Engagement is not a wellbeing programme you delegate to HR and forget. It is an operating input that shows up in your numbers: how many people quit this quarter, how much rework your teams generate, how quickly a new hire becomes productive, whether customers notice the difference. Treat it the way you treat throughput or unit cost, and it becomes something you can actually move.

The mistake most operations leaders make is buying a survey tool, running one annual questionnaire, and calling that a strategy. The survey tells you the temperature; it does not change it. What changes it is what your managers do on Tuesday afternoon — whether the person who flagged a broken process last month got an answer, whether recognition is specific or generic, whether a promising engineer can see a path that keeps them another three years.

This guide is written for the COO who owns the operating result. It covers what to measure and how often, what strong versus weak engagement looks like at the manager level, a table you can steal for your own scorecard, and a 90-day sequence to go from baseline to visible improvement.

Why the COO owns engagement, not just HR

HR owns the tooling, the policy, and the survey administration. The COO owns the consequence. Turnover shows up as vacant seats on your operating teams; disengagement shows up as slipped deadlines and defects. If you run operational excellence programmes, engagement is one of the raw materials — you cannot run stable, low-variation processes with a workforce that is halfway out the door.

The practical split is simple. HR designs the recognition platform; you decide that every department head reports an engagement metric alongside their output metric. HR runs the pulse survey; you make sure the results get discussed in the same operating review where you discuss cost and delivery. When engagement lives on the operations dashboard rather than in a separate HR deck, it gets managed.

Strong looks like: engagement scores appear in monthly operating reviews next to delivery and cost, and a declining team score triggers the same "what's the plan?" conversation a delivery miss would. Weak looks like: engagement is a once-a-year all-hands slide, owned by nobody, actioned by nobody, forgotten by week two.

What to measure — and what each number actually tells you

Pick a small set of indicators you will look at consistently rather than a large set you check once. The goal is a leading signal you can act on, backed by lagging outcomes that confirm whether your actions worked.

Two of these deserve emphasis. Separate regretted turnover from total turnover — losing a poor performer is not the same problem as losing your best process engineer, and a blended rate hides the one that matters. And track manager 1:1 completion: it is the cheapest early warning you have, because when one-on-ones quietly stop on a team, engagement is usually about to slide, and you can act before the survey confirms it. Strong looks like: you separate regretted from non-regretted exits, and you can name the top three reasons your best people gave in exit conversations this quarter. Weak looks like: a single turnover percentage on a slide with no distinction between who left, why, or whether you wanted them to stay.

The manager is the system — build the frontline habit

Most engagement variance in a company traces to one factor: the direct manager. The same job, pay, and office produce an engaged team under one manager and a demoralised one under another. So the highest-leverage thing a COO can do is make good management a defined, supported, measured practice rather than a personality trait you got lucky with.

Start with a non-negotiable rhythm: a weekly or fortnightly 1:1 that is the employee's meeting — their agenda, their blockers, their growth — not a status update the manager could get from a dashboard. Add specific recognition, because "your fix on the returns workflow cut our processing errors and saved support a pile of tickets" lands where "great job team" evaporates. And give managers a simple escalation path so that when someone raises a real problem, it goes somewhere and comes back with an answer instead of dying in a notebook.

This is where engagement connects to talent development and building high-performance teams. A manager who runs real 1:1s, recognises specifically, and unblocks quickly is running engagement and development at once — they are the same behaviours.

Strong looks like: every people manager has a scheduled recurring 1:1 that actually happens, and recognition in your channels references specific work and its impact. Weak looks like: 1:1s get cancelled whenever the week is busy, and recognition is a monthly generic "shout-out" nobody remembers by lunch.

Close the loop — the single most-broken step

Engagement dies the moment employees learn that speaking up changes nothing. You run a survey, people tell you workload is crushing them, and then nothing visibly happens. The next survey gets fewer responses and worse scores, because you have taught people that participation is pointless. This is the failure mode that turns a well-intentioned programme into a net negative.

Closing the loop means three things: report back what you heard, even the uncomfortable parts; name what you will act on, what you will not, and why; then show the change. You do not have to fix everything. A leader who says "the deploy process is painful, so we are rebuilding it this quarter — the coffee complaint is real but not a priority right now" earns more trust than one who promises everything and delivers vaguely. Transparency about trade-offs reads as respect. This connects directly to stakeholder engagement — your employees are stakeholders, and the same rules of honest, two-way communication apply.

