Every conference attendee typically seeks two things: a meaningful professional connection and a practical idea he can implement in his company. At Visier Outsmart earlier this month, Adam McKinnon, head of people platforms and analytics and Reece Group, offered the latter, providing a practical example of how analytics can drive real business value. To do so, two real-world use cases from the Reece Group were presented, demonstrating how people analytics has been applied to solve critical operational challenges.

Reece Group is a 100+ year old, predominantly family-owned Australian company that began in plumbing supplies and has grown into a multi-national organisation. With more than 900 branches across Australia and New Zealand, and a significant presence in the US following a 2018 acquisition, Reece generated $3.8 billion in ANZ revenue and $5.3 billion in the US in the most recent financial year.

Reece is a six-day-a-week retail business with a large frontline workforce. In such a context, business performance is deeply tied to having the right people in the right roles, at the right time. Recognising the need for strategic HR support, the company invested in building out its people analytics capabilities. What began as a one-person function evolved into a team of four, supported by two additional team members in compensation and benefits, built around a strong technology platform that enabled scalable delivery of insights and value.

Case 1: Recruitment performance and turnover

A pressing challenge in Reece’s workforce strategy was the increasing time to fill frontline roles and the high turnover of new hires within their first year. The HR team sought to address two core objectives through a data-informed intervention: reduce time to fill and reduce early tenure turnover. These were not theoretical goals — they represented urgent and tangible issues for the business.

Business case development: Opportunity cost

Reece needed to fill about 1,100 roles per year, averaging around 83 to 84 per month. At the time, the average time to fill a role was 33 days. With each branch open six days a week, this delay translated into significant missed revenue opportunities. On average, a frontline employee generates about $500 in sales per day. For business case purposes, however, a conservative assumption of $150 per day was used, acknowledging that new hires require time to reach full productivity.

Using these figures, reducing time to fill by just five days for 83 roles per month could yield over $62,000 per month in reclaimed opportunity cost. Assuming the new technology, Sapia, was used for only one in four roles, the six-month projected benefit came to over $93,000 — more than enough to cover a hypothetical $75,000 subscription cost.

Turnover cost analysis

The second and more critical side of the business case addressed turnover. In the prior financial year, Reece lost 277 new hires in its first year. The estimated total cost of this turnover exceeded $6 million, with a per-person cost of over $22,000. To justify the investment in Sapia, the team needed to save just four new hires, representing a 1.4% reduction in first-year turnover.

With this in mind, Reece proposed a 12-month proof-of-concept agreement with the vendor, committing to a highly transparent evaluation process. The vendor agreed.

Implementation and findings

After around ten months, a full evaluation was conducted. Results were mixed. When time came to fill, the combined use of Sapia and human decision-making sometimes led to longer turnaround times than using human assessment alone. Overall, there was only a modest improvement in time to hire, likely influenced by the Hawthorne effect, where individuals modify behaviour due to awareness of being observed. In some business units, however, time to hire did decrease noticeably. Upon deeper investigation, the team discovered inconsistencies in how the technology was being applied by recruiters and across departments, revealing a need for process standardisation rather than a fault in the technology itself.

Survival analysis: Impact on turnover

The more impactful result came from the analysis of turnover. Sapia’s recommendation engine rates candidates as ‘Yes’, ‘Maybe’, or ‘No’. The team examined the outcomes of hires made based on ‘Yes’ recommendations and compared them to hires made entirely without Sapia.

Using survival analysis, a statistical method from epidemiology, the study tracked the tenure of hires across 300 days. The findings showed clear separation between the two groups — those hired through traditional human decision-making were 1.8 times more likely to leave within 12 months compared to those recommended by the technology. This difference was both practically and statistically significant.

In total, 176 hires were made on the back of a ‘Yes’ recommendation. Based on historical turnover rates, 86 departures would have been expected. In reality, only 25 left. This equated to 61 net retained employees.

Using a conservative cost of $4,500 per leaver, the gross benefit of this retention was calculated at $275,000. After subtracting the $75,000 subscription cost, the net benefit was $199,000, yielding a 266% return on investment.

This result provided the kind of quantified, risk-reducing evidence that senior executives want to see. The intervention was no longer a gamble; it was a validated business decision.

Case 2: Automating employee case management

A second, simpler but equally revealing use case involved automation within the HR service delivery function. Reece’s centralised People Services Hub fields all HR inquiries from its over 900 branches. The six-person team handled over 30,000 tickets in the last financial year, representing a 42% increase over 18 months. No additional headcount was approved. The team was asked to innovate.

The data showed that each ticket required an average of 9.6 updates, and that update volume translated to 544 days of work annually, assuming one minute per update and continuous eight-hour workdays. With four to six months required to onboard and train a new team member, the current workload was unsustainable.

Using data to find automation opportunities

Using Visier, the HR analytics team worked with the People Services Hub to identify automation opportunities, specifically cases with high volume, high manual touch, and low complexity. One clear candidate emerged: roster requests. These accounted for about 1,200 tickets per year and, although seemingly straightforward, were surprisingly resource-intensive.

The team mapped the process end-to-end and identified that four teams and four systems were involved in a manual, handoff-heavy procedure supported by a rigid decision tree. While the process had a degree of operational complexity, the logic was highly rule-based and repeatable, making it a strong candidate for automation.

Turning concept into implementation

Finally, an European generative AI vendor was engaged to build a fully automated solution for the roster request process. Although not yet fully in production at the time of reporting, the projected improvements were substantial.

Previously, the process involved four teams, took up to three days, and required multiple handoffs. The new system reduced involvement to a single team, and once deployed, would allow the task to be completed in under one minute.

This initiative demonstrated the potential of combining data analysis, subject matter expertise, and automation to reduce manual work, free up team capacity, and deliver more timely outcomes to the business.

One tip for building a people analytics function

When asked what advice should be given to someone starting a people analytics function, McKinnon's response was clear: choose your battle. Identify a business challenge that is meaningful, measurable, and aligned to the organisation’s priorities.

The key takeaway from both examples — recruitment and case management — is that analytics must be connected to clear outcomes, and success should be defined upfront. For Reece, the initial focus was turnover, and through consistent measurement, the team was able to demonstrate real business impact.

The final thought offered was light-hearted: "That’s enough preaching — my own wife doesn’t listen to me for this long." But behind the humour was a firm conviction that people analytics, when applied thoughtfully and measured rigorously, is a powerful lever for organisational value.