Why Insurance Customer Service Teams Struggle to Deliver Consistent Agent Performance

There's a version of your contact centre that runs exactly as intended. Every advisor explains policy terms clearly. Cover is recommended based on actual customer need. Complaints are handled with patience and the right outcome in mind. Vulnerable customers are identified early and treated accordingly.

Most insurance customer service managers know that version exists because they've heard it. They've sat in on calls where everything goes exactly right. Where the advisor is calm, thorough, knowledgeable, and genuinely helpful. Where the customer ends the call feeling like they were looked after.

The problem isn't that your team can't deliver that. The problem is that they don't always deliver it consistently and most teams don't find out until something goes wrong.

Inconsistency in agent performance is one of the most persistent and underacknowledged problems in insurance customer service. It isn't usually about poor training or low standards. It's about the fact that a team of fifteen advisors, each handling dosens of calls a day, will inevitably develop slightly different habits. Different ways of explaining exclusions. Different thresholds for what counts as a "proper" vulnerable customer check. Different instincts when a customer pushes back on price.

Over time, those small differences compound. A customer who calls on Tuesday and speaks to one advisor gets a different experience from the customer who calls on Thursday and speaks to someone else. The outcome might be the same, the policy gets sold, the query gets resolved, but the quality of the interaction, and the risk embedded in it, can vary considerably.

Research backs this up. A study by McKinsey found that in customer service environments, top-performing agents can outperform average agents by a factor of eight or more on complex interactions, not because of talent differences, but because of habit differences that were never corrected or replicated at team level. In insurance specifically, where interactions routinely involve nuanced product explanations and regulatory obligations, that gap has real consequences.

For insurance firms operating under Consumer Duty, this variability matters more than ever. The FCA's expectations aren't just that firms have good processes. They're that firms can demonstrate consistently good outcomes, at the customer level, across every interaction, not just in the ones that were sampled.

Most contact centre quality assurance programmes run on random call sampling. A team leader or QA analyst listens to a small proportion of calls - typically somewhere between 2% and 5% of total volume - scores them against a framework, and flags issues for coaching.

The model works, up to a point. But it has a structural problem: by the time a pattern is identified through sampling, it's already been repeating for weeks. An advisor who consistently undersells key policy features, or who handles pricing objections in a way that increases churn risk, might not surface through random sampling for a month or more. The coaching conversation that follows is corrective rather than preventative. The behaviour has already affected real customers.

This isn't a failure of QA teams. It's a limitation of the tooling available to them. When you can only hear a small slice of what's happening on the floor, your coaching naturally becomes a response to what you happened to catch, not a proactive strategy built on a complete picture.

The Chartered Institute of Customer Management has noted that the majority of contact centre coaching remains event-driven - triggered by a complaint, a failed audit, or a supervisor overhearing a call - rather than stemming from systematic performance data. In sectors like insurance, where regulatory exposure compounds the risk of missed patterns, that reactive cycle is increasingly difficult to justify.

The downstream effects are significant. New starters embed bad habits before anyone notices. High performers don't get recognised or used as benchmarks. Team-wide patterns, a particular type of objection that everyone handles poorly, a disclosure that's frequently skipped under pressure, stay invisible until they become complaints or FOS referrals.

It's worth being specific about what inconsistency in agent performance can mean for an insurance contact centre.

Customer outcomes vary by advisor, not by situation. When quality of service is uneven, the experience a customer gets is partly luck, which advisor they happened to reach, what mood that advisor was in, how busy the queue was. That's a problem for firms trying to demonstrate fair treatment under Consumer Duty. The FCA's Financial Lives survey found that 7.4 million UK adults display one or more characteristics of vulnerability, a figure that rises to over half the adult population at points of stress or life event. The probability of one of those customers landing on a poorly performing advisor, on any given day, is not remote.

Complaints carry advisor fingerprints. FOS data consistently shows that complaint volumes aren't evenly distributed across teams. In 2023/24, the Financial Ombudsman Service received over 95,000 complaints about general insurance products alone, a 40% increase on the previous year. Internal analysis at many firms finds that a disproportionate share of upheld complaints traces back to a small number of advisors, often ones who weren't flagged as underperformers through standard QA sampling.