Strong looks like: within a few weeks of every survey, employees see a "here's what you said, here's what we're doing, here's what we're not and why" communication, followed by visible action. Weak looks like: survey results go into a report that only leadership sees, and employees never hear an outcome — so response rates fall every cycle.

Engagement in remote and hybrid teams

Distributed work removes the ambient signals a manager used to rely on — you cannot read the room when there is no room. Disengagement hides longer and surfaces later, often as a resignation that "came out of nowhere." The fix is not more mandatory video calls; it is deliberately designing the connection that used to happen by accident.

That means written clarity so remote staff are not guessing at decisions, protected 1:1 time that does not get sacrificed to timezone juggling, and a few genuine team touchpoints that are not status meetings. It also means watching different signals: a normally-active person going quiet in channels, declining meeting participation, or a drop in the small collaborative interactions that show someone still feels part of the team. The tools and practices for leading remote teams are a discipline of their own — the engagement question is whether managers spot fade-out early rather than discover it at the exit interview.

Strong looks like: remote managers actively track participation and check in on people who go quiet, and hybrid schedules are designed for collaboration rather than attendance theatre. Weak looks like: remote staff are managed purely on output with no relational contact, and the first sign of disengagement is a resignation letter.

Fund it and prove it — engagement as an investment

You will need a budget line, and you will need to defend it. Direct costs are visible: survey and recognition platforms, development programmes, the time managers spend on 1:1s. The return is real but shows up as costs you avoid — lower regretted turnover (each senior departure carries recruiting, lost productivity, and ramp cost for the replacement), faster time-to-productivity, and fewer quality escapes from teams that care.

Build the case the way you would any operations investment: baseline the current state, project the cost of the status quo, and track the metrics that would move. If regretted turnover is high in a critical role, model what a modest reduction is worth in avoided replacement cost, then measure whether your interventions bend the curve. Tying engagement work to COO success metrics is how you keep it funded when budgets tighten — a programme with no numbers behind it is the first thing cut.

Strong looks like: engagement spend is justified against modelled avoided-turnover and productivity cost, and reviewed like any other operating investment. Weak looks like: engagement is funded on faith and defunded the moment finance asks what it returns.

Key takeaways

• Engagement is an operating input — put it on the operations dashboard next to delivery and cost, not in a separate annual HR deck.
• Measure a small, consistent set: eNPS and monthly pulse as leading signals; regretted turnover, time-to-productivity, and internal mobility as lagging confirmation.
• The direct manager is the system. Non-negotiable 1:1s, specific recognition, and fast unblocking drive most of the variance.
• Closing the loop is the step that fails most often — report back, act, and be honest about trade-offs, or participation collapses.
• Remote and hybrid teams need deliberately designed connection and early-warning signals, because disengagement hides longer.
• Fund it as an investment and defend it with avoided-cost numbers, especially regretted-turnover cost.

Frequently asked questions

Where should a COO start if there's no engagement programme at all? Start with a baseline, not a big platform purchase. Run one short pulse survey (5–8 questions) and pull your regretted turnover for the last year. That gives you a starting point and one or two obvious problems. Then fix the manager 1:1 habit first — it is the cheapest, highest-leverage change you can make. How is engagement different from employee satisfaction? Satisfaction asks whether people are content — comfortable pay, decent perks, no major complaints. Engagement asks whether they are committed — do they care about the outcome, put in discretionary effort, and intend to stay. You can have satisfied but disengaged employees who coast, and highly engaged people who are frustrated because they care. As a COO, engagement is the one that moves your operating numbers. How often should we survey without causing fatigue? A short monthly or six-weekly pulse of a handful of questions, plus one deeper survey once or twice a year, is a workable rhythm. Fatigue comes less from frequency than from silence — people stop responding when nothing ever changes after they answer. Ask less often only if you cannot yet act on what you hear; otherwise close the loop and frequency stops being the problem. What's the fastest way to lose the trust I'm trying to build? Ask for feedback and then do nothing visible with it. A survey that produces no reported-back outcome teaches people that speaking up is pointless, and the next cycle gets fewer responses and worse scores. If you are not prepared to report back and act on at least some of what you hear, do not run the survey yet. How do I get middle managers to actually own this? Make it part of their job with a metric attached, not an extra ask. Put a team engagement indicator and 1:1 completion next to their delivery metrics in the same review, give them the training and a simple escalation path so the role is doable, and recognise the managers who do it well. Managers deliver what gets inspected; if engagement never appears in their operating review, it stays optional.