In May 2026, Visier published its annual trends report. The central argument is that the people analytics profession must "transform into something larger called workforce intelligence" and shift its north star "from insight to impact." When the strongest people analytics incumbent in the market publicly says the category needs a new name, the people running adjacent strategy decisions should read that as a structural signal, not a marketing line.
Two things are true at once. People analytics has matured into a respectable HR discipline, and that discipline cannot answer the questions a CEO, CFO, or CHRO is now being asked about AI return, workforce capacity, and skill. The first reality is why the category exists. The second is why it is splitting.
This post sits inside the Human Capital Operating System cluster and explains the split: what each side does well, why the boundary moved in 2026, and what an executive should expect from each layer.
The clearest definition of the line
Visier itself reached the simplest articulation this year. People analytics is largely defined by the data HR already knows: headcount, turnover, engagement, performance, compensation, demographics. Workforce intelligence has to understand the work itself: what gets done, how it gets done, which tasks are genuinely human and which are candidates for automation.
Stated bluntly: people analytics looks inward at the HR record. Workforce intelligence looks outward at the work.
The questions executives are now asking sit on the work side of the line. Where is AI actually producing measurable output. Which managers cannot absorb their teams' AI-elevated throughput. Which roles are now structurally different from the role descriptions in the HRIS. Where is workforce spend sitting in tools nobody uses. Every one of those requires signal HR does not own.
What changed in 2026, and why now
Three forces converged this year. They are independent of each other, which is part of why the category is moving.
The first force is universal AI use. The Aon 2026 Human Capital Trends Study, drawn from 2,361 board directors and senior leaders across 62 geographies, found that 73 percent of organizations have deployed or are piloting AI programs and 88 percent agree the workforce will need new skills. Only 18 percent report most of their workforce has actually participated in AI reskilling. Three quarters of the market is past the line where AI is a separate program. AI is now an embedded property of the workforce, and people analytics dashboards were not built to measure embedded properties of work.
The second force is that the return question has migrated to the CFO. PwC's 2026 AI Performance Study reported that 74 percent of AI's economic value is captured by 20 percent of organizations, and the gap is widening. When the CFO is asked whether the company is in the 20 percent or the 80 percent, the answer cannot come from an engagement survey. It requires sourced, auditable signal that ties AI use to work product and to cost. People analytics platforms were not built for CFO-grade numbers. Workforce intelligence has to be.
The third force is that the operating model is itself the lever. BCG's May 5, 2026 report, "Making AI Productivity Deliver Real Value," put it plainly: AI creates more output, but without deliberate choices that capacity is absorbed into existing complexity rather than reducing cost. Real impact comes from redefining roles, decision rights, and workflows. McKinsey's parallel finding: organizations seeing significant AI returns were twice as likely to have redesigned end-to-end workflows before selecting models. Gartner's CHRO survey of 426 leaders put a number on it: 29 percent of AI-driven productivity gains came from changing HR's operating model, not from improving employee AI skills or acceptance.
All three findings point at the same instrumentation gap. What separates AI leaders from laggards is not how many AI seats they bought. It is whether they can see across the work and act on what they see. That is workforce intelligence.
A side-by-side comparison
The clearest way to see the split is across the questions executives actually ask.
| Question | People analytics | Workforce intelligence |
|---|---|---|
| What is the question framed around? | The HR record | The work itself |
| Primary data sources | HRIS, surveys, performance platforms, compensation systems | HRIS plus the tools where work happens (Slack, Jira, GitHub, Salesforce, AI assistants, SaaS telemetry) |
| Primary buyer | CHRO and HR business partners | Shared: CHRO, CFO, CEO, transformation lead |
| What it answers well | Headcount, attrition, engagement scores, comp equity, demographic reporting | AI Impact, throughput, review capacity, behavioral skill, SaaS waste, cost per outcome |
| Time horizon | Backward-looking, cycle-driven | Real-time and forward-looking |
| Decision pace | Quarterly review cycle | Operational, weekly or faster |
| Confidence model | Implied; analyst trust | Explicit; every number drillable with a confidence score and audit trail |
| The 2026 question it must answer | What happened last quarter | Where AI is producing return, where it is not, and why |
The two are not competing for the same buyer. They are answering different questions for overlapping buyers. The mistake to avoid is buying people analytics and expecting it to answer workforce intelligence questions. That is the failure pattern the PwC study is measuring when it identifies the 80 percent of companies not capturing AI's value.
