This is a leadership problem before it's a frontline one
When resolution times creep up or satisfaction scores slide, the instinct in most organisations is to look at the frontline first: retrain the team, tighten the script, performance-manage the slowest agents. For an executive watching the number move on a quarterly business review, that instinct is understandable. It is also usually aimed at the wrong altitude.
Performance drift is almost never where it appears to start. By the time it shows up in a board pack or a quarterly review, it has been building for months in places nobody at the top was watching: how standards get explained to new starters, whether coaching still happens on a schedule or only when there is time, whether supervisors have any capacity left once queues and admin are covered.
Oliver has watched this pattern repeat across contact centres, sales floors and service teams for 25 years, and it is almost always a leadership and systems story before it is an individual one. Agents do what the environment around them rewards and what they were shown to do. When that environment drifts, so does the performance sitting inside it, and the fixes available to a supervisor working alone are limited if the real causes sit above their authority: ratios, budget for coaching time, how the role was designed in the first place.
That is the case for keeping drift on the leadership agenda rather than delegating it entirely. What follows sets out where it starts, what it costs before anyone notices officially, and the decisions in front of you once it is confirmed.
Where drift starts
Drift is rarely a single event. It is the accumulation of small failures that each look harmless on their own, and none of them will appear as a line on a budget report until they already have.
Standards nobody wrote down
Ask five supervisors in the same centre to describe what a good call sounds like and you will often get five different answers. If the standard was never written down clearly, or was written once at launch and never revisited, new starters learn it from whoever trained them, who learned it from whoever trained them before that. The standard does not so much erode as quietly mutate, one handover at a time, and nobody signs off on the drift because nobody decided it.
Coaching that stopped happening
Most coaching models are designed when a team is small: one supervisor, six or eight agents, room to sit in on calls every week. As the team grows, supervisor headcount rarely keeps pace, so the ratio climbs. Coaching has no fixed deadline the way a rota or a report does, so it is usually the first thing to slip, and it slips quietly, because nothing breaks the day it stops.
Supervisors carrying a job they were never trained for
Many organisations promote the strongest agent into the supervisor role, hand them a wider span of control, and add a queue to manage on top of it. There is rarely a deliberate handover of what the supervisor job requires, as distinct from what made them good on the phones. That is a different skill set, and without support for it, new supervisors default to firefighting: escalations, absence cover, reporting. Coaching becomes the first casualty because nobody is holding them to it.
Knowledge that has not kept pace
Products change, policies update, and systems get patched, but the knowledge base rarely keeps pace with any of it. Agents start improvising to fill the gaps.
Improvisation becomes habit. Habit becomes the unofficial standard, and the standard has moved without anyone in the building deciding it should.
What drift costs before it shows up as churn
None of the four patterns above look urgent from the executive floor. That is precisely the danger: drift stays invisible on the metrics most leadership teams are shown, right up until it is expensive.
The early cost is capacity. Every contact logged as resolved but not genuinely fixed becomes a repeat contact later, and each one consumes time that should go toward new volume or the next customer. Instead, that time gets spent reworking the last one. Every supervisor who has stopped coaching because the queue will not let them is spending their week managing symptoms instead of preventing the next one.
That capacity does not show up as a cost line anywhere. It shows up as a headcount request that looks reasonable in isolation, and is paying for drift nobody named.
The later cost reaches your desk directly. Escalations that used to stop at a team leader now arrive in your inbox, because the layer that should have absorbed them has been quietly hollowed out. Renewal conversations get harder when an account team cannot explain why service dipped, because nobody had been tracking the real cause, only the speed metrics that stayed flat. Reputational exposure builds the same way: one bad review rarely sinks an organisation, but a pattern of them, each individually explainable, is much harder to defend in a board discussion than a single clear incident with an obvious cause.
The honest reason drift is expensive is timing, not scale. Caught early, it is a coaching and standards conversation. Caught late, it is a restructuring exercise, a remediation program, or a client retention exercise: the same underlying problem, at a much higher cost to fix, because it has had months to compound.
The metrics that hide the problem
Most organisations can see speed in real time: average handle time, occupancy, queue length. They sit on a live dashboard, refreshed every few minutes, and they are easy to report upward because they are easy to see. That is exactly why they dominate the reviews that reach executive level, whether or not they are the numbers that matter.
Resolution quality is a different matter, and the gap between the two is where drift hides. First contact resolution is often presented as a single headline number in a leadership pack, but the definition underneath it matters more than the number itself: is a contact marked resolved because the issue was fixed, or because there was no repeat contact within a set number of days? Those are not the same thing, and a team under pressure to hit that number will find ways to close a case without necessarily closing the problem.
A team can hit every speed target on the board while the underlying service quietly gets worse: more repeat contacts, more escalations, more agents working around a policy gap rather than through it. None of that shows up in a standard report until customers start leaving, or until someone finally audits a sample of resolved contacts and does not like what they find.
The practical implication for anyone reviewing these numbers at leadership level is simple: ask what is underneath a headline metric before accepting it as evidence of health. "Resolved" is a policy choice as much as a fact, and it is worth knowing which policy your organisation is currently using.
Training, restructuring, or outsourcing: making the right call
Once drift is confirmed, the next decision is where to put the budget. This is where organisations most often reach for the wrong lever, because the three obvious options solve different problems: train harder, restructure, or hand the function to a partner.
Training helps when the honest diagnosis is that people do not know what to do, or have not had the chance to practise a skill they are missing. A team that has never been shown how to de-escalate a frustrated customer, or how to navigate a policy exception without guessing, genuinely improves with structured customer service training that pairs practice with feedback.
