The calls that matter most for regulatory risk, consumer duty obligations, and genuine customer outcomes are precisely the ones least likely to surface through traditional sampling.
What "treating customers fairly" actually sounds like under pressure
The FCA's Consumer Duty framework, which came into full force for insurance firms in 2023 and 2024, changed the stakes significantly. It's no longer enough to follow a complaints process. Firms are expected to demonstrate that customers receive good outcomes, including customers who are unhappy, confused, or actively disputing a decision.
What does good look like in a FOS-bound conversation?
-
The agent acknowledges the customer's frustration without becoming defensive
-
The firm's position is explained clearly, not recited from a script
-
The customer is told, genuinely, not performatively, that they have the right to escalate
-
There's no language that subtly discourages escalation, such as implying the process is long, expensive, or unlikely to succeed
-
The customer leaves the call (even if unhappy with the outcome) feeling heard and respected
What does poor look like?
-
Clipped, transactional language the moment FOS is mentioned
-
Passive deflection: "That's just our policy", "I've explained this already"
-
Technically accurate information delivered in a way that creates confusion
-
Tone that shifts from warm to cold when the customer becomes difficult
The difference between the two rarely shows up in a compliance checklist. It shows up in the language, the pace, the pauses, and the overall character of the conversation.
The regulatory direction of travel
The FOS published data in 2024 showing that complaints volumes across financial services, including insurance, remain persistently high, with motor and home insurance featuring prominently Ombudsman decisions consistently highlight communication failures as a root cause: not necessarily that firms made the wrong decision, but that they failed to explain it in a way the customer could understand or accept.
This is significant. It means a meaningful proportion of FOS cases aren't about whether the claim was right or wrong, they're about how the conversation went.
Firms that understand this are starting to look at complaints handling not just as a regulatory process but as a customer experience discipline.
The question isn't only "did we follow the procedure?" It's "did this person feel treated fairly?"
Those are not always the same question.
The intelligence gap
Here's the uncomfortable truth for many firms: you probably don't know what your agents are actually saying to customers who mention the FOS.
You might know what they're supposed to say. You may have scripts, training materials, a complaints handling policy. But whether that training translates into real conversations, under real pressure, with real customers is a different question entirely.
Closing that gap requires listening at scale. Not 2% of calls, but a far broader picture that captures complaint-related conversations specifically, surfaces language patterns, and gives quality and compliance teams the visibility they currently lack.
This is where technology has genuinely changed what's possible. Speech analytics and conversation intelligence tools can now identify the specific moments in a call where a complaint is escalating, flag language that falls short of Consumer Duty expectations, and give managers a factual, evidence-based view of how their teams perform when the stakes are highest, not just when the reviewer is watching.
What good oversight looks like
Firms that are ahead of this curve tend to share a few characteristics.
They treat complaint-handling quality as a distinct discipline, not a subset of general call monitoring. They track not just resolution rates and handling times, but customer sentiment and language quality throughout the conversation.
They use insight from FOS outcomes to feed back into training, not generically, but specifically, with real examples.
And critically, they don't wait for an ombudsman decision to tell them something went wrong. They have the visibility to identify the problem earlier, when there's still an opportunity to intervene.
That's not about catching agents out. Most agents, when they drift into poor practice under pressure, aren't doing it deliberately they're doing it because they haven't had the feedback, the coaching, or the support they need. Good oversight is what makes better performance possible.
