Executive summary
- About 80% of adviser-charge revenue is ongoing, from around 4 million clients. When the FCA asked 22 of the largest firms to prove the promised reviews took place, around 83% had been delivered, 15% were declined or went unanswered, and in fewer than 2% nobody had tried.
- Only 40% of advice firms reviewed their cost to serve a typical client in the past year, according to NextWealth's 2026 benchmarks. Fair value can't be judged on an average, and most firms don't have the figure that would test it.
- In the worked example below, £474k of ongoing fees — 10.7% of the total — came from 195 clients with no review that year. £49k of it came from 27 clients nobody tried to reach.
- The same firm's smallest clients pay 70p for every £1 they cost to serve; its largest pay £3.41. Both ends are fair-value questions, and only a segment view shows either.
Financial services at DataHexis
"Portfolio analytics, risk dashboards, FCA reporting, and client performance summaries." This piece is about the FCA reporting that advice firms find hardest to evidence: whether the ongoing service each client pays for was actually delivered.
— from the Industries section of the DataHexis websiteAsk an advice firm how many clients it has on ongoing service and the answer comes back in seconds. Ask how many of them had the review they paid for in the last twelve months, and how the firm knows, and the answer usually arrives a week later, from three systems and a spreadsheet.
That delay is the problem the FCA spent 2024 measuring. Fees are collected by the platform on a schedule. Reviews live in a diary, a back-office record and somebody's inbox. When the two drift apart, the firm is still paid, the client is still charged, and the only record that anything is wrong is the one nobody has built.
Charged by default, delivered on purpose
Review delivery rate = reviews delivered ÷ reviews contracted to fall due in the period. A review counts as delivered when there is evidence it happened — a meeting held and a record issued — not when an invitation went out.
Everything that isn't delivered falls into one of two very different buckets: declined or no response, where the firm offered and the client didn't take it up, and not attempted, where nobody offered at all. The first is an engagement question. The second is a redress question.
The FCA's multi-firm review used exactly that split. It wrote to 22 of the largest advice firms in February 2024, asked for seven years of data, and published the results a year later. The headline was reassuring, and the regulator said it did not currently view the issue as systemic. The detail was less so.
Among the poor practice the FCA listed was "insufficient management information to allow senior management to have adequate oversight of the services", alongside record-keeping that couldn't evidence delivery. Among the good practice was a policy to stop collecting fees from clients who hadn't engaged for a period of time. Both are reporting problems before they are compliance ones.
A fee is collected by default. A review only happens on purpose.
The FCA's sample was the largest firms, and it said plainly that the sample wasn't representative. Smaller firms weren't in it, so nobody knows their figure — often including the firms themselves. What a look-back to 2018 actually tests is not whether reviews happened, but whether the firm can prove it.
Fair value is a segment, not an average
Most ongoing advice is charged as a percentage of assets. The FCA's last full evaluation put the average ongoing charge at 0.8% a year; NextWealth's 2025 benchmarking found it up 9 basis points since 2023, with a growing share of advisers charging between 0.91% and 1%, and 70% charging on assets. A percentage fee scales with the pot. The work mostly doesn't.
Cost to serve = the adviser, paraplanner and administration time a client's ongoing service actually consumes, valued at cost rates, plus the platform, software and regulatory cost that travels with each client. Compare it with the fee the client pays and you get a fee-to-cost ratio: below 1.0, the client is subsidised; well above it, the fee needs a reason.
The FCA's 2024 letter to advice firms said it was concerned firms "may not be adequately considering the relevance and costs of these services for all clients". Its review of Consumer Duty board reports singled out, as good practice, firms that "considered the levels of fees against costs of providing a product". And yet only two in five firms told NextWealth they had reviewed the cost to serve a typical client in the past twelve months.
Two things follow. First, an average client doesn't exist: a firm-wide ratio of 1.7× can sit on top of a segment paying less than it costs and another paying three times more. Second, the answer isn't always to charge the small client more. Nearly half of firms (44%) have already turned off ongoing fees for clients who no longer fit their service, which is a fair-value decision made with data or made without it.
The average client is fairly charged. Almost no actual client is the average client.
What the receipt gap is worth
Put the two measures together and the arithmetic stops being abstract. In the worked example below, a sixteen-adviser firm collects £4.44m a year in ongoing fees from 1,180 clients.
Its review delivery rate is 83.5%, almost exactly the FCA's figure. That leaves 195 clients who paid an ongoing fee and didn't receive a review that year: £474k of fees, or 10.7% of ongoing income. Most of it is the declined-or-no-response bucket, which is defensible if the firm can show it offered. £49k of it came from 27 clients nobody tried to reach, which isn't.