Training does not help when the standard itself was never defined clearly enough to train toward, when there is no coaching structure in place to reinforce it once the session ends, or when the process actively punishes the behaviour being taught. Training a team to slow down and build rapport, while their handle time target penalises anyone who takes longer than average, sets them up to fail: the incentive wins every time, and the training gets blamed for not sticking.
Restructuring means redefining the supervisor role, resetting ratios and rebuilding the coaching model. It is the right call when the diagnosis points at the system rather than the skill. That is a harder conversation to schedule than a workshop, because it usually means a headcount or role-design decision that sits above a normal training budget. It is also, per the causes above, the one that most consistently stops drift from returning six months later.
Outsourcing part or all of the function is worth putting on the table when internal capacity genuinely cannot close the gap fast enough, or when building the leadership bench internally would take longer than the business can wait. It is not automatically the right answer: results still depend on the process, the location, the partner and the operating model built around it, and a partner inherits your standards problem if nobody has defined what good looks like before handover. Where it does make sense, outsourced customer service draws on a global network of 43 contact centre partners across the Caribbean, North America, the United Kingdom, South Africa and international markets, with capability matched to what the specific work needs rather than assumed from a location's general reputation.
A short way to sort the three:
- A missing skill. Train.
- A role, a ratio or a standard nobody defined. Restructure before you train.
- Capacity your organisation will not build fast enough on its own. Put outsourcing in the conversation once the diagnosis is done.
Where the fix goes wrong, and where to start instead
Common ways this gets fixed badly
Even once leadership accepts drift is real, a handful of responses reliably make it worse rather than better, and they are common enough to name directly.
Mandating more training without funding the coaching capacity to reinforce it spends budget on a session that decays within weeks, because nothing in the environment changed to hold the new behaviour in place. Chasing speed metrics harder, on the theory that the team simply needs to try more, treats a systems problem as a motivation problem, and usually makes resolution quality worse while the dashboard looks better. Promoting on technical strength alone, without preparing the new supervisor for what the role requires, restarts the exact cycle that produced drift in the first place. Our frontline leadership training exists specifically because that handover is a skill set, not something people absorb by osmosis once they have the title.
The costliest version is outsourcing a standards problem rather than fixing it. Handing a drifting function to a partner without first defining what good looks like does not remove the problem: it just moves the problem somewhere with less visibility and a contract attached, and the partner ends up managing to a standard nobody agreed on.
What to put on the agenda this quarter
None of the underlying causes above need a major program to fix. A realistic first quarter looks like this:
- Audit a sample of "resolved" contacts. Check them against what the definition requires, independent of what the dashboard already assumes.
- Calculate real coaching capacity. Work out the hours your current supervisor ratio genuinely allows in a week, and compare that against what your coaching model assumes happens.
- Test the standard directly. Ask every supervisor, independently, to describe what a good interaction sounds like, and see how far the answers diverge.
- Name an owner. Give one person clear responsibility for closing the gap between the written standard and the one currently being taught on the floor.
Getting honest answers to those four is worth more, this quarter, than committing to a training program or a vendor before you know which one you need.
Questions worth asking before you commit budget
Before committing budget to training, a restructuring exercise, or an outsourcing decision, a short set of questions tends to surface where the real problem is sitting, and who should be answering for it.
- Can every supervisor describe what good looks like for a call, chat or ticket in roughly the same words, or does it vary depending on who you ask?
- When did a supervisor last listen to a live interaction and coach on that specific interaction, rather than reading out a general score?
- What is the average number of direct reports per supervisor, and how much protected coaching time does that leave in a week?
- Is the knowledge base a living document updated against what has changed, or something built once and left alone?
- Are you measuring what good means to the customer as closely as you are measuring how quickly you got them off the phone?
If those answers are shaky, that is the real project, whatever the original request was. If you are already far enough along to be evaluating an outsourcing partner, three more questions belong on the list before a contract is signed: how they define "resolved" before you ever see a report from them, how coaching capacity is built into their staffing ratio rather than added afterward once quality slips, and how a service standard gets enforced consistently across a floor of hundreds rather than the handful of agents you met during the pilot. A contact centre consulting engagement, brought in before a shortlist is finalised, exists to pressure-test those answers with no attachment to any particular vendor's pitch.
Frequently asked questions
Who should own this problem: operations, HR or the executive team?
Whoever owns the budget the drift is quietly taxing, because that is who eventually pays for it in cost-to-serve, escalations or renewal risk. Operations owns the day-to-day fix, but the diagnosis and the funding decision belong at the level that can authorise a ratio change, a role redesign or an outsourcing conversation, which is usually above the operations team alone.
How quickly can drift be reversed once it is confirmed?
Faster than it built up, if the causes are addressed together rather than one at a time: a standards refresh without protected coaching time behind it will not hold, and coaching without a clearly defined standard to coach toward will not hold either. Tackled separately and in sequence, the same fixes take considerably longer, because each one keeps getting undone by the gap the others left open.
Do we need outside help, or can this be fixed internally?
It depends on whether the causes are a skills gap, a structural gap, or both, and whether your organisation has the bandwidth to run an honest audit of its own operation. Many of the diagnostic steps above can be run internally with the right discipline; an outside view earns its place when objectivity is the missing ingredient, or when the fix would otherwise compete for attention with everything else already on the leadership agenda.
About the author
Oliver Catt is the founder of The Cattalyst. He has spent 25 years inside contact centres, customer service, sales and training operations, and now helps organisations develop their leaders and build outsourced teams through a global partner network.
Start a conversation with Oliver
