Do HR teams score their own engagement higher than everyone else?

There is an old assumption in people teams that HR is the disengaged caretaker: the function that builds engagement programs for everyone else while quietly running on empty. It is a vivid story. The phrase even has a name borrowed from the proverb, "the cobbler's children have no shoes." But it is a story, not a measurement.

We measured it. Using 12 months of eNPS responses from companies on the Happily platform, we identified HR teams in each organization, computed their eNPS, and compared it to the eNPS of everyone else in the same company. The question was simple: do HR teams score their own engagement differently from the people they survey?

Across 21 organizations that met the sample thresholds, the answer is yes, and the direction is the opposite of the caretaker story. HR teams report eNPS scores that average 8.6 points higher than their non-HR colleagues. The median gap is +10.0. But the average hides a split worth taking seriously: in 8 of those 21 companies, HR scored below the rest of the organization.

How we measured it

eNPS, the employee net promoter score, is the standard one-question engagement metric: the share of promoters (scoring 9 or 10 on a 0-10 likelihood-to-recommend scale) minus the share of detractors (0 to 6). It runs from −100 to +100.

For each company we identified HR teams by matching team names against a keyword list in both English and Thai, covering HR, Human Resources, People Ops, People & Culture, Talent Acquisition, Recruitment, L&D, HRBP, and the Thai equivalents. Team names containing "People" but referring to non-HR functions, such as "People & Tech", were excluded as known false positives. We then computed HR eNPS from members of those teams and non-HR eNPS from everyone else in the same company.

To keep the comparison stable we required at least 3 HR responses and 10 non-HR responses per company. Twenty-one companies cleared that bar. The gap reported for each company is HR eNPS minus non-HR eNPS, so a positive number means HR scored higher.

One property of eNPS matters before reading any of the numbers below. The metric is volatile in small samples. HR teams in this sample average 2 to 5 respondents. With a 5-person team, one person moving from a detractor to a promoter swings the team's eNPS by roughly 40 points. That volatility is visible in the results, and it shapes how much weight any single company's gap can carry.

Finding 1 — HR scores higher, but the spread is wide

Across the 21 companies, the mean eNPS gap is +8.6 points and the median is +10.0. Both point the same way: HR teams, on balance, report engagement above their colleagues. The effect is consistent in direction but modest in size, with a Cohen's d of 0.283, which is a small effect by conventional benchmarks.

The standard deviation of 30.5 points is the number to sit with. It is more than three times the mean. A spread that wide tells you the +8.6 average is not a tight, reliable pattern that every company exhibits. It is a center of gravity around which companies scatter widely, with the first quartile sitting in negative territory at −11.2.

Why might HR tend to score higher? The data cannot prove a mechanism, but four are plausible. People drawn to HR roles may carry higher organizational attachment to begin with, a selection effect. HR teams often sit close to leadership, and proximity to decision-makers is a known driver of engagement. HR are the cultural custodians who design the engagement programs and may internalize the values they promote. And HR teams are small and cohesive, and small close-knit teams tend to score well on engagement measures, an effect the eNPS formula amplifies in small samples.

Finding 2 — The gap is largest in small companies

Splitting the 21 companies by headcount shows the HR-above-average pattern is strongest in the smallest organizations and narrows as companies grow.

Two readings sit behind this. One is genuine: small companies put HR physically closer to leadership and to the whole workforce, which amplifies the proximity and cohesion effects. The other is mechanical: in a small company the HR team is a larger fraction of a smaller whole, and small-sample volatility is highest, so the +14.1 in the small segment also carries the widest standard deviation at 37.3. With only 3 large companies in the sample, the large-company figure in particular should be read as directional, not settled.

Finding 3 — The bellwether question stays open

The more interesting hypothesis behind this study was a lead-lag one. If HR teams sit closest to the culture, do changes in HR eNPS arrive before changes in the rest of the company? An HR team that is the first to feel a downturn would be a useful early-warning signal, a bellwether.

We could not test it. Building a monthly HR and non-HR eNPS time series per company, with a minimum of 2 HR responses and 5 non-HR responses per month and at least 6 valid months, left exactly one qualifying company. HR teams are too small and eNPS surveys too infrequent to produce reliable month-by-month HR scores. Most monthly bins held 0 to 2 HR responses.

The honest conclusion is that the data granularity required, monthly HR-specific eNPS with enough respondents, exceeds what most organizations produce. A future study could use quarterly windows instead of monthly, or pool several people-facing functions together (HR plus Admin plus L&D) to reach a testable sample. Until then, the bellwether claim is a hypothesis, not a result.

What this means

The data tells a clear story about levels and an unfinished story about trajectories. On levels: HR teams are, on balance, not the disengaged caretakers the stereotype assumes. They score moderately higher than their organizations. But the 38% where HR scores lower is where the attention belongs, because that is where the cobbler's-children problem is real.

Limitations

This study measures a 12-month snapshot of eNPS levels, not their causes and not their movement over time. Several caveats shape how far it travels.

• Small HR team sizes. The average HR team here has 2 to 5 respondents. eNPS is designed for samples of 50 or more. Small-sample volatility inflates the spread and makes any single company's gap unstable.
• Team classification accuracy. Keyword matching can miss creatively named HR teams or misclassify HR-adjacent ones. Known false positives such as "People & Tech" were excluded, but edge cases remain.
• Selection bias. Every company in the sample uses the Happily platform and is already investing in engagement measurement. HR teams here may be more engaged than HR teams at organizations that do not measure engagement at all.
• Cross-sectional, not longitudinal. This is one 12-month window. We cannot tell whether HR consistently scores higher or whether the gap fluctuates within companies.
• Correlation, not causation. The gap may reflect team size, selection effects, or response bias rather than a genuine engagement difference.
• Bellwether analysis underpowered. Only one company qualified for time-series testing. The lead-lag hypothesis remains untested, not disproven.
• Sample of 21. Adequate for detecting an average effect, but it limits subgroup analysis. The size segments hold only 3 to 10 companies each.