.
In social cloud
The cloud – a fuzzy concept at best – isn’t to be taken literally when it comes to combining and analysing multiple sources of mobile data. Much of what is so far being achieved occurs at the intersection between mobile devices, the internet and firms’ own physical servers.
But the movement for combining this data to extract and apply useful information is where the twin concepts of big data and the cloud are transforming how businesses understand their customers’ preferences and behaviour.
“The key aspect here isn’t where the data is, but that we have access to it,” says Mike Saliter, director of global market development at QlikTech. “Twitter, for example, allows us to draw from data across geographies and boundaries, from all locations and from people on the go to help brands better analyse customer behaviour.”
Throw in position data and the social media data turns into something valuable. In September, Telefonica, which owns O2, started selling retailers “heat maps” of its 23 million mobile phone users’ movements throughout the day. Crunching the data from this with information from Facebook or Twitter should let retailers understand what people said about their shops before and after they left. Add in transaction data from individuals’ banks and you have a watertight account of what people bought and why.
By collating position data from its devices, Ericsson’s Swedish business can already text its customers if they are approaching areas of high congestion. Big data and the cloud promise much for the future.
Dr Linus Bengtsson, from the Department of Public Health Sciences at the Karolinska Institutet in Sweden, is looking to the “smart cities” initiative for innovation – a global movement towards better-informed, tech-savvy urban development. “Combining the number of free parking spaces with the number and location of road users could inform the programming of traffic lights to secure smoother traffic flows through cities, saving time and cutting emissions,” he explains.
But, for the time being, vendors will have to wait. Having spent years dissuading Londoners that Oyster heralded Big Brother surveillance techniques, TfL are adamant that they’ve no plans to start selling our personal travel patterns to third parties.
In the days following the Haiti earthquake, researchers in Stockholm and New York used historical and real-time data from two million SIM cards to model the flow of 600,000 displaced people from the Haitian capital Port-au-Prince. The information helped relief organisations on the ground provide shelters and, when the cholera outbreak soon followed, distribute supplies to where they were most needed.
“If you understand where people have their social security net, you have a good chance to understand where they will go during a crisis and we showed that anonymous cell phone data was a very good tool for mapping this,” says Dr Linus Bengtsson, at the Karolinska Institutet in Sweden, who conducted the research with colleagues.
San Francisco-based Global Viral crunches internet and mobile data to isolate the locations, sources and drivers of local outbreaks before they become global epidemics. Chief innovation officer Lucky Gunasekara says the firm can successfully predict outbreaks a week before the World Health Organization.
“We can implement better policies by analysing this data exhaust,” says Dr Bengtsson. “For example, how people move about in a city is likely to impact on the spread of a new influenza outbreak. Analyses of anonymous mobile phone data by Telefonica from the H1N1 influenza outbreak in Mexico City made it possible to understand how people’s movement patterns changed when the government closed down schools and workplaces to prevent spread.”