Can you measure management quality without an engagement survey?
Leadership · 9 min read
Can you measure management quality without an engagement survey?
Most organizations measure management once or twice a year, through a survey, and then act on the results a quarter later. Imagine running finance that way.
The problem isn’t that surveys are bad — it’s that they’re the wrong instrument for anything you want to act on quickly. They’re slow, they’re seasonal, they’re gameable, and they’re anonymous in a way that tells you a score without telling you where to look.
Three things you can measure continuously
Coverage. What proportion of people have a live one-on-one cadence with their manager? This is embarrassingly simple and almost nobody knows their number. Coverage below 80% means a meaningful slice of your organization has no regular channel to raise anything.
Cadence compliance. Of the 1:1s that should have happened this month, how many did? Measured per manager, per department, over time. The absolute number matters less than the trend and the outliers.
Follow-through. Of the commitments made in those conversations, how many were closed on time? This is the closest available proxy for whether the conversations have consequences.
All three are structural. None requires reading anyone’s private conversation, and none needs anyone to fill in a form.
What they tell you that a survey can’t
They’re current. A cadence drop shows up within a fortnight of it starting, not at the next survey cycle.
They’re specific. “Engineering’s compliance fell to 54% this month, concentrated in two managers” is actionable. “Manager effectiveness scored 3.4” is not.
They’re hard to game in a harmful direction. Gaming cadence compliance means… holding your one-on-ones. That’s a metric whose failure mode is the desired behaviour.
They don’t cost the organization anything to collect. No survey fatigue, no comms campaign, no three weeks of nagging people to respond.
Where these metrics mislead
Worth being honest about the limits.
- Frequency isn’t quality. A manager can hold every 1:1 and run each one badly. Cadence measures whether the channel exists, not whether the conversation is good.
- Targets corrupt. The moment compliance becomes a number in someone’s performance review, you’ll get compliance and lose the point. Use it to find where to look, never as a target to hit.
- Context beats comparison. A support team’s healthy pattern looks nothing like a research team’s. Compare a team to its own history, not to another department.
How to use them properly
Treat them as a smoke alarm, not a scorecard. When a team’s cadence drops, the response is a conversation with the manager about what’s going on — usually span, workload or a reorg — not a note in their file. Publish the numbers openly, including your own. And keep the engagement survey: it tells you about sentiment, which these metrics can’t. It just shouldn’t be the only instrument on the dashboard.
These three measures are exactly what Cadentia AI reports — coverage, cadence compliance by department, and follow-through on commitments — without any survey and without anyone reading a private conversation. If you’d like to see what your own numbers look like, four seats are free.
Keep reading
The 1:1 questions that actually get an honest answer
Most one-on-one question lists read like an interview script. Seven questions that reliably surface a real problem, four that never do, and why the difference is structural.
How often should you hold one-on-ones?
Weekly, biweekly or monthly is not a matter of taste. A simple rule based on tenure, autonomy and how fast the work changes — plus what to do when the calendar fights back.
Skip-level visibility without turning into a surveillance company
Executives need to know whether conversations are happening. There’s a version of that which destroys trust and a version that builds it — the difference is entirely in what you choose to measure.
Reading about 1:1s is not the same as holding them.
Four seats free with unlimited one-on-ones. Put this idea to work this week.
