If you run compliance for a growing UK fintech, Consumer Duty is rarely the problem as written. The FCA's outcomes (products fit for purpose, fair value, customer understanding, customer support) aren't hard to agree with. The problem shows up where the outcomes meet the marketing workflow. A promotional email that's "fine in policy" but reads at 18+ for a product targeting first-time savers. A landing page headline that's technically accurate but materially misleading to a reasonable consumer. A risk warning that's present but doesn't survive the mobile feed it ships in.
The FCA intervened in 19,766 financial promotions in 2024, roughly double the 2022 volume. Most of those interventions weren't because firms ignored Consumer Duty. They happened because teams were checking the wrong things, at the wrong stage, with the wrong people in the room.
The commercial reality: Consumer Duty is an operational standard, not a policy
Consumer Duty failures don't announce themselves as Consumer Duty failures. They arrive as campaign delays, unexpected FCA interventions, and subjective rejection queues where the same promotion gets approved twice and rejected once depending on who reviews it.
In one mid-size wealth platform we've reviewed, the marketing team was working to an explicit target reading age of 11-13 on promotional materials. Ads, landing pages, emails, push notifications, all checked against that standard at submission, not at review. The target wasn't aesthetic. It was operationalised Consumer Duty. If the average UK reading age sits around 9-11, a financial promotion aimed at a non-expert audience has to clear that bar by design, not by hoping reviewers catch it downstream.
Most workflows still treat readability as a nice-to-have. The approval checklist checks claim evidence, disclosure accuracy, COBS 4.2 alignment. Then a communication passes substance review and gets rejected in a second round because the paragraph is 280 words long and the average sentence spans 34 syllables. That's not a compliance failure. It's a workflow that has its checks in the wrong order.
Six places Consumer Duty quietly breaks the marketing workflow
1. Confusing T&Cs compliance with communications compliance. Your terms are watertight. That's not the same as a fair, clear, not-misleading promotion. The T&Cs disclose. The promotion persuades. The bar is different, and the FCA does not accept "it's covered in clause 8" as an answer to what the promotion itself said.
2. Treating Consumer Duty as a policy problem, not an operational one. Most firms updated a policy document in 2023 and called it done. The operational question (how do we embed this into every promotion, every campaign, every customer-journey touchpoint) got left as an exercise for later. One team we worked with moved from roughly 40% first-time approval to around 80% by tightening the operational standard at the entry point, not by hiring more reviewers. That's a systems win, not a compliance win.
3. Subjective review standards that shift by reviewer. "Fair" and "clear to the ordinary consumer" aren't binary states, and even experienced reviewers disagree at the edges. In one enterprise workflow, roughly half of first-time rejections weren't genuine compliance breaches. They were workflow and admin failures. Missing context. Unclear submission standards. Inconsistent application of house style. Different reviewer, different outcome. That's not a defensible control.
4. Readability isn't checked at submission. Most approval workflows check substance first and readability second, if at all. This is backwards. Reading-age and sentence-density checks are mechanical. They should fail the ad before it reaches a reviewer. The reviewer's time is for judgment calls: is this claim fair, is this prominence adequate, does this framing mislead. Not "this paragraph is too long."
5. Approval is siloed between Compliance and Marketing. Compliance owns regulatory substance. Marketing owns brand and commercial impact. Neither team owns the intersection, the moment a regulatory expectation becomes a customer-facing message. Under Consumer Duty, that intersection is where outcomes are decided. If no one owns it, feedback loops stay slow and late.
6. No systematic capture of real-world failures. When a promotion gets pulled by the FCA, or pulled internally after launch, it rarely makes it back into the approval workflow as a data point. Teams make the same mistake in different campaigns because there's no feedback loop. The next reviewer has no way to know the pattern exists.
What Consumer Duty actually requires of marketing communications
Three practical tests that the FCA's outcomes translate into for a promotional asset:
- Consumer understanding. Not "the information is present." The test is whether a reasonable customer in the target audience can understand what is being offered, what it costs, and what the risks are, at the right time, in the form it was presented. Truncated risk warnings in a mobile feed, footer-buried fees, jargon that doesn't land for the audience. All of those fail this test regardless of whether the information was technically available.
- Avoiding consumer harm by design. The FCA's framing is that firms must act to deliver good outcomes, not merely avoid bad ones. For marketing, that means the absence of a specific rule breach isn't enough. If a layout, ordering or emphasis pattern predictably reduces understanding, even when the text is accurate, the firm is expected to have caught and corrected it.
- Proportionate information to the decision. Pre-contract information must support a considered decision. For a high-risk product, the prominence of risk information has to reflect the weight of the decision the customer is being asked to make. A small, footer-weight warning under a headline offer is not proportionate and has been repeatedly named in FCA reviews as a failure mode.
None of these is radical. What is different under Consumer Duty is that the firm has to be able to demonstrate how each was considered and decided, at asset level. That demonstration burden is where most workflows still don't have the plumbing.
Quick diagnosis: is your workflow ready?
Answer these six questions about your current approval flow:
- Readability at submission. Does your submission template check reading age, sentence length and paragraph density before human review?
- First-time approval rate. What percentage of promotions clear review on first submission? Below 70% suggests entry standards are loose; below 50% suggests they're non-existent.