Onboarding takes longer than it should. Without clear benchmarks derived from your best performers, new advisors develop their own approach by trial and error. Industry estimates suggest it takes the average insurance contact centre agent between four and six months to reach consistent performance, a timeline that extends further when coaching is infrequent or not grounded in real call data.

Regulatory exposure is harder to manage. Consumer Duty requires evidence of good outcomes across your customer base, not just in the calls you chose to listen to. The FCA has been explicit that firms must be able to demonstrate this evidentially, not just assert it. Firms that can't show consistent quality across interactions are increasingly exposed when supervisory attention turns their way.

Most insurance customer service teams have well-defined quality standards. They have a call framework. They have a scoring rubric. They know what a good call looks like.

What they often lack is a way to measure how consistently those standards are being met across the full volume of interactions.

A QA framework without comprehensive data is a bit like a speed limit on a motorway without any cameras. Most drivers will probably stick to it. Some won't. And you won't know which is which until something happens.

The advisors who consistently fall short aren't usually aware they're doing so. The habits that cause problems, a slightly rushed risk disclosure, a tendency to move past objections rather than address them properly, a default script that doesn't flex for vulnerable customers, feel normal from the inside. They've been working that way for months, and nobody has told them otherwise, because nobody heard it.

The advisors who consistently exceed expectations often go unrecognised too. Their approach isn't captured, replicated, or used to inform how the rest of the team develops. According to research from Gallup, employees who receive regular, strengths-based feedback show up to 12.5% greater productivity, yet most contact centre coaching is still weighted toward remediation rather than recognition and replication of what works.

The shift from reactive to proactive coaching isn't a management philosophy question. It's a data question.

To coach proactively, you need to know what's happening across your team's interactions, not just in a random sample, but at the kind of scale that allows you to identify patterns.

That kind of visibility requires the ability to analyse calls at scale, against criteria that reflect your specific quality standards, not generic industry benchmarks, but the behaviours that matter to your firm, your products, and your regulatory obligations.

This is where platforms like Insights360 become relevant. Insights360 analyses customer interactions across the full volume of calls, not just the sampled ones, using scoring criteria that firms define themselves.

That means QA teams can move from reviewing a handful of calls per advisor per month to understanding patterns across everything. Coaching becomes grounded in what's actually happening across the team, rather than what happened to get picked up in a sample.

The goal of any customer service quality programme isn't to produce the occasional excellent call. It's to make that standard consistent, across advisors, across call types, across the working week.

That means having the data to know when you're hitting it and when you're not. It means building coaching conversations around what the evidence shows, not what a manager happened to overhear. And it means using the performance of your best advisors as an active input into how the rest of the team develops, rather than leaving that knowledge locked inside individual habits.

Insurance customer service is under more scrutiny than it's ever been. Consumer Duty has changed the regulatory expectation from "we have a process" to "we can show it works." Firms that can demonstrate consistent, evidenced quality across their customer interactions are in a fundamentally stronger position, both from a regulatory standpoint and a customer retention one.

Gartner research suggests that reducing customer effort is the single biggest driver of loyalty in service environments, not delight, not going above and beyond, but simply making the experience reliably good every time. That's a consistency problem as much as a performance one. And it starts with being honest about how much of what happens on your team's calls you can actually see.

Insights360 is an AI-powered conversational intelligence platform built for insurance contact centres. It analyses customer interactions at scale across every call, not just the ones that get sampled and scores them against criteria that your team defines. That means your quality framework, your compliance obligations, your coaching priorities, reflected in the data rather than imposed by a generic template.

For customer service leaders, that translates into something straightforward: visibility into what's actually happening on the floor, in a volume and at a speed that manual monitoring can't match. Where performance is drifting, you see it early. Where advisors are excelling, you can identify what they're doing and build it into how the rest of the team develops. And when the FCA asks how you're evidencing consistent customer outcomes under Consumer Duty, you have something concrete to point to.

Insights360 is developed by Conversant Technology, working with insurance firms across the UK to close the gap between the standards they set and the experience customers actually receive.
If you'd like to see how it works in practice, we'd be happy to talk.