Can an algorithm predict which businesses will close?

A closed store in New York City. Image: Getty.

Over the past decade, changes in the way people shop have led more and more businesses to close their doors, from small music venues to book shops and even major department stores. This trend has been attributed to several factors, including a shift towards online shopping and changing spending preferences. But business closures are complex, and often due to many intertwined factors.

To better understand and account for some of these factors, my colleagues at the University of Cambridge and Singapore Management University and I built a machine learning model, which predicted shop closures in ten cities around the world with 80 per cent accuracy.

Our research modelled how people move through urban areas, to predict whether a given business will close down. This research could help city authorities and business owners to make better decisions, for example about licensing agreements and opening hours.

Pattern spotting

Machine learning is a powerful tool which can automatically identify patterns in data. A machine learning model uses those patterns to tests hypotheses and make predictions. Social media provides a rich source of data to examine the patterns of its users through their posts, interactions and movements. The detail in these datasets can help researchers to build robust models, with a complex understanding of user trends.

Using data about consumer demand and transport, along with ground-truth data on whether businesses actually closed, we devised metrics which our machine learning model used to identify patterns. We then analysed how well this model predicted whether a business would close, given only metrics about that business and the area it was in.

Our first dataset was from Foursquare, a location recommendation platform, which included check-in details of anonymous users and represented the demand for businesses over time. We also used data from taxis trajectories, which gave us the pickup and drop-off points of thousands of anonymous users; these represented dynamics of how people move between different areas of a city. We used historic data from 2011 to 2013.

Taxiiii! Image: Sunset Noir/Flickr/creative commons.

We looked at a few different metrics. The neighbourhood profile took into account the area surrounding a business, such as the different kinds of businesses also operating, as well as competition. Customer visit patterns represented how popular a business was at any given time of day, compared with its local competitors. And business attributes defined basic properties such as the price bracket and type of business.

These three metrics enabled us to model how closure predictions differ between new and established venues, how the predictions varied across cities and which metrics were the most significant predictors of closure. We were able to predict the closure of established businesses more accurately, which suggested that new businesses can face closure from a bigger variety of causes.


Making predictions

We found that different metrics were useful for predicting closures in different cities. But across the ten cities in our experiment – Chicago, London, New York, Singapore, Helsinki, Jakarta, Los Angeles, Paris, San Fransciso and Tokyo – we saw that three factors were almost always significant predictors of a business’s closure.

The first important factor was the range of time during which a business was popular. We found that businesses which cater to only specific customer segments – for example, a café popular with office workers at lunchtime – are more likely to close. It also mattered when a business was popular, compared with its competitors in the neighbourhood. Businesses that were popular outside of the typical hours of other businesses in the area tended to survive longer.

We also found that when the diversity of businesses declined, the likelihood of closure increased. So businesses located in neighbourhoods with a more diverse mix of businesses tended to survive longer.

Of course, like any dataset, the information we used from Foursquare and taxis is biased in some ways, as the users may be skewed towards certain demographics or check in to some types of businesses more than others. But by using two datasets which target different kinds of users, we hoped to mitigate those biases. And the consistency of our analysis across multiple cities gave us confidence in our results.

We hope that this novel approach to predicting business closures with highly detailed datasets will help reveal new insights about how consumers move around cities, and inform the decisions of business owners, local authorities and urban planners right around the world.

The Conversation

Krittika D'Silva, PhD Candidate, University of Cambridge.

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

 
 
 
 

These maps of petition signatories show which bits of the country are most enthusiastic about scrapping Brexit

The Scottish bit. Image: UK Parliament.

As anyone in the UK who has been near an internet connection today will no doubt know, there’s a petition on Parliament’s website doing the rounds. It rejects Theresa May’s claim – inevitably, and tediously, repeated again last night – that Brexit is the will of the people, and calls on the government to end the current crisis by revoking Article 50. At time of writing it’s had 1,068,554 signatures, but by the time you read this it will definitely have had quite a lot more.

It is depressingly unlikely to do what it sets out to do, of course: the Prime Minister is not in listening mode, and Leader of the House Andrea Leadsom has already been seen snarking that as soon as it gets 17.4m votes, the same number that voted Leave in 2016, the government will be sure to give it due care and attention.

So let’s not worry about whether or not the petition will be successful and instead look at some maps.

This one shows the proportion of voters in each constituency who have so far signed the petition: darker colours means higher percentages. The darkest constituencies tend to be smaller, because they’re urban areas with a higher population density.

And it’s clear the petition is most popular in, well, exactly the sort of constituencies that voted for Remain three years ago: Cambridge (5.1 per cent), Bristol West (5.6 per cent), Brighton Pavilion (5.7 per cent) and so on. Hilariously, Jeremy Corbyn’s Islington North is also at 5.1 per cent, the highest in London, despite its MP clearly having remarkably little interest in revoking article 50.

By the same token, the sort of constituencies that aren’t signing this thing are – sit down, this may come as a shock – the sort of places that tended to vote Leave in 2016. Staying with the London area, the constituencies of the Essex fringe (Ilford South, Hornchurch & Upminster, Romford) are struggling to break 1 per cent, and some (Dagenham & Rainham) have yet to manage half that. You can see similar figures out west by Heathrow.

And you can see the same pattern in the rest of the country too: urban and university constituencies signing in droves, suburban and town ones not bothering. The only surprise here is that rural ones generally seem to be somewhere in between.

The blue bit means my mouse was hovering over that constituency when I did the screenshot, but I can’t be arsed to redo.

One odd exception to this pattern is the West Midlands, where even in the urban core nobody seems that bothered. No idea, frankly, but interesting, in its way:

Late last year another Brexit-based petition took off, this one in favour of No Deal. It’s still going, at time of writing, albeit only a third the size of the Revoke Article 50 one and growing much more slowly.

So how does that look on the map? Like this:

Unsurprisingly, it’s a bit of an inversion of the new one: No Deal is most popular in suburban and rural constituencies, while urban and university seats don’t much fancy it. You can see that most clearly by zooming in on London again:

Those outer east London constituencies in which people don’t want to revoke Article 50? They are, comparatively speaking, mad for No Deal Brexit.

The word “comparatively” is important here: far fewer people have signed the No Deal one, so even in those Brexit-y Essex fringe constituencies, the actual number of people signing it is pretty similar the number saying Revoke. But nonetheless, what these two maps suggest to me is that the new political geography revealed by the referendum is still largely with us.


In the 20 minutes it’s taken me to write this, the number of signatures on the Revoke Article 50 has risen to 1,088,822, by the way. Will of the people my arse.

Jonn Elledge is the editor of CityMetric. He is on Twitter as @jonnelledge and on Facebook as JonnElledgeWrites.

Want more of this stuff? Follow CityMetric on Twitter or Facebook.