Part 1 showed what ten reports said about 2026. Part 2 asks the harder question: why do they all say the same thing?

When ten independent houses converge on a single agenda, it looks like the truth. That's exactly when to look twice.

A trend report isn't a measurement of the workplace. It's a product — and like any product, it's built to find what's new, not to test what's true. What you get is a map of the conversation, not the work: some of it essential, some of it this decade's version of last decade's fad.

When everyone agrees, be suspicious.

That's The Noise — Part 2 of the trilogy.

— Tanguy

Link to heatmap → issue #033 (or the PDF download).

Part 2 of 3 — why ten reports converge on one story, and what their consensus leaves out.

By Tanguy Dulac, Founder & Managing Director, PeopleCentriX

Part 1 left a question: a whole field moved to the same AI vocabulary inside twelve months — is it reading the world, or reading each other? Mostly each other. That is a structural claim, not a cynical one, and it has a test.

The fad isn’t AI — it’s that everyone says the same thing

AI is not the problem — it is the real shift of the decade, and Part 1’s evidence, a fifth of all jobs churning by 2030, says so. The problem is narrower and stranger: every report now says almost the same thing. Ten publishers, one vocabulary, four-fifths of the people agenda dropped to make room.

How do you tell a real shift from a fashion? Not by its subject, and not mainly by its speed: AI is a genuine general-purpose shock, and real shocks can move fast. The tell is displacement. A real shift adds; a fad substitutes. Between 2025 and 2026 the AI-framed share of trends leapt from 11% to 75%, and as it surged the rest of the agenda fell silent — wellbeing, fair pay and worker protection dropped off the page, with nothing in the evidence to say those problems were solved. That is the signature of a management fad: not novelty but uncompensated displacement, prior themes abandoned in unison with no underlying shift to justify it (Miller and Hartwick, 2002; Briner, 2024). The speed and uniformity corroborate it; the displacement is what gives it away.

Absorbed, or abandoned?

What has dropped is the whole people-sustainability agenda, not wellbeing alone. We take wellbeing as the illustration because its evidence of outcomes is the strongest; diversity and inclusion, fair pay, or any other strand of that agenda would make the same case.

Wellbeing did not merely slip down the list; it fell out of the ranked field altogether. It had held a top-ten place in both 2024 and 2025 — forced up the agenda by the pandemic and, tellingly, staying there well after the crisis eased — yet by 2026 it sits outside even the top twenty. Perhaps, the defence runs, it left not because the field tired of it but because it was absorbed, too embedded to count as news. It is a fair hypothesis.

The international evidence refutes it: absorption predicts the problem is easing, yet across psychological, physical and financial dimensions it is worsening. Gallup puts global engagement at 20%, its lowest since 2020, with 40% of workers stressed for much of the previous day and more than one in five lonely at work; PwC finds 59% financially stressed; Deloitte’s 2025 survey across 44 countries finds 40% of Gen Z anxious or stressed most of the time. The decisive signal is the gap between activity and outcome: employers have expanded wellbeing rhetoric, platforms and programmes for years, and these numbers have not improved. A problem that had been solved would not deteriorate as it left the page. This is abandonment, not absorption; on the evidence, wellbeing belongs near the top of the agenda, not off it. The same dynamic can scratch any still-important theme from the top twenty the moment a buzzier story arrives — its fall a measure of fashion, not of whether it still matters.

None of this is a new accusation; it is a documented mechanism. Forty years ago Eric Abrahamson named the market that produces it: a fashion-setting community — consultancies, analysts, business media and the business schools — that competes to sense, label and broadcast the next management technique faster than its rivals, then moves on, whether or not the last one worked (Abrahamson, 1991, 1996). The trend report is that machinery, running on an annual cycle.

I have watched the cycle turn for two decades, and it runs back further still. Each era has its compulsory themes, announced as transformational, cresting across every report at once, then receding as the next arrives.

Four decades of HR fashion — each cohort announced as transformational, most have since receded.

Few were false. They were fashions — real enough, oversold, abandoned before the work was finished. AI in 2026 is the same shape at greater amplitude: the newest entry in a long catalogue, not a break from it.

So the reports do not map the workplace. They map the conversation about it — what the field is discussing, not what is happening inside organisations. Call it the consensus mirror: a profession that, looking for the workplace, sees only its own reflection. The reflection is useful, as long as you do not mistake it for the room. Most readers do. Four reasons the gap matters.

One — The forecast is written by the people selling the cure

Nearly every publisher monetises the problems it forecasts. Research houses sell subscriptions; consultancies sell transformation; platforms sell courses; technology vendors sell the systems. A trend report is an analysis and, structurally, the top of a sales funnel. That does not make it false — it makes it selected: themes that win engagements get amplified, themes that don’t get quietly retired. A market converging on what currently sells is not the same as a workplace changing.

Two — Not all evidence is equal

Stacked in a carousel, ten logos look like ten equal data points. They are not. Some reports rest on global surveys of thousands; others on member panels, single-analyst syntheses, or editorial timed to a publishing calendar. Averaging a 7,000-person survey and an opinion piece into one “consensus” is how an echo becomes evidence. A theme carried by nine surveys and one carried by a single essay are not the same finding, however identical they look in a list.

