Are We Finally Getting Better at HR Technology ROI?

Individual systems aren't the same as an ecosystem

My role has involved working across an exceptionally complex and disconnected technology roadmap spanning multiple back-office services, and it reinforced something I've seen repeatedly: individual systems can work perfectly well while the overall process works badly.

We naturally organise technology around functions – HR, payroll, finance, recruitment, procurement, IT – but processes and data don't respect those boundaries. A single employee event can trigger activity across several functions, systems and external providers. If each part is designed independently, we end up joining them together afterwards with interfaces, spreadsheets, manual intervention and reconciliation. The systems may all work. The end-to-end process doesn't.

Individual systems can work perfectly well while the overall process works badly.

That matters more now than it ever has. We've spent our careers investing in HR technology, and the latest research suggests we may finally be getting better at generating a return from it.

The numbers are improving

There are encouraging signs. ISG reports that 52% of organisations can now demonstrate measurable HR technology ROI, up from 46%. Mercer's 2026 research is even more positive, with 72.7% saying they're satisfied with their HR technology ROI. Compare that with Mercer's 2024 research, when 49% were dissatisfied or very dissatisfied, and the direction appears positive.

These aren't identical studies, so we shouldn't pretend they form a perfect trend line. But the evidence does suggest we're getting better at extracting value from HR technology.

One number caught my attention, though. Only 14% of organisations say their HR technology is coherent and interoperable. So the interesting question isn't whether we're improving – it's how much value is still being left on the table. Perhaps the biggest opportunity isn't the next technology investment at all, but getting much more from what we've already bought.

And I think that starts with something deceptively simple: design the best end-to-end process first, then design the organisation and technology around it.

Enter once. Serve all needs.

For me, this should be a basic principle of back-office transformation. Enter data once, at the right source – the person whose data it is, or who made (and is accountable for) the decision – and use it to serve every legitimate downstream need.

To achieve that, process design has to be agnostic about who does what today. Start with the outcome. What is the simplest end-to-end process? What data does it require? Where should that data originate, and where does it need to flow? What governance is genuinely necessary? Only then decide who – or what – should perform the work.

Too often we do the reverse. We start with existing organisational responsibilities and existing systems, and try to optimise within those constraints. That may improve individual activities, but it rarely creates the best end-to-end solution.

Integration isn't necessarily interoperability

This is why Mercer's 14% figure matters. We've spent years integrating systems, but connecting System A to System B doesn't automatically create an effective ecosystem.

Real interoperability means the process, data and governance work together. Information should originate in the right place, flow to wherever it is legitimately required, and trigger the necessary actions without repeated entry, unnecessary approvals or manual reconciliation.

The objective isn't to connect every system. It's to create the best possible flow of work, data and information through the organisation. Technology should enable that flow, not dictate it.

That's where the opportunity sits

ISG reports that fully integrated HR technology ecosystems generate approximately twice the ROI of siloed ones. That makes sense, because much of the hidden cost isn't sitting inside the systems themselves. It's sitting between them – in the manual work, duplication, reconciliation and organisational capacity required to compensate for fragmentation.

And this isn't uniquely an HR issue. Deloitte reports that only 50–75% of organisations believe they're getting value from major technology investments across ERP, cloud, data architecture and AI. As back-office technology becomes increasingly interconnected, optimising one function at a time becomes less effective. We need to follow the process and the data across the organisation, rather than stopping at functional or system boundaries.

Maybe optimisation is the next transformation

We've become much better at implementing technology, and the evidence suggests we're also getting better at generating returns from it. But with only 14% of organisations describing their HR technology as coherent and interoperable, there is clearly much more value available.

Perhaps the next question shouldn't be "What technology should we implement next?" but "What value is still trapped in the technology landscape we've already paid for?"

Start with the best end-to-end process. Design how the data should flow. Build the governance around it. Then decide what combination of people, organisation and technology should deliver it.

Enter once. Serve all needs. Simple in principle – and potentially transformational in practice.

Related insight

The Governance Gap: Why HR and Payroll Need More Than a One-Way Integration

The Automation Business Case Starts With the Work, Not the Technology