MEMA Guides / Consumer Duty
Consumer Duty outcomes monitoring: a practical guide
How to build outcomes MI, test what customers understand, assess fair value, write the annual board report and act on what you find.
- Annuallyminimum board review and approval of the Consumer Duty reportSource
- 180first-year board reports reviewed by the FCASource
- 56firms surveyed in the FCA's July 2026 outcomes monitoring reviewSource
- 79%of surveyed firms said they used complaints MI for customers in vulnerable circumstances, against 36% using behavioural insights dataSource
The guide in 30 seconds
Practical guidance- 01Define good before you measureWrite what a good outcome looks like for each product and journey, then pick the indicators that would show harm.
- 02Cut the numbers by customer groupAverage results hide customers who do worse. Look at vulnerability, channel and length of relationship.
- 03Show the fix workedEvery finding needs a root cause, an owner, a date and a re-test. The board report should show all four.
Explore the chapters
Read them in order, or jump to the one you need.
01What the FCA expects you to monitor, and whyThe Consumer Duty is not met by having good policies. It is met by being able to show, with data, what your customers actually experience. Monitoring is how you do that.
02Building your outcomes MI: metrics, thresholds and customer groupsGood MI starts with a plain statement of what a good outcome looks like, then picks a few indicators that would show harm. Starting from the data you already have gets it the wrong way round.
03Testing customer understanding and supportThe absence of complaints does not show customers understood anything. Light, repeated tests with real customers do.
04Price and value: fair value assessments and what the FCA has foundA fair value assessment is evidence that the price is reasonable compared with the benefits, for every group of customers. The FCA's findings show where assessments fall short.
05The annual board report: what goes in, who signs, what the FCA foundThe board report is the evidence the FCA asks to see first. A strong one reports what the MI shows, what went wrong, what you did and whether it worked.
06Acting on what you find: root causes, fixes and proofFinding a problem is the easy half. The FCA's complaint is that firms often show they acted but cannot show why they chose the action or whether it worked.
Chapter 01 · The starting point
What the FCA expects you to monitor, and why
The Consumer Duty is not met by having good policies. It is met by being able to show, with data, what your customers actually experience. Monitoring is how you do that.
You already know the Consumer Duty has four outcomes: products and services, price and value, consumer understanding, and consumer support. This guide is about the next step. How do you check, month after month, that customers are really getting those outcomes, and how do you prove it to your board and to the FCA? The rules sit in the Consumer Duty part of the FCA's Principles for Businesses (PRIN 2A), made by policy statement PS22/9. The FCA's guidance on how to apply them is FG22/5.
firms assess, test, understand and are able to evidence the outcomes their customers are receiving
That sentence from FG22/5 chapter 11 is the whole job. Four verbs: assess, test, understand, evidence. A policy that says you treat customers well does none of them.
Chapter 02 · Data and MI
Building your outcomes MI: metrics, thresholds and customer groups
Good MI starts with a plain statement of what a good outcome looks like, then picks a few indicators that would show harm. Starting from the data you already have gets it the wrong way round.
The FCA's July 2026 review highlighted a firm that defined good outcomes for each key product, across joining, using and leaving, and linked them to its customer journeys. Firms with weak ones had broad outcome statements, used operational proxies such as conversion rates or review completion, and could not explain their thresholds. Management information (MI) is simply the numbers you report to managers and the board, so the quality of your outcomes MI decides whether your board report means anything.
Four steps from outcome to MI pack
Stage 1 of 4
1. Define
For each product and journey stage (joining, using, leaving), write one or two sentences a customer would recognise. For example: a customer understands the cost before they apply and can reach a person if they are stuck.
- List your products and the journey stages for each
- Write the good outcome in plain words, per outcome
- Ask front-line staff whether they agree
Stage 2 of 4
2. Choose
FG22/5 11.33 lists the types of data to consider. Pick the few that would move first if customers were being harmed, and mix customer-level data with testing and feedback.
- Choose 3 to 5 indicators per outcome
- Include at least one forward-looking indicator, not only lagging ones
- Name the data owner for each
Stage 3 of 4
3. Threshold
The FCA found thresholds and RAG (red, amber, green) ratings that were not justified. It suggests baselining against past performance, or explaining the outcomes on either side of a threshold.
