Save content
Have you found this content useful? Use the button above to save it to your profile.
Garmin Vivosmart fitness tracker
Garmin Vivosmart

PwC fitness trackers: Is Big Brother watching?


PwC has been offering real-time biometric trackers to employees in a project to monitor fitness and stress levels. Maddy Christopher investigates what PwC is monitoring and how beneficial the trackers are to employees.

8th Sep 2021
Save content
Have you found this content useful? Use the button above to save it to your profile.

In 2019, PwC offered 50 Garmin Vivosmart fitness trackers to employees as part of a scientific project to monitor and improve staff stress and fitness levels. After a successful pilot, 1,000 trackers were offered to employees in 2020.

When the scheme was offered more widely, it was inundated with applications from twice as many employees as the first trial.

The firm partnered with human performance consultancy iHP Analytics, which usually operates in elite sports including the Olympics and Formula One to support competitors’ performance, health and wellbeing. PwC wanted see if iHP Analytics’ advanced fitness tracking methods would work in a business environment. 

“I can see why people would potentially think it could be an intrusion until they hear that the level of care and caution and time we’ve gone through this,” PwC director of artificial intelligence Rob McCargow told AccountingWEB. 

But “we can go through the nuanced the detail, step by step in terms of how data is collected, stored, deleted, and utilised.”

Former PwC consultant and founder of flinder Alastair Barlow could see the purpose behind the wellbeing project. “It’s about being informed as to where we are against a target and taking action,” he said. 

“Without it, we’re guessing – it’s exactly the same in business too. Without knowing how we’re tracking against budget, then we’re relatively blind until the end of the month, quarter or even year. With real-time data and insights, we can take much more proactive and timely decisions towards our goals.”

Phase one

PwC put together a team that included a full-time medical doctor alongside a data ethicist, experts in regulation, lawyers, communications experts, technologists and data scientists.

The 1,000 applicants were selected to reflect PwC’s composition by grade, office location, sex and so on. Participants included senior executives including chairman and senior partner Kevin Ellis. 

Participants were psychometrically baselined to provide personality profiles, in addition to working environment. “The critical bit was [seeking] their permission to anonymously access their timesheet data,” said McCargow. 

“There were so many consent emails sent to us,” reasoned PwC director Victoria Broadhurst, another volunteer for the project. “If I had any concerns, there were many moments I could have backed out. I knew what information they were going to look at, what they were using it for and how it would benefit me,” she added.  “And there were so many people doing it that from an aggregate perspective, they wouldn’t have been able to know it was me.”

The information harvested from particiants included:

  • biometric data 

  • cognitive function data

  • psychometric baseline data

  • contextual data.

Over four months, the trackers picked up information on heart rate variability, sleep, rest and recovery. The platform also got participants to play games each day to gather data on their ability to switch tasks, pattern memory, response times and other cognitive function indicators

Monitoring, security and personal data

Personal data can only be viewed by the respective participant – no individual data can be viewed by PwC. Instead, a single, anonymous aggregation of all data is made available to the analysts along with overall hours worked, project types and how busy diaries were. 

“It’s not tracking things like where they’re working from a location perspective, that feature is disabled,” commented McCargow.

The wellbeing score compiled from this information is an “aggregated, anonymised, organisation-wide” measure of all the fitness tracker users. Applying a set of proprietary algorithms to the data allowed the project team to examine the key factors affecting staff wellbeing during the four-month period.

Phase two

The second phase focused on direct needs of the Garmin users. Participants could access their personal dashboard to view biometric outputs like sleep per night the consequent cognitive effects.

Participants could see, for example, if a day of 15 back-to-back video calls impacted sleep or their ability to switch tasks at 4pm on a Thursday Was this pattern of work having a specific impact on exhibited stress?

“The idea is that we equip our staff with real-time insights on their wellbeing within the context they are operating in, [to] understand how to make positive change. And then we can tailor interventions to support them,” explained McCargow. 

“Wellbeing became a huge focus. Last year it felt like a blunt instrument to tick the wellbeing box by saying that you'd offered a meditation app to everybody. Job done. What we now believe, with our human-science experiment is that you've got to find a way of empowering people individually within a specific situation they're facing and give them a targeted level of support.” 

In phase two, the Garmin data goes through more aggregation and anonymisation to create organisation-wide, real-time viewpoints of PwC staff. iHP’s data helped PwC assess the impact of hybrid working based on unbiased staff data.

“The intentions for hybrid working really could be profoundly positive for staff across the profession,” McCargow said. 

“We’re calling it empowered flexibility, giving staff autonomy on where and when they work. We’re also trying a compressed four and a half-day working week. If we [ensure] a real-time assurance mechanism is in place, we can correct the policy and test if the experiment is actually leading to positive outcomes for staff.”

Preliminary findings

Participants in the pilot scheme were asked how stressed they actually felt for a comparison with biometric stress results. The poor correlation of these results indicated that staff often felt more stressed than they actually were. “So we’re working on providing training on how to reduce feeling stress during periods of stress,” said McCargow.

“You get a bit obsessed when you first get it, checking your sleep and so on,” admitted Broadhurst. But after seeing that she was sitting at her desk all day every day, she decided to take up the Couch to 5k running challenge, which she continues to keep up. “Where I thought I was quite stressed, I wasn’t – but my body battery was quite low where I wasn’t relaxing. 

“It gave me information to think about whether I getting enough rest, sleep and was overthinking from a psychological perspective.”

The experiment definitely helped with mindfulness and being present, she added.

Users voluntarily disclosed how the data increased their awareness of what makes them tick. “For the first time, we’re starting to quantify wellbeing,” said McCargow. “We can now provide a score on wellbeing which we can benchmark and track, rather than asking people and getting varying answers. 

“The data has demonstrated that we have to do a far better job of catering to people’s particular needs, their environmental conditions at home and their own preferences in work style [to] personalise the way they deliver work for the organisation.” 

Control and release

PwC has made its stance on using the data from the offset. “There is no way technically, or from a regulatory and data privacy perspective to allow this to be used in any performance management decision,” insisted McCargow.

PwC maintains fitness tracking will always be voluntary. Users can leave the scheme at any time and have their data wiped if they choose. Equally, they can dip back in and look at their private data and then leave. PwC’s AI director emphasised that the project was about staff empowerment with no compulsion or hidden implications in taking part. 

“My primary focus has been developing responsible AI and responsible technology,” said McCargow. “For me, getting this right was critical. The people around me include medical doctors, ethicists, regulatory experts and 60 representatives across all grades in the organisation in the UK and I openly consult them about the stages of the project.

“If there’s no inherent trust between the participants and the employer, these things just simply don’t work. Our technology will always be strictly governed, entirely voluntary and transparently co-created with workers. I won’t deviate from that, ever.

Replies (0)

Please login or register to join the discussion.

There are currently no replies, be the first to post a reply.