Where Levos sits
Levos is a Human Capital Operating System. It is the workforce intelligence layer above the existing stack, not a replacement. Visier, Lattice, Workday and the rest are good at what they were built for. Levos sits above them, pulls behavioral signal from the tools where work happens, and answers the cross-functional questions the HR stack alone cannot.
Two design choices distinguish it inside the workforce intelligence layer.
The AI Impact signal family is treated as a first-class measurement category, to our knowledge the first product to do so. AI use, AI-attributable output, and AI return on workforce investment are measured as their own family, not derived from engagement surveys or self-report. Every other 2026 source of AI productivity data either reports on access (seats and licenses) or self-report. Levos reports on demonstrated use joined to demonstrated output.
Behavioral skill is derived from demonstrated work, not from taxonomies or self-assessment. People analytics infers skill from survey data and performance review tags. Workforce intelligence infers skill from the operational artifacts of the work itself: what someone has built, shipped, written, closed, or reviewed. The behavioral method is the one that holds up in a board meeting when a CFO asks why the company is paying for a capability nobody can demonstrate.
The privacy posture is the same as the rest of the Levos platform. Individual data flows only to direct managers. Aggregated team views require five or more people. Executive-level exceptions are escalation-based and audited.
What this means for the next four quarters
Three things follow.
The buyer is no longer HR alone. Workforce intelligence has a CHRO sponsor, a CFO co-sponsor, and a CEO consumer. The companies that get this right in 2026 will run joint CFO-CHRO selection.
The people analytics function does not disappear. It moves up the value chain. Day-to-day operational measurement migrates to workforce intelligence, and the people analytics team becomes the partner who interprets patterns and runs harder strategic studies. The function gets more strategic, not less.
The vendor map will sort itself by 2027. Platforms that started as people analytics will either build a workforce intelligence layer above their stack or be acquired into one. Platforms that started as workforce intelligence will absorb the people analytics motion below them where the data is cheap to add. The split stays at the question level for as long as AI is the dominant productivity story, which is at least the rest of this decade.
The category is splitting because the question executives are asking has changed. The companies that can see across the work, on a CFO-grade basis, with a defensible confidence score behind every number, are the ones who will be answering it.
Frequently asked questions
What is the difference between workforce intelligence and people analytics? People analytics reports on HR data and tells you what happened. Workforce intelligence captures signal from the tools where work happens and tells you what is happening now and what to do.
Why is the category splitting now? AI is now embedded in the workforce, the return question has migrated to the CFO, and the operating model itself is the lever that separates leaders from laggards. People analytics was not built for any of the three.
Do we still need people analytics if we have workforce intelligence? Yes. The two answer different questions. People analytics owns the HR-data questions. Workforce intelligence owns the across-the-work questions. Run both and be clear which question each is answering.
What does workforce intelligence measure that people analytics does not? AI Impact, SaaS Stack Intelligence, throughput and review capacity, financial workforce intelligence, behavioral skill from demonstrated work, and the connector signal from the operational stack.
Who should own workforce intelligence inside the company? It is shared instrumentation. CHRO, CFO, CEO, and the manager layer all consume it. Most companies are placing it with a transformation office or chief of staff so the same numbers are visible to all four.
The Levos position
Workforce intelligence is the layer where AI return becomes a measurable property of the work itself. Request early access to the Levos platform. Design partner cohort today: 150 to 500 employees. Expanding to 500 to 2,000 in the second half of 2026.
For finance leaders specifically, the AI Impact signal family is the layer where AI investment becomes auditable workforce return.