- Collect two to three months of baseline data
- Write down why each threshold sits where it does
- Agree what each colour triggers
Stage 4 of 4
4. Cut
PRIN 2A.9.10R requires you to find out whether any customer group gets different outcomes. Start with customers who have told you about a support need or characteristic of vulnerability, then channel, region and length of relationship.
- Add the group fields to each metric
- Accept small numbers: show them and explain, do not drop them
- Review the cuts monthly
| Outcome | Example indicators | Data source | Cut by |
|---|---|---|---|
| Products and services | Sampled files where the product matched stated need; decline and cancellation rates by lead source; sales from channels with poor completion | File reviews, CRM, lender feedback | Lead source, product, support need |
| Price and value | Total cost to customer by product; charge-related complaints per 1,000 customers; our commission against lender cost to customer | Lender data, finance system, complaints log | Lender, product, customer segment |
| Consumer understanding | Drop-off at fee and eligibility pages; comprehension test scores; questions asked about the same page | Web analytics, user testing, call tags | Device, support need, age band |
| Consumer support | First-contact resolution; repeat contacts within 7 days; time to reach a person; complaint themes | Contact centre and chat data, call QA | In-house or outsourced, support need |
FG22/5 11.36 expects MI on outcomes for customers with characteristics of vulnerability, of enough quality and depth to show which products and processes cause detriment. You do not have to collect new data on protected characteristics such as ethnicity (11.38), but if you already hold it the FCA expects you to use it to monitor differences in outcomes where possible. The FCA and the Information Commissioner's Office (ICO) issued a joint statement in March 2026 saying data protection law does not stop firms using vulnerability-related data where it is appropriate and necessary to support customers. You still need a lawful basis, and for health data an additional condition, so involve your data protection lead.
Illustrative example Not FCA figures. Use your own.
Chapter 03 · Testing in practice
Testing customer understanding and support
The absence of complaints does not show customers understood anything. Light, repeated tests with real customers do.
The FCA's March 2026 findings on consumer understanding are blunt. Testing was often superficial, one-off or poorly documented. Some firms relied on sales data or the absence of complaints as proof that customers understood, which the FCA says gives no reliable assurance. The payments review made the same point about weak proxies such as email open rates, which it said does not indicate understanding.
Testing the FCA found, and testing it praised
Poor practice
- One-off tests with no record of what changed
- Small, unrepresentative groups
- No testing with people who have accessibility needs
- Cosmetic changes with no check that they worked
- Open rates or lack of complaints used as proof
Good practice
- Test before and after a change
- Short surveys, comprehension checks, A/B tests, callbacks
- Document what changed, why, and the impact
- Test with different customer groups
- Set a target, for example at least 80% correct recall of key points (one firm's own target)
A light test cycle for a small team
Pick the moment that matters
Choose the page, letter or call where a misunderstanding would cost the customer money: fees, eligibility, cancellation, what happens if they miss a payment.
Ask three to five recent customers
The FCA gave fortnightly tests with 3 to 5 recent customers, phone read-backs and post-sale comprehension calls as proportionate examples for smaller firms. Ask them to say in their own words what they agreed to.
Include people who may struggle
Include at least one customer with an accessibility or support need each round, and record preferences once so you do not ask again.
Change one thing, then re-test
Fix the wording or layout, release it, and measure the same indicator again. Write down the before and after.
Illustrative data. Analytics showed about a third of visitors left at the fee explanation page. Testing with five customers showed they did not understand how Brightloan was paid. The page was rewritten in plain words with a short FAQ and released at the end of month 6. Drop-off fell and stayed down.
Illustrative example Not FCA figures. Use your own.
Chapter 04 · Fair value
Price and value: fair value assessments and what the FCA has found
A fair value assessment is evidence that the price is reasonable compared with the benefits, for every group of customers. The FCA's findings show where assessments fall short.
The price and value outcome (PRIN 2A.4) requires that the price a customer pays is reasonable compared with the overall benefits of the product. The FCA's price and value findings, updated in July 2026, apply to firms of all sizes. The Duty allows a proportionate approach that reflects the nature, scale and complexity of the business and product, but assessments must still be sufficiently robust. The FCA does not set prices or require them to be low. It does expect evidence.
The payments review found many assessments fell short mainly in supporting analysis. Some relied on price benchmarking alone and did not consider non-financial benefits such as consumer support, and many listed price elements but never concluded whether fair value was provided.
