Automate Advisory Fee Reconciliation

Why advisory fee reconciliation still creates risk

Advisory fee billing should be routine. The firm has a signed agreement, a fee schedule, household records, custodian data, and a quarterly or monthly billing cycle. Yet the reconciliation work behind that process is rarely routine.

Someone exports custodial billing files. Someone else pulls data from the CRM, portfolio management system, billing platform, spreadsheets, and occasionally a folder full of fee agreements. They compare records line by line. Exceptions get marked up in email, a workbook, or a notes field that only one person understands.

Then the billing period closes and the team starts again.

For an owner or partner, the problem isn’t just the hours spent matching data. It is the exposure created when household relationships are missed, negotiated rates are applied incorrectly, accounts are excluded, assets are double-counted, or a fee override is never carried forward into the billing system.

A small mistake can be fixed. Repeated exceptions, late reviews, and unclear audit trails create a process problem.

Across financial advisory firms doing $1M to $25M in annual revenue, we usually see operational leakage in the $70K to $200K annual range. That figure can include unbilled work, billing corrections, avoidable write-offs, compliance effort, and senior staff time absorbed by manual controls. Fee reconciliation is often one of the clearest places to find it because the process touches revenue, client trust, and compliance at once.

The aim isn’t to hand fee decisions to an AI model. The aim is to automate the comparison work, document the exceptions, and give a responsible person a clean review pack before bills are finalised.

You can see the broader operating model in Omni for financial advisory firms. The practical opportunity starts with mapping how your team reconciles fees now.

What manual fee reconciliation looks like in practice

Most firms don’t have one fee reconciliation process. They have several versions built around custodians, adviser preferences, legacy client arrangements, and software limitations.

A billing administrator may receive a custodial file containing account numbers, market values, billing dates, fee debits, and transaction descriptions. The file is loaded into a spreadsheet or billing system. Then begins the manual work:

• Matching custodian account numbers to client and household records
• Confirming the account is assigned to the correct billing group
• Comparing the billed rate against the signed fee arrangement
• Checking breakpoint tiers and minimum fees
• Identifying accounts that should be aggregated for household pricing
• Reviewing cash, managed accounts, held-away assets, or excluded assets
• Checking for account closures, transfers, or recent onboarding
• Investigating zero-fee, negative-fee, duplicate, or unusually high charges
• Preparing an exception list for adviser or operations approval
• Saving evidence of what was reviewed and why a change was made

The work becomes harder when the source records disagree. The CRM might show a household with four accounts. The custodian file might list five, including a new account with no household ID. The fee agreement may state 90 basis points on the first $1 million and 70 basis points thereafter. The billing platform may contain a historical 80 basis point override because a former adviser made an arrangement years ago.

A person can work that out. The issue is that the person has to find it first.

Many firms rely on experienced operations staff to spot the odd records because they know the clients and remember past billing issues. That expertise is valuable, but it doesn’t scale well. It also creates a key-person dependency every time a billing cycle arrives.

The same team may already be under pressure from advice documentation and file notes. The Advice Document Agent can help prepare SOAs, ROAs, and notes from approved source material, which frees paraplanners from repetitive drafting. That capacity matters when billing exceptions need proper review rather than a rushed check at month-end.

The right role for AI in billing reconciliation

An AI-assisted fee reconciliation process works best when it has clear boundaries.

It should collect and structure information. It should match records across systems. It should calculate expected amounts using approved fee rules. It should identify anomalies and missing data. It should produce a review summary with source references.

It should not independently approve a fee change, amend a client agreement, initiate a fee debit, or decide how a regulatory obligation applies. Those actions remain with authorised people under your firm’s controls.

Think of the agent as a billing operations analyst that never gets tired of comparing records. It prepares the work. Your billing owner, adviser, or compliance lead makes the decision on exceptions.

This is how an AI agent can handle the process end to end.

1. Collect billing-cycle inputs

The workflow starts with defined inputs, not a vague instruction to “check the fees.”

Typical inputs include:

• Custodial billing files for the relevant period
• Billing platform export
• CRM household and account data
• Portfolio values and billing-date balances
• Current fee schedules and approved client-specific overrides
• Signed advisory agreements or structured agreement extracts
• Prior-cycle exception log
• Account opening, closure, transfer, and death-notification records
• Any approved exclusion rules for accounts or assets

The agent checks file completeness before any matching begins. If a custodian file has 2,340 records but the expected account count is 2,410, it flags the gap. If a file is stale, has an unexpected date range, or lacks a required column, the workflow stops and requests a replacement.

That sounds basic. It is exactly the kind of basic control that gets skipped when staff are trying to finish a billing run before a deadline.

2. Create a common account and household view

Custodians, CRMs, and billing systems often use different identifiers. A clean workflow builds a controlled matching layer.

The agent first matches accounts using exact account IDs. It then applies approved secondary rules, such as matching a masked account number, account title, tax ID suffix, or known cross-reference table. Anything below the confidence threshold stays unmatched. It doesn’t guess.

Next, it rolls accounts to the household level based on the CRM’s approved household structure. This is where billing errors often hide.