That £49k is the figure a redress conversation starts from. The £474k is the figure a fee-off policy should be written against. Neither appears in a standard back-office report, because the fee lives on the platform and the review lives in the back office.
Choose the review frequency, and expect to defend it
In March 2026 the FCA proposed, in CP26/10, "replacing the annual suitability review with periodic reviews based on clients' needs", alongside clarifying "how firms support disengaged clients, underpinned by the Consumer Duty". The consultation closed in May; the policy statement will follow.
Read that as flexibility and it sounds like less work. Read it as a reporting requirement and it's more: a firm that sets review frequency by client need will need to show which clients are on which cycle, why, and whether the cycle was met. A fixed annual review is easy to count. A justified one has to be evidenced.
The FCA's reviews of Consumer Duty reporting point the same way. Board reports were criticised because "some firms did not have sufficient data quality to justify conclusions" and because they "often failed to provide well-reasoned justifications for setting thresholds". Its July 2026 review of outcomes monitoring added that "some firms collected relevant MI but could not show how it helped them make decisions or improve customer outcomes".
When the regulator lets you choose the frequency, it is also letting you be asked why.
A worked example: the ongoing service view
Here's what the reporting looks like for an independent advice firm with sixteen advisers across two offices in the South East, 1,180 clients on ongoing service and £612m under advice. It's a composite of typical firm structures rather than a single client, and the figures are illustrative — but they're calibrated against the benchmarks cited throughout, and every number reconciles to the others.
The conventions are stated rather than assumed: all figures are a rolling twelve months; percentages are of total fee income (£5.46m, of which £4.44m ongoing and £1.02m initial); ongoing fees are tiered by assets, from 1.0% under £150k down to 0.6% above £1m.
| Line | % of fee income | Note |
|---|---|---|
| Adviser and paraplanner pay | 44% | Salaries, bonus and on-costs, including principals' salaried pay |
| Client services and administration | 15% | Review scheduling, valuations, onboarding, platform admin |
| Regulatory | 8% | FCA fees, FSCS levy, PI cover and compliance support |
| Technology, platforms and premises | 14% | Back office, planning and research tools, offices |
| Operating profit | 19% | Before owners' dividends and tax |
Cost to serve below is the part of the first three lines that ongoing clients consume, plus their share of technology: £2.56m, or 47% of fee income. Premises and management overhead are not allocated, so segment contribution is before those costs. Four sources feed the model:
The operations director opens this view every Monday; the compliance lead uses the same model for the quarterly Consumer Duty pack. Nobody rebuilds the numbers by hand for either.
Ongoing service view: whole firm
Rolling 12 months, refreshed overnightDelivery leads the row: of the five tiles, it's the one the operations team can move before the next board meeting. The 90% threshold is written into the board report with its reason — it's the level at which every undelivered review can be individually explained to a reviewer. Statuses are spelled out in words as well as colour, for anyone reading in greyscale or on a projector.
Reviews falling due, next seven months
693 reviews in seven months, bunched into January to March, when the tax-year conversations land. The capacity question is answered before the diary fills, not after the misses appear.
Engagement against value: where each client sits
Reviewed, and the fee covers the cost. Keep the service level and keep the evidence.
Reviewed, but the fee is below cost to serve. Re-tier the service or reprice it.
Fee above cost, no review. The receipt gap. Re-engage and document it, or stop the fee.
Low fee, not engaged. Ask whether ongoing advice still fits, or offer a lighter service.
On segment-level cost to serve, the four boxes hold 772, 213, 108 and 87 clients. The 108 paying for silence account for £391k of the year's ongoing fees.
The board view: five charts for the Consumer Duty report
The Monday view is for the people running the diary. The board needs to see whether delivery is improving, where the misses concentrate, whether each segment receives fair value, and what the firm is collecting without a review behind it. Every chart below answers a question the FCA has said board reports struggle with.
Review delivery rate by month
From 78.2% to 89.8% in a year. December and August dip with the holidays; the step change came when the segment view went live in January and the fee-off policy in June. A threshold only means something if the board can see which month crossed it and why.
Reviews not delivered, by segment
The firm-wide 16.5% belongs to nobody. In the smallest segment 29% of reviews didn't happen, and 5% were never offered; above £1m, 5% and none. Aggregated MI is exactly what the FCA says hides differences between groups of clients.
Fee paid for every £1 of cost to serve
Clients under £150k pay £950 a year and cost £1,350 to serve. Clients over £1m pay £13,980 and cost £4,100. One is a viability question and the other is a fair-value one, and a board report that only shows 1.74× answers neither.
Ongoing fees with no review that year
£474k in total. The middle two segments carry two-thirds of it: fewer misses, larger fees. The redress-relevant slice, where no attempt was made, is £49k.
