How to turn people insights into business impact with AI

Explaining why it’s happening is much harder.

At AI in HR 2026, Marian John Müller, Senior Account Executive at Eletive, shared how HR teams can use AI to move from the what to the why, and turn people insights into outcomes the business actually feels.

HR knows the what. The why is harder.

If you are using our Eletive platform, you as HR already track plenty of hard KPIs: turnover, absence, hiring rates, and sometimes profitability through correlation analysis in tools like Power BI. They’re easy to measure and easy to trend over time.

The problem is that hard KPIs are lagging. By the time turnover rises, the reasons behind it have usually been building for months. Rely on them alone and HR stays reactive.

That’s where soft KPIs come in. Engagement scores, sentiment and eNPS tell you what’s driving the hard numbers, early enough to do something about it. In John’s words, understanding the what isn’t enough. Understanding the why is what moves the hard KPIs.

And acting on it doesn’t have to mean big initiatives. Often the most powerful thing is helping managers have better conversations with their teams, faster and with the right focus.

A SMART way to set AI goals

John used a familiar framework with one addition for AI:

• Specific. Be clear about the problem before you build a process around it. Start with the hard KPIs or business objectives you’re accountable for.

• Measurable. Combine quantifiable scores, which give you trend lines and internal benchmarks, with qualitative signals like sentiment in open-text comments.

• Actionable. Only measure what you can act on. Collecting data and doing nothing with it creates frustration, not a better culture. Link every action to a driver and let AI flag where to focus.

• Relevant and responsible. Move with purpose, but not so fast that you risk anonymity breaches or data issues. Regulations differ between the EU, Switzerland, the US and elsewhere, but they all point the same way: handle people data responsibly.

• Timely. Shorten the time from insight to action. Survey results that take weeks to analyse end up acted on when nobody remembers what they answered.

Bring your own AI is already here

78% of employees who use AI at work bring their own tools. It’s even more common in small and mid-sized companies. Meanwhile, 60% of leaders say their leadership lacks a plan and vision for implementing AI.

Put those together and you have a real risk. Without clear frameworks, people paste sensitive data into whatever tool they like, build their own agents and run their own experiments. Knowledge isn’t shared, and people data ends up where it shouldn’t.

John’s recommendation: put clear processes in place, train people on both the uses and the risks of AI, and agree policies on which models you use, where they run, and what data goes into them. Because the technology changes so fast, a one-off reminder isn’t enough. This needs ongoing education, with a human in the loop who knows enough to fact-check what AI produces.

It’s also why Eletive builds AI into the platform. Aggregated survey results, lifecycle data and comments can be analysed where they already are, without being exported to another tool.

Make the business case in business language

Getting budget for AI isn’t just about a licence fee. More and more tools also charge based on usage. So HR has to show what the investment is worth.

The clearest way to do that is to flip the question. What does it cost the business when engagement drops? Lower engagement leads to lower performance, then higher turnover, lost productivity and disrupted teams. Some of this is hard to put a number on, but the long-term impact is real.

Speak the language of the board. Frame AI as an investment in faster, better decisions about people. Then start small and build from there.

Three signals, three business risks

John shared three examples of soft signals that turn into hard business outcomes if no one acts on them.

1. eNPS is dropping. Fewer people would recommend you as an employer. Left alone, that hurts hiring and retention. With AI, you can ask Copilot why eNPS is falling and get a correlation analysis across engagement and lifecycle surveys, showing which drivers matter most to your detractors.

2. Manager support drops in customer success. It’s one team, not the whole company, but it matters. Customer-facing teams shape whether customers would recommend you, and that feeds straight into revenue. AI can analyse the trend over time and check whether manager support is the only issue, or whether workload, autonomy or team relationships are also playing a part.

3. Attrition risk rises among senior sales reps. Engagement and sentiment decline quietly. The business risk is losing high performers, customer relationships and momentum in new markets, plus a harder time onboarding replacements. Spot the soft signal early and you have time to act before it shows up in the hard numbers.

Five takeaways

• Start from the hard KPI you already own, then work back to the soft signals.

• Understand why. Let AI help you find the why, not just the what.

• AI surfaces the action, but a person owns it. Make someone accountable for moving the score.

• Handle sensitive data safely. Governed by design, not retrofitted.

• Reframe the budget conversation. The real cost is what happens when a signal sits unused.

Watch the full session

Missed it? Watch John’s session