FCA Consumer Duty Call Monitoring: A Practical Checklist for UK Firms
Since the Consumer Duty came into force, call monitoring stopped being a quality nicety and became an evidence requirement. Under Principle 12 and PRIN 2A, regulated firms have to deliver good outcomes for retail customers — and, crucially, demonstrate that they do. For any firm whose customers are won, served, or retained over the phone, the calls are where much of that evidence lives.
The problem most contact-centre operations run into is simple: the monitoring process they built for the old "treating customers fairly" world — a QA analyst spot-checking one or two per cent of calls against a scorecard — was never designed to evidence outcomes across an entire customer base. This guide sets out what the Duty actually expects you to monitor on calls, and a practical checklist to measure your current process against.
This is general information for operational and compliance teams, not regulatory or legal advice. Read it alongside the FCA's own guidance — FG22/5 and PRIN 2A — and take your own compliance view.
What the Consumer Duty asks of call-based firms
The Duty is built around four retail-customer outcomes — products and services, price and value, consumer understanding, and consumer support — sitting under three cross-cutting rules: act in good faith, avoid foreseeable harm, and enable customers to pursue their financial objectives.
For a firm that sells or services over the phone, three of those translate almost directly into things you can hear on a call:
- Consumer understanding — did the customer actually understand the product, the price, the risks, and what they were agreeing to?
- Consumer support — was it as easy to cancel, complain, or change their mind as it was to buy? Or did the call introduce friction at exactly the wrong moment?
- Avoiding foreseeable harm — including how the firm handles customers in vulnerable circumstances.
The FCA has been explicit that this is an ongoing, evidenced obligation, not a one-off implementation project. Outcomes monitoring is one of the regulator's stated multi-firm review areas for 2025/26 — it is actively examining how firms monitor consumer outcomes and whether they can show it. Firms also have to feed this into an annual board report on whether they are delivering good outcomes. Calls are a primary source of that evidence.
Why 2% sampling no longer holds up
The uncomfortable maths of traditional QA: a human analyst can realistically review a tiny fraction of calls — often one to two per cent. That was defensible when the goal was coaching agents on technique. It is much harder to defend when the question becomes "show me that customers across this product understood what they were buying, and that vulnerable customers were handled correctly — all of them, not the handful you happened to listen to."
A 2% sample tells you about 2% of your outcomes. It says almost nothing about the 98% of calls nobody listened to — which is exactly where an unnoticed pattern of poor outcomes would hide. The reason firms are moving toward monitoring 100% of interactions is straightforward: partial sampling cannot evidence a whole-population obligation.
The practical checklist
Measure your current call-monitoring process against the following. For each, the test is not only "do we do this?" but "could we show the regulator the evidence across our whole call base, not just a sample?"
Consumer understanding
- Are key terms, prices, exclusions and risks disclosed clearly, without jargon or rushing?
- Does the agent check and confirm the customer has understood — not just recite a script?
- For any sale, is there a clear, unpressured explanation of what the customer is committing to?
Consumer support
- Are opt-outs, cancellations and "I'd like to think about it" handled as easily as the sale itself — no friction added at the point of leaving?
- When a customer asks not to be contacted, is that honoured immediately and logged?
- Are complaints recognised on the call and routed correctly, with empathy?
Vulnerable customers
- Are vulnerability signals — bereavement, confusion, financial distress, health conditions, age-related need — being detected on calls?
- When they appear, does the agent adapt: slow down, offer a call-back, apply the firm's vulnerability process, and flag the customer for the next interaction?
- Can you evidence how vulnerable customers were treated as a population, not anecdotally? The FCA's 2025 review told firms specifically to strengthen outcomes monitoring and support here.
Conduct and foreseeable harm
- Are pressure selling, age- or fear-based persuasion, and misleading or overstated claims being caught?
- Is there a reliable way to surface the worst conduct calls for review, rather than hoping they fall into the 2% sample?
Evidence and audit trail
- Are calls recorded, indexed and searchable, with a retention policy that matches your regulatory obligations?
- If the FCA asked for every call where a customer raised an affordability concern over the last 18 months, could you produce them — quickly?
- Does your monitoring produce management information that can feed the annual Consumer Duty board report?
The data-protection tension nobody mentions
There is a catch in monitoring calls at scale. The same recordings that evidence good outcomes are full of personal and special-category data — names, dates of birth, bank details, health disclosures. The FCA and the ICO have acknowledged this directly, publishing a joint statement on how firms should balance vulnerability and outcomes monitoring against their data-protection obligations, with further guidance expected.
In practice that means the answer to Consumer Duty cannot be "record and keep everything, in the clear, forever." Monitoring at full coverage has to come with proportionate controls: personal data redacted by default in transcripts, access to the raw detail role-gated and audited, and a clear retention regime. Evidencing outcomes and protecting data are not in conflict — but only if the monitoring system is designed for both from the start.
Moving from sampling to coverage
The shift most firms are making is from sampling for coaching to monitoring for outcomes — from a human listening to a handful of calls, to automated transcription and analysis across every call, with human time spent on the calls the system flags. That is what makes a whole-population obligation answerable: every call transcribed and searchable, scored for the outcomes and conduct signals above, with personal data handled safely and a clear audit trail behind it.
Telebyte builds exactly this kind of monitoring into the contact-centre estates it runs — transcription and analysis across the full call base, conduct and vulnerability flagging, semantic search, and PII redaction with audited reveal — on infrastructure designed for FCA-regulated firms. The reference deployment it operates is a live, FCA-regulated brokerage running the whole stack.
Ready to pressure-test your own process? We've turned the checklist above into a free, more detailed Consumer Duty Call Monitoring Readiness Checklist you can work through with your compliance and operations teams — no sign-up wall.