- Geolocation data, which is collected by pinging cell phones to track people’s movements, trailed more widely used alternative datasets like credit-card receipts and web-scraping feeds — until the pandemic.
- Hedge funds, banks, and other investors are tracking how people are moving around towns and cities to get a sense of what the recovery from the pandemic looks like in real time.
- Geolocation-data providers are seeing an influx of demand, and bigger players have crept into space as popularity has risen.
- “They’re taking geospatial data more seriously,” said one provider to financial services clients.
Always the bridesmaid, never the bride.
For years, that was the reality for geolocation data, which tracks people’s movements by pinging their cell phones.
The data sets were viewed by Wall Street as a secondary need in the world of alternative data, as they usually had to be paired with other information — credit-card receipts, for example — to create an investment signal. Clientele for the data was often limited to a handful of well-resourced funds and banks with serious data-science chops.
The pandemic changed all that.
With the virus shutting global economies and restricting travel — both internationally and locally — information on how consumers are slowly returning to their past patterns, or have changed them entirely, is in high demand. As a result, geolocation data now has its time to shine.
“We are coming off of one of our best quarters ever,” said Jason Richman, director of enterprise sales at SafeGraph, a geolocation data provider that counts Goldman Sachs and Jefferies among its clients.
Before the pandemic, financial services were not the biggest sector of SafeGraph’s clientele — hotels, governments, and geographic information system (GIS) companies all outnumbered them — but the coronavirus forced investors to take another look at the data.
“They’re taking geospatial data more seriously,” he added.
Firms looked to location data to get a handle on the pandemic
Foursquare, which sells feeds of location data it collects, also saw a significant uptick of interest in its data set this spring from financial firms. Over the past few years, the majority of financial customers were large, systematic funds, Gavin Mohrmann, senior director of business development at Foursquare, told Business Insider.
And while a wider range of investment firms, and even research groups at banks, began showing interest in the data toward the end of 2019 and early 2020, things really took off when the coronavirus hit, Mohrmann said.
As firms scrambled to understand what was going on, Mohrmann said inquires regarding the dataset increased 100x — albeit many were knee-jerk reactions, he added — before eventually settling down to an uptick of around 3x to 5x.
The use cases for the data stemmed from wanting to understand how lockdowns were impacting consumers, he added.
“One, are people actually listening to stay-at-home guidance? Are they actually doing what they’re supposed to be doing?” Mohrmann said. “Second, which vertical/categories are responding more quickly? We understand that travel is affected, but how is travel affected in Atlanta, say, versus travel effected in LA.”
Big tech has noticed the increased attention
The pandemic has shown no signs of being close to ending, with cases once again spiking in states across the country, and geolocation data providers don’t believe the demand will end anytime soon.
Alternative data, in general, is expected to continue to boom as an industry, and the pandemic has sped that evolution up.
A new report from Grand View Research pegs the industry as a $17.4 billion space by 2027, despite only being worth slightly more than $1 billion today. Geolocation data specifically will be a significant driver of that growth, going from an estimated $154 million sector today to more than $1 billion by 2027.
“Location data has likely been put in the spotlight because of what it can do in a period like this,” said Ethan Chernofsky, VP of marketing at Placer.ai, a provider of consumer foot-traffic, during a talk named ‘The renaissance of location data’ at a conference hosted by alt-data provider Neudata.
“We think this is going to be a big market for us in the coming years.”
Increased interest in the space has led some bigger players to wade into the waters. Data from Apple and Google on mobility trends has started to be published in sell-side reports.
“I think to the extent that Google and Apple provide what somebody wants for free, it’s a little bit of a complication for folks who are trying to get paid to provide that thing,” Ryan Preclaw, who co-leads Barclays’ data-science research group, told Business Insider.
But a rise in usage of Apple and Google data from the banks has not concerned the big players in the geolocation space.
The data from the massive tech companies is not at the level a serious finance player would need, SafeGraph’s Richman argued.
“It’s not as simple as knowing how many people are going to this Starbucks at this time,” he said. SafeGraph can tell you where they are coming from, and going to, and people can use Census data to get a sense of household wealth.
“It’s not just the number of visitors, but the types of visitors, the profile,” Richman added.
Foursquare’s Mohrmann said the tech giants simply don’t offer the same accessibility.
“I’d say the biggest differentiator is just ease of use. The way that Foursquare has productized its visit data versus the way Google and Apple is quite different,” he said. “They’re very restricted in how you can interact with them versus Foursquare is more flexible.”
James Crawford, the founder of Orbital Insight, allows users to go onto a platform where all his firm’s satellite and foot traffic data is kept and find their own answers to their questions instead of selling prepackaged datasets. The “diversity of questions” from users, he said, eliminates overlap between different clients and allows them to scale.
There was engineering work needed at the beginning of the pandemic, he added, as the entire platform is run on Amazon Web Services, but they haven’t needed any additional work since, despite a serious uptick in usage — and he expects interest to remain steady.
“Our expectation is that change is a part of our society,” Crawford said.
Already clients like Unilever were monitoring the effects of climate change using the platform.
“We don’t anticipate any reduction in usage on the platform post-pandemic,” he said.
The increased attention on the datasets has also put additional pressure on some geolocation players, many of whom are smaller, niche players. The need to scale up quickly in order to meet the rise in demand has led some to wonder if there is a potential for consolidation in the space.
“People have reset, recalibrated, and refocused on location data,” said Mike Shimerda, director of strategic partnerships for 7Park Data.
“I think it’s inevitable” there’s M&A activity in the space.
Geolocation data still requires some smart engineering
For all the momentum geolocation has gained in recent months, it still has its limitations.
For one, Barclays Preclaw made it clear to get the most precise estimates on spending, geolocation data needs to be coupled with transaction data. Just because a person enters a store or restaurant doesn’t guarantee they will purchase anything. And if they do pay for something, geolocation data can’t provide insight into how much they spent.
“Geolocation data is farther away from the actual revenue and the purchase-intent funnel,” he said.
“Geolocation is always going to be structurally worse [on its own] if your only goal is to basically estimate the revenue for some thing in the world,” he added.
And while there is definite potential in using geolocation data on its own to understand new customer behavior, serious data science is required to pull those insights.
Adam Kelleher, chief data scientist for research for Barclays, said one problem stems from not knowing where a person goes between pings. For example, pings could increase only when a person is using a certain app, leading the data to be driven by a specific factor and biased towards certain locations.
It’s what Kelleher described as a sampling issue, and one that needs to be taken into consideration when using the data.
Users of the data also need to realize that different segments of the population could be over or under represented: Older people may use their phone less than younger people, leading to less pings.
For the data to be useful, these biases need to be sussed out, Kelleher added. Making adjustments is essentially applying a form of social science at a larger scale, something most people don’t have extensive experience in.
“It’s certainly a very specialized skillset. Not one that I would expect a lot of people who are just dipping a toe into data science to have in-house,” Kelleher added. “It’s actually quite hard to hire for, because it’s a relatively new set of technologies that are used to do this type of work.”
Foursquare’s Mohrmann echoed a similar point.
“It’s not at the point where it’s plug-and-play,” he said. “It takes a sophisticated buyer. You need the teams in place. You need to have experience, preferably even having looked at location data at some point. It needs to mesh well with your alternative-data strategy. You kind of need to have questions already in mind that you’re trying to answer.”