Three — The consensus is not your context

In our own practice, the pattern holds in every people strategy we build from primary evidence — interviews, focus groups, workforce data: the consensus top ten captures at most half of what actually matters to the organisation. We recently built one for a large European research organisation, using desk research and consultations with executives, managers, and employees, conducted through 35 interviews and 25 focus groups. We then audited its ten objectives, line by line, against the publishers’ 2026 consensus. Four sat inside the consensus top ten; four fell into the 11-to-20 band, which barely registers; and two did not appear in the top twenty at all — among them, wellbeing and performance, the themes the field had deleted.

Not because the organisation is behind the curve, but because real priorities come from context — a demographic cliff, a funding cycle, a regulatory regime, a culture with its own history. The reports describe the average of a global conversation, and no organisation is average. A map that is half right, with no legend telling you which half is yours, has no decision value.

Four — A technology agenda in a people vocabulary

The 2026 “people trends” fail the WEF’s own test of good work: only employability advances, while fair pay, protection, and wellbeing retreat, and responsible technology deployment remain nearly absent in the year AI saturates the agenda. Strip the language, and they are, in substance, technology trends in a human-capital vocabulary.

The employee — their health, pay, security — has left the page. That should trouble a profession whose first name is Human, because these unfashionable disciplines are where human risk concentrates. The one in nine from Part 1 (the workers who need reskilling and never receive it), the wellbeing line that went dark: not abstractions, but where private strain becomes organisational cost. Disengaged, exhausted and financially stressed people make more errors and leave faster; in safety- and conduct-sensitive work, that strain becomes operational and regulatory exposure. The reports cause none of this — they describe a conversation. The danger is in the reading: a leader who treats the agenda’s silence as guidance inherits its blind spot and shuts down the early-warning system for the hard risks, in the year those risks are climbing. That is the difference between studying the map and walking the territory.

In the publishers’ defence

The strongest reply is one the publishers could fairly make: our job is to surface what is new, not to be your strategy checklist. True — and exactly the point. A trend report is a novelty tracker; it reports what changed, not what matters. Wellbeing did not stop mattering in 2026 — it stopped being new. The error is the reader’s, for mistaking a novelty tracker for a strategy input, and the industry’s, for profiting from being read that way.

Our own glasshouse

Full disclosure, before anyone offers it for us. This series is published by a firm with an agenda: we sell people strategy work, and these articles are structurally the very thing we have just dissected. We cannot offer purity. We can offer verifiability — an open dataset behind every number, a method anyone can check. This is our reading of the evidence; judge it by whether it survives inspection. That is the standard we ask you to apply to everyone, ourselves included. And the same scepticism applies to our own numbers: this is a hand-coded reading of 64 trends across ten publishers, the theme mapping and the AI flags are judgements, the earliest year is a provisional baseline, and the publisher roster shifts between editions. Treat the precise percentages as indicative, not exact — the argument rests on the direction of travel and on what the agenda displaced, neither of which turns on a decimal point. See the full analysis for yourself in our 2026 HR Trends & Priorities heatmap.

From compass to map

None of this argues for ignoring the reports. I have read them every year since 2006. It argues for demoting them from compass to map. A map of the conversation tells you what your board has read, what your peers will benchmark, and where the vendor pressure will come from. But a compass can only be calibrated from inside: your workforce, your tensions, your risks — including the human ones the conversation has stopped naming.

The trends tell you what everyone is talking about. Your evidence tells you what is true. Build on the second; quote the first. How to do that — and how to keep score of everyone, ourselves included — is Part 3, The Filter.

Mini-lexicon

The consensus mirror — a field that, reading the same reports, maps the conversation about work rather than the workplace itself — and increasingly sees only its own reflection.

Management fad — a still-important theme that drops out of the rankings because the buzz moved on, not because the problem was solved (Abrahamson).

Displacement — the tell of a fad: a real shift adds to the agenda; a fad substitutes, pushing prior themes off the page.

Absorption — the defence that a dropped theme left because it was solved or embedded — refuted here by evidence that the underlying problems are worsening.

Novelty tracker — what a trend report is built to be: it reports what changed, not what matters.

Map and territory — the reports are a map of the conversation; mistaking them for the territory — the workplace itself — is the core error (Korzybski).

The full traceability workbook — per-trend mapping, AI tags, evidence classifications and borderline notes — is open to inspection on request.

Abrahamson, E. (1991). “Managerial Fads and Fashions.” Academy of Management Review, 16(3): 586–612; and (1996). “Management Fashion.” 21(1): 254–285.

Briner, R. (2024). Evidence-Based HR: A New Paradigm. Corporate Research Forum.

Deloitte, Gen Z and Millennial Survey 2025

Gallup, State of the Global Workplace 2026.

Miller, D. and Hartwick, J. (2002). “Spotting Management Fads.” Harvard Business Review, 80(10).

PwC, Employee Financial Wellness Survey 2026

World Economic Forum, Good Work Framework and Future of Jobs Report 2025.