- Handoff points. How many times does a single promotion move between teams before launch? More than 3 points to a silo problem.
- Feedback loop. When something fails in the real world (FCA intervention, internal pull, customer complaint), does it feed back into the approval workflow as a tagged, searchable data point?
- Consistency. Could two reviewers in your team approve different versions of the same promotion? If yes, your standards are too subjective to be defensible.
- Median time to approval. What's the median from submission to approval for a standard asset? Over 5 business days points to workflow friction, not review depth.
If three or more of those answers are uncomfortable, the issue is your workflow your compliance team operates in.
Practical fix: four steps
1. Establish a single source of truth for what "Consumer Duty compliant communication" means at your firm. Not a policy document. A set of working standards: what clarity requires, what readability thresholds apply, what evidence standard each claim type needs, what the escalation matrix looks like. Reviewers should be able to reference it at decision time, not ask colleagues.
2. Move mechanical checks upstream to submission. Reading age, sentence length, prominence-ratio of warnings, disclosure placement, claim-to-evidence linkage. All mechanical. Build them into submission, not downstream review. Make it impossible to submit an asset with a reading age above 16 without an explicit documented override.
3. Standardise the approval template. Who reviews? What do they check? In what order? What counts as a blocker vs. a flag? Document it in the approval system so every reviewer sees the same gates. Consistency is what makes the audit trail defensible under FCA scrutiny.
4. Close the feedback loop. Log every real-world failure (FCA intervention, internal pull, customer complaint traced to communication) and tag it against the approval record that shipped it. Use the log to retrain reviewers, tighten entry standards, and flag patterns before they recur. The feedback loop is what turns reviewers from gate-keepers into pattern-finders.
Objection: "Governance first, tools later"
This objection comes up in every scaling compliance conversation and it is partly right.
A platform that enforces bad standards at speed is just bad standards at scale. Get the standard right before you automate it. That's true.
What's also true: most teams use "governance first" as a reason to delay the operational build indefinitely. Six months later the policy document is tighter and the workflow is still the same. The point isn't to tool before you govern. It's to recognise that governance without operational plumbing is theory, not defence.
The sequence that works is simple enough. (1) Define the standard and write it down. (2) Pilot it manually on a small sample to verify it's reviewable at all. (3) Once you can describe the mechanical checks precisely, automate the 80% that's repeatable so the reviewer has time to apply the 20% that's judgment. Automating before step 2 is premature. Staying at step 1 indefinitely is cowardice dressed as caution.
When manual works and when it doesn't
Manual approval works for firms producing fewer than ~50 promotional assets a quarter, with a small stable compliance team, a single channel, and relatively simple product structure. The decision volume is low enough that consistency holds by habit.
It breaks when any of those change. Volume crosses that threshold. Channels multiply. Products layer. The team grows beyond the point where three people can review the same way by default. At that point manual workflows create bottlenecks faster than people can clear them, and first-time approval rates drop because reviewer fatigue compounds with inconsistency.
The trigger for dedicated tooling isn't "more than X assets." It's "we can no longer predict whether a specific asset will be approved." When reviewers can't confidently predict their colleagues' decisions, the standard isn't operating as a standard anymore. That's the moment to automate the mechanical gates and free the humans for the judgment calls.
FAQ
What first-time rejection rate should we expect?
Across regulated marketing workflows, first-time rejection rates typically sit 20-30%. Above 40% suggests entry standards are too loose or submission guidance is unclear. Below 10% suggests either a very mature workflow or inadequate checking.
Can AI-powered readability tools replace compliance review?
No. Readability tools catch mechanical issues like reading age, sentence length, passive voice, and jargon. They don't catch unfair framing, misleading claims by omission, or prominence failures. Use them as a first gate before human review, not as a substitute for it.
Should compliance or marketing own the approval workflow?
Neither alone. Shared ownership with written standards works better. Compliance owns the regulatory gates. Marketing owns the commercial gates. Operations owns the workflow itself: routing, SLAs, audit trail.
Does Consumer Duty require us to test readability with actual customers?
Not explicitly, but the FCA expects firms to demonstrate that target customers can understand material communications. Some firms use reading-age metrics, some run small customer panels, some both. The expectation is evidence-based, not process-based.
How do we handle vulnerable customer requirements under Consumer Duty?
Under the vulnerable customer guidance, firms must consider whether communications reach and serve customers in vulnerable circumstances. For promotional assets, this usually means plain language, clear next steps, accessible design, and disclosure placements that don't rely on fine print or hover states.
What's the audit trail the FCA expects?
Per-asset: approver, approval date, evidence supporting claims, disclaimer variant used, final published version, and a record of the review decisions made. If a promotion can't be traced to an approval record at variant level, that's a finding.
Related reading
- Why Meta Rejects Financial Ads and How UK Fintech Teams Fix It
- $3 Billion in Marketing Compliance Fines: The 2026 Operating Model US Firms Need
Adclear is automated pre-submission marketing-compliance software for FCA-regulated firms. This article is guidance, not legal or compliance advice. Firms remain responsible for their own financial promotions under FCA rules including COBS 4.2, SYSC 9.1R, the Consumer Duty (PRIN 2A) and FG24/1. Facts reflect public FCA guidance as of April 2026.
Once you're booked in, we'll send you a free playbook on Financial promotions compliance for FinTechs.
