Where engagement variance actually lives

You have probably seen the chart. Some version of it has been on your CHRO's deck at the offsite, or in a Gallup webinar, or buried in a footnote of last year's engagement vendor's pitch:

Managers account for at least 70% of the variance in employee engagement scores across business units.

The figure comes from Gallup's State of the American Manager (2015), a variance-decomposition analysis on Q12 data covering 27 million employees and 2.5 million work units over two decades. You can argue the methodology, you can argue the precise number — other analyses put manager effects somewhat lower, in the same direction. But you cannot really argue with the shape of the claim, because you've watched it happen on your own teams.

Two teams in the same function, same compensation band, same workload. One has thriving people, low attrition, strong promotion pipeline. The other has steady turnover and middling reviews. The manager is the difference. Everyone in HR knows this. The question is what to do about it.

What every People program has gotten wrong

Most engagement programs in the field today were designed for a different problem. They were designed for the problem of "we don't know how our employees feel." That's the problem the survey solves.

The actual problem in 2026 is "we know roughly how our employees feel, and the variance is in the manager, and we don't have a lever to pull on that variance at scale." It's a different problem. The survey doesn't solve it.

Worse, most of the second-order programs make the situation more brittle:

• Manager training: useful at the margins, but you can't train a great manager out of a person who isn't one. Variance survives the workshop.
• Calibrations: address performance review consistency, not the daily relationship that drives engagement.
• OKR rollouts: assume the manager is competent at goal-setting; when they aren't, the OKR is the wrong layer to intervene at.
• More frequent pulse surveys: re-measure the same thing at higher resolution. The variance is still in the manager. Now you have monthly data confirming it.

None of these are bad. All of them target levers other than the one that explains 70% of the outcome. That's the pattern People functions have been stuck in for a decade.

The lever that's actually moveable

You cannot make a manager good by company-wide intervention. What you can do is improve the input to the conversation that drives the variance — the 1:1.

The 1:1 is the unit of management where engagement is won or lost. It's also the unit that's least defended by HR programs, because it's not a program — it's a recurring meeting between two people. The shape of that meeting is mostly determined by what the employee walks in with.

When the employee walks in with nothing — no list, no evidence, no asks — the meeting defaults to whatever the manager wants. If the manager is great, the conversation goes well. If the manager is mediocre, it doesn't. That's exactly the variance the Gallup data identifies.

When the employee walks in with a prepared 1:1 agenda — specific wins they want noted, specific challenges they want help on, specific open loops from last conversation — the meeting changes shape. The manager has to engage with what the employee brought. The conversation tilts toward the employee's agenda, which is the agenda their engagement actually responds to.

This is the leverage point. You can't fix every manager. You can equip every employee to make the 1:1 work for them regardless.

The mechanism in plain language

Kahn's 1990 work on personal engagement identified three conditions: meaningfulness (do I think this matters?), psychological safety (can I take an interpersonal risk?), and availability (do I have the resources to engage?). The 1:1 is one of the few corporate touchpoints that operates on all three. It's where the employee gets to articulate what matters, take the small risk of voicing a concern, and ask for the resources they need.

Self-Determination Theory adds the autonomy lens: when the employee walks in with their own list, they exercise autonomy over the conversation, not just within it. That single shift — owning the agenda — does more for the felt sense of autonomy than any org-wide policy change.

The Job Demands–Resources model points the same direction. Feedback and manager support are "job resources." Employees who consistently receive both predict higher engagement than those who don't. A recurring 1:1 with a structured agenda is, structurally, a feedback-and-support resource the employee can rely on.

None of this is novel. What's novel is the willingness to design the program around it instead of around the survey.

What this looks like, structurally

A People function that has internalized this stops trying to push engagement up from the top and starts removing the conditions that suppress it from the bottom. The interventions look like:

• Equip the employee, not the manager. Give every employee a private workspace where they capture wins, log challenges, draft weekly reflections, and prepare 1:1 agendas for important conversations. The employee owns the workspace. The company sees seat metadata and aggregate adoption only.
• Stop trying to read the conversation directly. Whatever HR thinks they would learn from listening to the 1:1, they'd learn faster and more honestly from watching the behavior pattern — is the employee creating 1:1 agendas before conversations? Are open loops getting closed? Are weekly reflections happening?
• Measure the cadence, not the content. Aggregate counts at the team level (gated below 5 active members to prevent inferring about individuals) tell you whether the program is actually being used. That's the only engagement signal you need at the org level, and it's the only one that doesn't degrade the system by being measured.
• Accept that the manager variance is not yours to eliminate. It's the manager's to address, with their own manager and their own coaching. Your job is to ensure the employee isn't at the mercy of it.

The honest separation between mechanism and product

The Gallup variance estimate, Kahn's engagement conditions, JD-R, and SDT collectively underwrite a clear theory of change: the 1:1 conversation is the highest-leverage intervention point, and improving the employee's input to that conversation produces better conversations across the variance distribution.

What the literature does not yet prove is that this specific product, in your specific organization, achieves that improvement at the magnitude the theory implies. That's what the pilot measures. We are not telling you that Nela will move your engagement scores by N points. We are telling you that the mechanism is well-grounded, the structural design pulls the lever the literature identifies, and a six-month pilot will tell you whether the bet pays in your environment.

This is the only honest pitch for a product that targets the highest-variance point in the engagement system without trying to retrain every manager. It's also the only pitch that survives a research-literate buyer asking "show me your evidence" — because the evidence is for the mechanism, not the product, and we say that out loud.

How Nela Helps

Use Nela to log your wins, track your challenges, and build a private 1:1 agenda from your own evidence for your next conversation. Your data is owner-only at the database — enforced by Postgres Row-Level Security, not just hidden in the UI — and only you can read it back through the app. Request pilot access.

Further reading

• Gallup. (2015). State of the American Manager.
• Kahn, W. A. (1990). "Psychological Conditions of Personal Engagement and Disengagement at Work." Academy of Management Journal.
• Bakker, A. B., & Demerouti, E. (2017). "Job Demands–Resources Theory: Taking Stock and Looking Forward." Journal of Occupational Health Psychology.
• Deci, E. L., & Ryan, R. M. (2000). "Self-Determination Theory and the Facilitation of Intrinsic Motivation, Social Development, and Well-Being." American Psychologist.