Where each £1 of fee income goes
Pay is 59p of every pound. That's why cost to serve is a time question first: the model needs to know where adviser and paraplanner hours went, not just what the levies cost.
Under the hood (for the data team)
The charts are the easy half. Getting a back office, four platforms and a set of practice accounts to agree on what a client, a service year and a fee are is the work.
Six tables and the service yearPlatform charges, versioned agreements, and why January looks overdue
Everything lands in one small warehouse before Power BI sees it. Whichever back office holds the client record — Intelliflo Office, CURO, Xplan or another — it ends up in the same six tables:
| Table | Grain | Why it matters |
|---|---|---|
| Fact Review Cycle | Client × service year | Due date, first offer, attempts, meeting held, record issued, outcome — the denominator for delivery |
| Fact Adviser Charge | Client × platform × month | The charge actually taken, from the platform feed, not the rate on the agreement |
| Fact Time | Staff member × client × activity × week | Adviser, paraplanner and admin hours, so cost to serve isn't a guess |
| Dim Client | Client, with segment and vulnerability history | Segments at the start of the service year, so clients don't jump bands mid-review |
| Dim Service Agreement | Agreement version | Fee basis, rate, review frequency and start date; versioned, because terms change |
| Dim Staff | Person, role, cost rate by month | A pay rise shouldn't rewrite last year's cost to serve |
The join that breaks first is time. Platforms take fees monthly; reviews are due on the anniversary of each agreement. Measure delivery on the calendar year and every client who signed in the autumn looks overdue in January. The service year has to come from the agreement, not the calendar.
The second is the fee itself. A client with three platform accounts generates three charge lines, and a household on a combined fee generates one for two people. Conform charges to the client before anything is divided by anything.
The measuresSample DAX for delivery rate, fees without a review, and fee-to-cost
Reviews Due =
CALCULATE ( COUNTROWS ( 'Review Cycle' ), 'Review Cycle'[Due Date] <= MAX ( 'Date'[Date] ) )
Reviews Delivered =
CALCULATE ( [Reviews Due], 'Review Cycle'[Outcome] = "Delivered",
NOT ISBLANK ( 'Review Cycle'[Record Issued Date] ) )
Review Delivery Rate =
DIVIDE ( [Reviews Delivered], [Reviews Due] )
Fees Without Review =
CALCULATE ( SUM ( 'Adviser Charge'[Amount] ),
'Review Cycle'[Outcome] IN { "Declined", "No response", "Not attempted" } )
Fee to Cost Ratio =
DIVIDE ( SUM ( 'Adviser Charge'[Amount] ), [Cost to Serve] )
The detail that decides whether a compliance officer trusts the number is in Reviews Delivered: a review counts only when a record has been issued, not when a meeting was booked. Count invitations as delivery and the rate will look better than any file review can confirm.
Keep "declined" and "no response" as separate outcomes in the source data even if they're reported together. A client who said no and a client who was emailed once are not the same evidence, and a look-back will ask which was which.
Already on a back-office data warehouse, or buying Consumer Duty MI?Where this adds something, and where it doesn't
The main back-office systems now offer data feeds, and some offer a warehouse that Power BI can read directly. There are also Consumer Duty MI products built around a single back office, and compliance consultancies with good board report templates. If your clients, reviews and fees all live in one system and your board pack already shows delivery by segment against a justified threshold, you have what this article describes, and we'd tell you so.
Where they stop is the edge of the system that owns them. The fee actually taken sits on the platforms. Cost sits in payroll and the practice accounts. Several years of history may sit in a back office you migrated away from. Delivery against fees, and fees against cost, needs all three. That join is what we build.
What we need, and what we don't
A compliance officer's first question is the right one, so here is the answer up front. This reporting needs operational metadata: client reference numbers, segment, service agreement terms, review dates and outcomes, charge amounts, hours by activity, and pay and overhead totals. It doesn't need names, addresses, fact-finds, suitability reports or the advice itself, and client identities can be pseudonymised before anything leaves your systems.
Contractually, DataHexis acts as a data processor under UK GDPR: a signed data processing agreement, a published list of sub-processors and UK-hosted storage. The reporting layer is read-only, so it can't move client money or place instructions. The firm keeps its own regulatory responsibility for anything it outsources, so the sensible order is your due diligence first and data second.
Sole advisers and two-adviser practices
The FCA's 2024 data counts 5,016 advice firms and 37,441 advisers — fewer than eight advisers per firm on average — so this is the size most firms actually are. A single adviser with a hundred clients doesn't need a warehouse. They need the same three columns that matter to a sixteen-adviser firm — review due, review delivered with a record, fee taken — kept in one place and looked at monthly. For many sole advisers that's a well-built spreadsheet fed by the back office, and we'd say so.
