When Niall Silvestro, team leader reward and analytics at Aberdeenshire County Council, first took on his role, benefits strategy at the local authority was based mainly on intuition. A decade later, things are very different. The council, which employs around 16,000 people, uses a highly data-driven approach.
The team developed data-driven streamlined workflow design, reducing administrative burdens. This makes it easier to prove a business case in a cash-strapped environment, says Silvestro. “The conversation [about any new benefit] usually goes: how much is it going to cost to administer? Data-driven workflows mean our costs for running benefit schemes are lower,” he adds.
One example is the workflow for salary sacrifice applications. If an employee wants to buy a new appliance through the council’s home and electronics scheme, offered via Perkbox Vivup, eligibility checks are required. “We dynamically pull information out of our HR and payroll system and other sources and put it in one place,” Silvestro explains. “Historically, those schemes were more manual. People had to go to multiple different systems. Now the team knows in a couple of seconds if they can approve the application.”
This improves the employee experience of benefits schemes, he adds: “[Employees] don’t want [their] employer to be sitting for a week on a decision when [their] fridge is broken.” Such decisions can now be made within one working day. The next step will be to automate the approvals process, shortening the time even more.
Data is used to analyse how benefits contribute to areas like retention in pinch points in the organisation, such as social care. “We look at the impact benefits have on our different job cohorts,” says Silvestro. “For example, is there evidence we can retain a group longer if they actively engage with the benefits?”
The team is also using sentiment analysis to track how people feel about benefits via HR service desk interactions, exit interview data and other sources.
Silvestro’s advice to others that want to become more data-driven is to “find your own starting line”. “Finding that baseline is important,” he says. “Once you’ve got that, the data itself is not that scary.”
