We used big data to divide city-dwellers into six “tribes”

A generic picture of New York City. Image: Getty.

Whether tapping a contactless card to take the subway, buying a gift for a loved one or using a smart phone to find the way home, your everyday actions leave behind breadcrumbs of digital information. When these traces are collected and analysed en masse, they can help urban planners to pick up on the daily rhythms of the life of a city and uncover trends about the people who live there.

In a recent study, we developed a new mathematical framework which extracts various spending habits or lifestyles from the digital traces left by credit cards and mobile phone data in Mexico City. We discovered that purchase habits not only related to socio-demographic characteristics such as age, gender, income and mobility – they also related to the places people visit, and the people they call.

Over the past decade, smart phone data has helped researchers to understand and plan cities. For example, location data is used in transport planning to identify which stations and routes are busiest at different times of the day. For example, scientists have found that people don’t necessarily take the optimal route to their destination – instead, they have a favourite route for trips they perform routinely, and a few alternative routes which they take less frequently.

Meanwhile, credit card data and spending behaviour data have been used from the retail companies to build a profile of consumers, based on their set of purchases. This is illustrated by the parable of the beer and nappies. As it turns out, on Friday evenings, young men who buy nappies also have a predisposition to buy beer – this group became parents.

When analysed across the years, these digital traces can help scientists and governments to understand at an unprecedented scale how societies in different parts of the world cope with major events, such as recessions or major policy changes.

The six tribes

From the detected spending habits, we identify six groups or “tribes” while protecting people’s privacy by aggregating the data and ensuring they were anonymous. Each of these groups has a core purchase, which is the most frequent of their spending activities, indicated by the yellow arrows in the figure below.

Our six lifestyle groups: the colour of the arrows represents the frequency of the transactions, from yellow (most frequent) to red (less frequent). Image: Riccardo Di Clemente et al./author provided.

The “commuters” tribe is mainly made up of adult men who live far from the city centre. They commute by car (paying tolls), earn above average incomes and enjoy eating at restaurants.

Middle-aged women dominate the “household” tribe and tend to have the least expenditure and mobility. Their core transaction is grocery shopping, they have lower incomes and live in the suburbs. On average, members of this group receive more phone calls from the other tribes.

Young people are split between two tribes. The “young”, who are under 30 years old, live in the city centre, mainly use taxis as means of transportation and have an average income.

The second group, called “high tech”, is slightly older, with an average age of 35. Their core transactions are on technology such as smart phones and computers. They could be young professionals, since they have higher than average expenditure, a wide range of contacts in their mobile phones and do most of their activities downtown.

The “dinner out” group lives closer to the city centre. Their core transactions are restaurants. They also have more contacts in their social network than average.

We also created an “average” group from a random sample of citizens, to use as a benchmark, against which we could compare the spending habits and socio-demographic information of the other tribes.


Helping urban challenges

By generating useful information from these digital traces, we hope to help cities to embrace the digital revolution and resolve important urban issues, such as how to make cities more inclusive.

For example, the commuters tribe may be hit hardest by fuel price increases which may affect their spending and commuting. The creation of inexpensive and efficient public transport systems may be an important investment in urban areas, especially those where low-income residential areas and job opportunities are not in proximity. And to help the household tribe, we could consider introducing nutrition subsidies – programmes to incentivise low-cost grocery stores in areas where it’s difficult to find affordable, good quality food.

What’s more, our study shows that it is possible to gather information about daily mobility, social contacts and lifestyles by collecting the data that people already produce, rather than undertaking surveys which are expensive, time consuming and very limited in sample size.

This kind of analysis can help city authorities make informed decisions based on data about an unprecedented number of people and lead to more tailored policy to the specific needs of different groups and lifestyles.

The Conversation

Riccardo Di Clemente, Newton International Fellow of Royal Society, UCL and Marta Gonzalez, Associate Professor of City and Regional Planning, University of California, Berkeley.

This article is republished from The Conversation under a Creative Commons license. Read the original article.

 
 
 
 

Does it matter that TfL are renaming White Hart Lane station Tottenham Hotspur?

New White Hart Lane. Image: Getty.

Pretend for a moment that you’re travelling in the London of 1932. You’re taking the Piccadilly Line northbound and alight at Gillespie Road station. The name should be obvious: it’s inscribed in bespoke brown tiling on the platform.

But that 31 October, following an intense campaign by the eponymous football club, the London County Council changed the station’s name to Arsenal (Highbury Hill). The area’s growing association with the name “Arsenal” ended in a lengthy negotiation that changed maps, signs and train tickets alike. Football had acquired so much power that it changed the name of not just a Tube station but an entire suburb, even before the era of Wenger or the Emirates.

Now the spectre of name changes is on the horizon once again. As Tottenham Hotspur FC inches closer to completing its new stadium, the club is clamouring for a renamed Overground station. Despite the fact the new stadium is located on almost exactly the same site as the old just off White Hart Lane, and fans have long been calling the scaffolding-laden mess “New White Hart Lane”, the club’s executive director is adamant that the station’s existing name cannot stand. White Hart Lane station, on the Overground line leaving Liverpool Street, is set to be renamed “Tottenham Hotspur”, at a cost to the club of £14.7m.

Little has been made of the fact that this peculiar PR kerfuffle is tied to Spurs’ failure to convince Nike to sponsor the venue. Some sources have even claimed that the sponsorship is yet to be finalised because it is somehow contingent on the renaming of the Overground station; beyond the ridiculous Johnson-era vanity project that was the Emirates Air Line, it seems improbable that TfL will allow any more corporate-flavoured information pollution. There will be no “Nike Stadium” station on the way to Enfield, much as there is no “Emirates” on the way to Cockfosters, especially if public consultation gets a look in.

The scene of the crime. Image: TfL.

But there’s a problem with the new name, all the same. “White Hart Lane” already means “football stadium”, in the same way Loftus Road or Stamford Bridge do. Changing it to “Tottenham Hotspur” risks opening the floodgates to an “O2 North Greenwich” or a “Virgin Euston” at some point in future, names as banal as there are dystopian. The Greater London Authority has promised to spend the £14.7m fee on community programmes in the local area – but that’s not much money to set the precedent that a private company can mess about with the Tube map.


What’s more, as CityMetric has often observed, there are plenty of station names across London that could do with a tidy up. Picking one that’s perfect already and asking for £14.7m to change it is adding insult to injury. How much would it cost a community group if they asked to change the name of Goodge Street to Fitzrovia? Why does a vast corporate entity backed by international sponsors and thousands of season ticket holders get to set the standard?

Back in Arsenal’s day, changing names on the Tube must have been easy; changes could be accommodated gradually without bothering the every day traveller. But in our world of online information, maps and apps, name changes are rather more complicated.

The question is – if TfL can bring itself to balefully accept this particular proposition, why can’t it accept ours? Why sort out a single non-issue on the Tube Map when you can catch lots of real ones in one go? A day’s pandemonium might just be a price worth paying to fix the Bethnal Greens problem once and for all.