The cost-to-serve question matters more at that size, not less. NextWealth's 2026 benchmarks put the average minimum ongoing fee at £1,935 a year. Below that level, a client who takes a full review, a valuation pack and a couple of calls is often a client the adviser is paying to keep. That can be a perfectly good decision, as long as it's made on purpose and the board report — even a one-person one — says so.
Frequently asked questions
What did the FCA find in its review of ongoing advice services?
The FCA asked 22 of the largest advice firms for seven years of data and published its findings in February 2025. Promised reviews were delivered in around 83% of cases; in a further 15% clients declined or didn't respond; in fewer than 2% the firm hadn't attempted a review, and the FCA said redress is likely to be due there. It did not currently view the issue as systemic, but said firms carrying out proactive reviews should look back to 2018.
Do financial advisers still have to carry out an annual review?
In March 2026 the FCA consulted (CP26/10) on replacing the annual suitability review with periodic reviews based on clients' needs, and on clarifying how firms support disengaged clients. The consultation closed on 22 May 2026 and a policy statement will follow, so check the final rules before changing your service terms. Either way, firms will need to evidence that the reviews clients pay for are delivered.
What MI should an advice firm include in its Consumer Duty board report?
At minimum: review delivery against what was promised, split by client segment; outcomes for clients with characteristics of vulnerability, shown separately rather than aggregated; complaints and their root causes; and fees compared with the cost of providing the service. The FCA has criticised reports for weak data quality, thresholds without justification and MI that didn't lead to decisions, so state each threshold's reason and the action it triggers.
How do you calculate cost to serve for a financial advice client?
Add up the adviser, paraplanner and administration time each client's ongoing service consumes in a year, valued at cost rates rather than charge-out rates, then add the platform, software and regulatory costs that travel with each client. Compare it with the fee the client actually paid. Doing this by segment rather than for an average client shows which groups are subsidised and which pay well above cost. Only 40% of firms told NextWealth they had reviewed it in the past year.
How many clients should a financial adviser (advisor) look after?
NextWealth's 2026 benchmarks put the average at 88 clients per adviser, while Dynamic Planner's 2026 survey found 134, so treat any single ratio with care. The right number depends on what each client's service includes: a firm that delivers 83% of its reviews at 74 clients per adviser has a capacity or process problem, not just a ratio problem. Forecasting reviews falling due by month is the more useful measure.
Sources
- FCA, Ongoing financial advice services: multi-firm review, 24 February 2025 (22 firms; 83% delivered, 15% declined or no response, under 2% not attempted; 80% of adviser-charge revenue from around 4 million clients). fca.org.uk
- FCA, Portfolio letter to financial advisers and intermediaries, 7 October 2024 (4.4 million consumers paid for advice; 90% of new clients into ongoing advice; ongoing share of revenue 60% in 2016, 80% in 2023). fca.org.uk
- FCA, Consumer Duty board reports: good practice and areas for improvement, 11 December 2024, updated 9 March 2026. fca.org.uk
- FCA, Outcomes monitoring: good practice and areas for improvement, 27 July 2026 (including a survey of 56 firms). fca.org.uk
- FCA, CP26/10: Simplifying the pensions and investment advice rules, 25 March 2026 (closed 22 May 2026). fca.org.uk
- FCA, Evaluation of the impact of the RDR and FAMR, December 2020 (average ongoing charge 0.8% a year; initial 2.4%). fca.org.uk
- NextWealth, Financial Advice Business Benchmarks, reported by Professional Paraplanner, 10 September 2026 (318 respondents; 40% reviewed cost to serve; 44% turned off ongoing fees for some clients; average 88 clients per adviser; average minimum ongoing fee £1,935). professionalparaplanner.co.uk
- NextWealth fee benchmarking, reported by Professional Adviser, 25 March 2025 (average ongoing fee up 9 basis points since 2023; 70% of advisers charge on assets, 29% fixed fees). professionaladviser.com
- Dynamic Planner, Advice 2026, reported by Financial Planning Today, 21 May 2026 (530 advisers; 134 clients per adviser, up from 118). financialplanningtoday.co.uk
- FCA retail mediation data for 2024, reported by Financial Planning Today, 19 November 2025 (5,016 adviser firms; 37,441 advisers; £5.7bn revenue). financialplanningtoday.co.uk
Related reading
How wide is your receipt gap?
In a free 60-minute discovery call, we'll look at where your service agreements, review records, platform charges and practice costs live, and tell you what a delivery-and-value view by segment would take. Or start with our fixed-price Ongoing Service Health Check, from £750, credited in full against any build.
Book a free discovery call → Pseudonymised client data · Delivered by a partner · First dashboard usually within about two weeks