Fair value assessment: weak and strong
Weak
- Compares our price with competitors and stops
- Treats being in line with the market as proof of value
- One assessment for products with different terms
- Does not look at customers with vulnerability characteristics
- Concludes nothing
Strong
- Looks at total expected cost by customer segment
- Shows the benefits customers actually receive
- Groups only similar products and customers
- Asks whether customers with vulnerability characteristics are less likely to get fair value
- States a conclusion and the action taken
Illustrative data. Complaints about charges were three times higher on one lender's product than on the others. Brightloan shared the data with the lender and reviewed its panel position. This is a finding in the worked example, not real data.
Illustrative example Not FCA figures. Use your own.
Chapter 05 · Governing body
The annual board report: what goes in, who signs, what the FCA found
The board report is the evidence the FCA asks to see first. A strong one reports what the MI shows, what went wrong, what you did and whether it worked.
PRIN 2A.8.3R requires you to prepare a report for your governing body on the results of your monitoring under PRIN 2A.9, and any action needed as a result. At least annually, the governing body must review and approve the report, confirm whether it is satisfied the firm is complying with the Duty, and assess whether future business strategy is consistent with the Duty (PRIN 2A.8.4R). FG22/5 10.14 says you should expect to give the FCA the report and the MI behind it on request.
What the FCA found in 180 first-year board reports
Areas for improvement
- Conclusions not supported by the MI. One firm described extensive vulnerability support while its MI showed fewer than half of customers receiving good outcomes
- Thresholds and RAG ratings without justification
- Action plans with no owners or timescales, and no test of whether actions worked
- Board challenge not evident, for example in the minutes. The FCA says the board should not be a rubber stamp
- Reports written almost entirely by compliance
- No evidence that outsourced support delivered good outcomes
Good practice
- A clear focus on the four outcomes
- Commentary backed by good quality MI
- Analysis of different customer groups, including vulnerability
- A defined process to produce the report in time for board review
- Complaint trends linked to root causes
- Specific, measurable actions with an accountable executive, tracked through governance
Read the report four ways
Your job is to challenge, and to be seen to.
- Ask for target market and user testing information if it is missing
- Keep a tracker of the board's requests and the answers
- Check that every red indicator has an owner and a date
You own the outcomes and the strategy link.
- Check the report says how strategy and incentives support good outcomes
- Make sure business leaders, not only compliance, wrote their sections
- Own the action plan
You assemble the evidence and keep it honest.
- Define what good looks like before presenting the MI
- Acknowledge data gaps and the plan to close them
- Do not let the report claim more than the MI supports
Keep it proportionate but evidenced.
- A single senior lead, clear decision notes and simple logs can be enough
- A knowledgeable impartial adviser can act as a critical friend on the approach and report
- Use the few indicators you can sustain: complaints, feedback, missed service standards, file checks
| Section | What it says |
|---|---|
| Overall conclusion | Outcomes were good for most customers but not all. Two red findings and three amber findings were found in months 4 to 6. Actions are in the tracker; two red findings have been fixed and re-tested. |
| Consumer understanding | 31% of visitors left the fee page, threshold 20%, set from the first three months' baseline. After a plain-language rewrite, tested with five customers, drop-off is 16% in month 12. |
| Consumer support | Customers with a recorded support need completed applications at 58% against 71% for others. A callback option and staff training closed the gap to 2 points. Outsourced chat repeat contacts fell from 27% to 14% after new MI terms and call QA. |
| Price and value | Charge complaints on Lender B's product were 12 per 1,000 against a panel average of 6. The lender cut one fee after review. Complaint rate to be re-checked at month 15. |
| Strategy | The board reviewed the plan to grow through affiliates and agreed to cap volumes from the one source with poor completion until its promotion is corrected. |
| Challenge recorded | The champion asked for target market evidence and for support-need data by age band. Both are appended with the minutes. |
Self-check: does your board report contain these?
Tick each area your plan already covers. Your ticks stay in this browser only.
Chapter 06 · Making it work
Acting on what you find: root causes, fixes and proof
Finding a problem is the easy half. The FCA's complaint is that firms often show they acted but cannot show why they chose the action or whether it worked.
