The Engagement Cascade: how far each management layer reaches

When a leader becomes more engaged, does that energy stop with their own team, or continue through the organization? Across 72 organizations, the clearest answer is: most leaders influence the people who report directly to them. Only the top leader reliably reaches the next team down as well.

Think of each manager-to-employee relationship as one reporting link. A top leader's direct reports are one link away. The people reporting to those managers are two links away. By measuring each distance separately, we can see where the relationship remains strong and where it breaks.

We followed 7,176 monthly observations of leaders and studied 4,155 month-by-month handoffs between a manager and another manager reporting to them. The strongest relationship starts at the top. The clearest breakdown happens between layer 1 and layer 2, where the next manager still influences their own team but is not reliably activated by the leader above.

What we measured

In this study, engagement means observable behavioral engagement: how consistently people take part in everyday workplace interactions such as check-ins, open feedback, and recognition. These signals show what people did, not whether they reported feeling happy, satisfied, healthy, or committed. The interactions were observed through Happily, but the measure represents the workplace behaviors rather than use of a particular product. It is also separate from Happily's DEBI measure.

We call the highest leader in each reporting tree layer 0, or the top. Their direct reports are layer 1, the next level is layer 2, and so on. A leader's direct team means immediate reports. Their full branch means everyone below them in the reporting tree.

We compared each leader with their own usual engagement level. For layers 1 and below, we also compared managers experiencing the same company month, which helps remove shared events such as a company campaign. Every main result kept the same direction when we removed each organization one at a time.

Finding 1: Most leaders reach one team; the top reaches two

A broad branch average can hide where the relationship stops because it mixes direct reports with everyone below them. We therefore measured each reporting distance separately. We count engagement as having traveled only when the evidence stays reliable at every link along the way.

When top-leader engagement is one typical step above its usual level, direct-team engagement is 3.20 points higher and engagement in the next team down is 2.08 points higher. At the third link, the relationship becomes too weak to distinguish from normal variation. For leaders starting at layers 1 and 2, the relationship also becomes unclear after the direct team.

Finding 2: The next manager determines whether engagement continues

To study the handoff, we isolated 677 relationships where one manager directly supervises another manager. We call the second manager the relay because engagement must pass through them to reach the team below. Neither manager is counted in that team's result.

Top engagement carries through the next manager

When the top leader is more engaged than usual, the manager below them also tends to be more engaged (β = 0.22), as does that manager's team (β = 0.30). Once we account for the next manager's engagement, the top-to-team relationship becomes 22% smaller. That pattern is consistent with the manager carrying part of the signal, although it does not prove causation.

The layer 1-to-layer 2 handoff is the clearest break

When a layer 1 leader becomes more engaged, we do not see a reliable change in the layer 2 manager or that manager's team. Yet the layer 2 manager's own engagement remains strongly connected to their team (β = 0.20). The local manager still matters; the handoff from the leader above is what fails.

The top-leader relationship is stronger when the next manager is usually engaged. By contrast, wider teams and deeper reporting structures do not consistently weaken the relationship. The clearest obstacles in this study are therefore a manager-to-manager handoff that fails and a signal that fades quickly, not organization size or depth by itself.

Finding 3: Broader branch averages still highlight the top and layer 2

We also looked at each leader's full branch: everyone below them combined. This is useful for understanding the overall area of the organization connected to a leader, but it does not tell us how many management links the relationship crossed. A beta of 0.20 means that when manager engagement rises by one typical step, downstream engagement rises by one-fifth of a typical step.

* Statistically reliable after correcting for repeated comparisons. † Early evidence only, because fewer than 10 organizations contribute.

The top aligns the widest system

Top-leader engagement has the largest same-month relationship: β = 0.38 for both the direct team and the full branch. The point increase is larger for the direct team because an organization-wide average changes more slowly than a small team's average.

The top result needs extra caution. Most companies have only one top leader, so we cannot compare them with another top leader experiencing the same company events. A company-wide campaign could move the leader and organization together. The next-month analysis below tests whether any relationship remains after the immediate month.

Layer 2 matters locally, but does not automatically move the next team

Layer 2 is the only level below the top with reliable same-month relationships for both its direct team and its full branch. Its direct team is 1.37 points higher for a typical increase in manager engagement, while the full branch is 0.87 points higher. But the exact-distance analysis shows that managers below layer 2 still need to be engaged directly.

Layers 1 and 3 also matter to their direct teams. Their wider-branch results are not clear enough to rule out normal variation, so their dependable reach appears local.

Finding 4: Most of the relationship fades by the next month

Movement in the same month tells us that managers and teams rise and fall together. It does not tell us whether the manager moved first. To test persistence, we followed the same employees into the next month while accounting for their current engagement.

Only the direct-team relationships at the top and layer 2 remain statistically reliable in the next month. Their branch estimates stay positive, but the evidence is not strong enough to call them dependable after correcting for repeated tests. Layers 1 and 3 shrink almost completely to zero.

The top-level handoff is the exception. Top engagement remains connected to both the next manager (β = 0.16) and that manager's team (β = 0.23) in the following month. We do not see a similarly reliable manager-to-manager handoff starting at layers 1 or 2.

Finding 5: Layer 2 shows the clearest six-month branch relationship

We also compared manager engagement in the second half of 2025 with employee engagement in the first half of 2026, while accounting for those employees' earlier engagement. Fewer leaders qualify for this longer test, but it shows whether any relationship lasts across a six-month operating cycle.

No other six-month result is statistically reliable. Too few top leaders qualified for a dependable top-level estimate. Because the branch measure includes direct reports, the layer 2 result shows lasting alignment across the branch as a whole. It does not prove that engagement crossed another manager.

Finding 6: The result is more than active managers versus absent managers

Managers recorded no participation in about one quarter of eligible months. That matters, but it could make the result look stronger if the analysis merely separated managers who showed up from those who did not. We therefore repeated the analysis using only months when managers participated at least once.

The direct-team relationship remains reliable at layers 0, 1, 2, and 3. The full-branch relationship remains reliable at the top and layer 2. Showing up matters, but differences in how engaged managers are also matter.

What leaders should do

The practical conclusion is simple: start at the top for broad movement, but manage every handoff. Do not assume that engagement will pass through an intermediate manager without direct reinforcement.

Limitations

• This measures behavior, not feelings. A higher score means more consistent participation in check-ins, open feedback, and recognition. These signals were observed through Happily, but they represent workplace interactions rather than product usage. The score does not mean employees are happier, healthier, or more satisfied.
• The study shows relationships, not proof of cause. Engaged teams may encourage managers to participate, and shared events can move both at once. Looking at later months improves the timing evidence but does not create an experiment.
• The top-level comparison is less controlled. Most companies have only one top leader, so there is no same-company top-level peer for comparison.
• Reporting lines come from June 2026. We cannot see every team change or reorganization that happened during the study year.
• Some organizations have more than one top reporting tree. In those cases, we measure the leader's branch rather than the entire legal organization.
• Layer 4+ includes too few organizations. Those estimates are early evidence, not a dependable conclusion.
• Deeper distances include fewer organizations. We only call engagement "traveled" when every link along the path is supported.
• The relay analysis does not prove the mechanism. An unmeasured event inside a branch could move the senior manager, the next manager, and the team together.