Meet Movvo, the Portuguese startup that provides footfall analytics without collecting any personal data at all.

Initially developed by Diana Almeida, Roberto Ugo and Suzy Vasconcelos at Oporto University, Movvo specialises in location intelligence, using mobile phone radio signals to track the movement of people around shopping centres in an extremely simple way.

Credit: Movvo
Credit: Movvo

How it works

Movvo collects two sets of signals to create an accurate dataset: cellular (2G, 3G, 4G) and WiFi on an almost real-time basis, so it sees each device between twenty and thirty times a minute.

"The system sees every consumer device close enough to real-time to be real-time, locating each signal within a square metre of where it came from," says Movvo CEO Cyrus Gilbert-Rolfe.

Retailers need to understand how the flow of traffic around shopping centres works, and many use analytics to locate which area receives the most footfall to determine where an ideal shop location would be.

"It is a passive and anonymous system" says Gilbert-Rolfe, meaning you don't have to opt in or download an app. The only thing needed is that your mobile phone is on.

The idea that mobile phones can be used as a proxy is not new in business. However the ability to see not only all mobile devices, but also see how many people there are and where a person is remains extremely valuable to retailers.

"Data like this can be used to segment the market. We can find out which shops a person who eats at Nandos goes in and how long they spend in there relatively easily."

But what is the cost to personal privacy? While no personal data is collected or stored, the fact that there is no opt out function may leave some feeling a little vulnerable.

Data privacy laws

What sets Movvo apart from other location tracking services is that "Movvo collect no personal data at all", says Gilbert-Rolfe.

"When we see the map address of a phone we one-way encrypt it then we throw it away, so when people will ask questions like "how often do people come to the shopping centre?" we can't answer that because at least once a day we've erased everything"

"If they say "do more people come on a Tuesday than a Thursday?" then we do know that because we have the top line tabulated numbers", he added.

In Portugal, strict data protection laws meant that in order for Movvo to enter the data collection and location tracking market, they must get certified to do so.

Enter EuroPriSe: the 'European Privacy Seal' that assures all data collected is in line with European data protection laws.

EuroPriSe monitors and audits Movvo regularly to make sure they are in accordance with EU law and has even carried out randomised unannounced checks.

At one point Movvo had to surrender all its database usernames and passwords to EuroPriSe to prove they were following EU regulations properly.

Movvo's algorithm

To avoid revealing a person's identity through the shopping centre's CCTV being cross-referenced with a person's location derived from their phone, Movvo created an algorithm that scrambles the data collected.

Gilbert-Rolfe says: "Firstly it has to decrypt all of the headers on all of the data packets that it sees."

"We install proprietary hardware inside the places that we work with and reduce that to a set of devices and then you look at the locations that have been captured with each ping", he added.

Movvo can generate a real-time count that estimates the amount of people inside a participating shop.

A challenge to Movvo's real-time headcount is the number of devices that a person will have on them at one time.

For example, while the majority of people will only have one smartphone on their person, a number of people will have a phone, a tablet and perhaps a laptop also. This initially made tracking the amount of people in an area difficult and insufficiently accurate.

To combat this, Movvo's algorithm "will look at the path that every device took and it will say if there's a 99.999% overlap between the paths these devices took, we are going to assume both devices are in the same person's pocket" says Gilbert-Rolfe.

Equally, this can be used for devices following an almost exact path that at some point do separate. For example, a couple could be holding hands or walking very close to one another, the algorithm will detect two devices but will ultimately deem they are two different people by tracking the devices to when they separate.

What's next for Movvo?

"It's a fast evolving industry but the questions are really old...I think what will happen next is we'll figure out which questions people would like to ask," says Gilbert-Rolfe.

"There is a lot of noise around context based marketing. We are trying to get to a point where there are all kinds of services that we would provide if you knew where people were and what they needed."

"We see a large part of the market for Movvo being around health and safety. Because we are in real-time we could tag people at the start and make sure you have the right people in the right place based on a difficult situation."

If we think about crisis situations and disaster relief, tracking the location of people would be extremely valuable.

"It's easy to imagine a crisis situation because we could locate a lot [of people]," he added.