FG22/5 11.8 lists the kinds of intervention you might use: adapt or stop a product, change design, fees or charges, change operations, update support or distribution channels, make communications easier to understand, or provide redress where customers have suffered harm. The July 2026 findings add the test. Firms often showed they took action but did not explain the cause or whether the fix worked.
From finding to proof
Rate and log the finding
Record the finding, the outcome affected, the rating and the evidence. Keep the MI you considered, what you challenged and what you decided. The FCA criticised a firm that could not show that trail.
Find the root cause
Ask why until you reach something you can change. Complaints reviews should look at themes and customer groups, not only volumes and handling times.
Choose and justify the action
FG22/5 11.9 expects you to explain how you reached the decision on the most appropriate intervention. Training alone is rarely a root-cause fix if the process is the problem.
Name an owner and a date
Put it in one tracker overseen by the champion, with status updates.
Re-test with the same measure
Check again after the change has had time to take effect. If it has not worked, record what you did next.
| Finding | Outcome | Rating | Action and owner | Due |
|---|---|---|---|---|
| 31% of visitors leave at the fee page; customers could not say how Brightloan is paid | Consumer understanding | Red | Rewrite page in plain words, add FAQ, test with five customers. Head of product | Month 7 |
| Customers with a support need complete at 58% against 71% for others | Consumer support | Red | Add callback option, train staff to offer it, add flag to CRM. Head of operations | Month 8 |
| 27% of outsourced chats need a repeat contact, against 12% in-house | Consumer support | Amber | Add MI and QA terms to contract; sample 20 chats a month. Head of operations | Month 7 |
| Charge complaints on Lender B product twice the panel average | Price and value | Amber | Share data with lender, review panel position, decision at committee. Head of partnerships | Month 9 |
| One affiliate source sends applicants who are declined at three times the rate | Products and services | Amber | Require affiliate to correct promotion; cap volumes meanwhile. Head of marketing | Month 8 |
| Complaints reported by volume only, not by customer group | Monitoring | Amber | Add group fields and theme tags to dashboard. Data analyst | Month 5 |
| 48 of 50 sampled files showed the product matched the stated need | Products and services | Green | No action. Repeat quarterly. Compliance lead | Quarterly |
Illustrative data. The gap between customers who had told Brightloan about a support need and other customers fell from 13 points to 2 after a callback option and staff training. The FCA's own examples include a lender that cross-referenced declines with recorded support needs and reviewed its decision policy.
Illustrative example Not FCA figures. Use your own.
Questions for your board
- What does a good outcome look like for each of our key products and journeys, in words a customer would use?
- Which of our indicators are about customers, and which only describe our own activity?
- How were our thresholds set, and what happens at red?
- Do we know how customers with characteristics of vulnerability fare against others, and do we act when they do worse?
- What did we find last year, what was the root cause, and did the fix work?
- Do outsourced providers and distributors give us the MI we need to see their outcomes?
- Where in the minutes can a reader see that we challenged this report?
- Is our strategy, including growth and incentives, consistent with what the MI is telling us?
Where this comes from
Sources
Every number and rule in this guide comes from these pages. Check them before you rely on anything, as rules change.
- FCA Handbook PRIN 2A.8: Governance and culture (annual board report)
- FCA Handbook PRIN 2A.9: Monitoring of consumer outcomes
- FCA PS22/9: A new Consumer Duty
- FCA FG22/5: Final non-Handbook guidance for firms on the Consumer Duty (chapters 10 and 11)
- FCA: Outcomes monitoring, good practice and areas for improvement (July 2026)
- FCA: Consumer Duty board reports, good practice and areas for improvement (December 2024, updated 2026)
- FCA: Price and value outcome, good and poor practice (updated July 2026)
- FCA: Consumer understanding, good practice and areas for improvement (March 2026)
- FCA: Consumer support outcome, good practices and areas for improvement (March 2025, updated July 2026)
- FCA: Delivering good outcomes for customers in vulnerable circumstances (March 2025)
- FCA: Complaints and root cause analysis, good practice and areas for improvement
- FCA: Payments Consumer Duty multi-firm review (October 2024)
- FCA and ICO: Joint statement on firms' approaches to vulnerability-related data (March 2026)
- FCA: Our Consumer Duty focus areas (September 2025)
- FCA: Consumer Duty requirements review update
This guide is general information, not legal or regulatory advice. Rules and FCA expectations change; check the sources and take advice on your firm's circumstances.