For example, a client might have an individual account, joint account, trust, IRA, and a family entity. Their agreement may state that assets are aggregated for tiered pricing. If two accounts sit outside the household in the CRM, the calculated fee may be too high or too low.

The reconciliation agent produces three lists:

1. Accounts matched cleanly to a client and household
2. Accounts matched with a review note
3. Accounts that could not be matched and need human action

That separation saves time. Staff don’t need to manually inspect every record. They start with the records that actually need judgement.

How the agent checks fee calculations

The next step is not simply comparing one fee figure to another. It is calculating the expected fee from the rules the firm has approved.

For each account or household, the agent applies the relevant billing method. That might be:

• A percentage of assets under management
• Tiered household pricing
• A fixed annual or quarterly fee
• A minimum fee
• A blended rate
• A negotiated client rate
• A fee waiver for a defined period
• A prorated charge for a newly funded account
• An exclusion for certain asset classes or managed sleeves

The calculation needs an effective date. It also needs a clear source for the rule. A client agreement signed three years ago may differ from the standard schedule now shown in the billing platform.

The agent should retain both values: the fee billed by the custodian or billing system, and the expected fee under the approved record. It then calculates the variance in dollars and basis points.

A useful exception policy might flag:

• Any missing fee agreement or missing rate source
• Any account with no linked household
• Variances above a defined dollar threshold
• Rate differences above a defined basis-point threshold
• Fee changes from the prior cycle outside expected movements
• Households with accounts on conflicting fee schedules
• Accounts billed at zero where no waiver exists
• Accounts with an active fee agreement but no fee charge
• Newly opened accounts without a billing treatment
• Closed or transferred accounts that were still billed

The thresholds should be set by your firm. A $15 variance may not deserve the same attention as a $500 variance, though a small difference can matter if it points to a systemic rule issue.

This is where a well-designed workflow reduces errors without pretending every exception is an error. A variance may be legitimate. It just needs a documented explanation and approval.

Preparing the reconciliation summary people will use

The final output should not be a 20-tab workbook that senior people won’t read.

The agent prepares a billing-cycle reconciliation pack with enough detail for operations and enough clarity for a partner or compliance reviewer. It can include:

• Total accounts and households reviewed
• Total fees billed and total fees expected under the approved schedule
• Number and value of exceptions by category
• Accounts missing household assignment
• Households with rate or tier conflicts
• New, closed, and transferred accounts requiring review
• Exceptions carried forward from the prior period
• Proposed actions, with a named owner and due date
• Evidence links back to the source file, agreement, or CRM record
• A final approval log

A partner may only need a one-page summary and the high-value exceptions. The billing team needs the account-level detail. Both can come from the same workflow.

The important point is traceability. If someone asks six months later why a household was billed at 65 basis points, the team should be able to find the agreement reference, the calculation, the reviewer, the approval date, and any associated correspondence without searching old inboxes.

This is also where Omni apps can be useful. A simple internal review app can present exceptions by adviser, household, custodian, or status. It gives reviewers a controlled place to approve, reject, assign, and comment rather than circulating multiple spreadsheet versions.

Where the process still needs human review

AI-assisted reconciliation improves the preparation work. It does not remove accountability.

Your firm should define who can approve each type of decision. For example, a billing operations manager may resolve an account mapping issue. A fee rate exception may require the responsible adviser and a compliance or operations approver. A change to a client fee arrangement may need documented client authority before it reaches the billing platform.

Human review is particularly important for:

• Non-standard client agreements
• Discretionary fee discounts
• Fee refunds or credits
• Estate and trust account changes
• Vulnerable client situations
• Complaints or disputed charges
• Ambiguous household aggregation rules
• Significant variances that may affect multiple clients
• Any exception where source documents conflict

The workflow should make review easier, not hide decisions behind automation. Every unresolved item needs a clear owner. Every approved adjustment needs a reason code. Every fee cycle needs a final sign-off before submission.

That control structure is useful beyond billing. The Client Onboarding Agent can collect KYC documents, guide fact-finding, and prepare an adviser-ready onboarding pack. If the onboarding workflow captures household structure and agreed fee terms correctly from day one, there are fewer exceptions to untangle later.

A practical rollout for an advisory firm

Don’t start by trying to automate every custodian, billing rule, and historical exception in one project.

Start with one billing cycle, one custodian feed, and a defined client segment. A sensible first phase usually looks like this:

1. Document the current billing process from file receipt through final approval.
2. Identify the systems of record for household, fee agreement, account status, and billed amount.
3. List your current exception types and estimate how long each takes to resolve.
4. Set matching rules, variance thresholds, and escalation rules.
5. Run the AI-assisted process in parallel with the existing manual process for one or two cycles.
6. Compare results, correct data issues, and tune exception categories.
7. Expand to other custodians or more complex fee arrangements once the first workflow is trusted.

A good pilot does more than prove the technology works. It exposes weak data ownership. For example, you may find that household relationships are inconsistently maintained in the CRM, or that fee overrides lack expiry dates. Those are operational issues worth fixing even before automation.

If you want an outside view of the workflow, Book a call with Sam. We use the session to identify where the billing team spends time, what data is available, and where controls need to stay human-led.

The wider operational payoff

Fee reconciliation is a focused use case, but it often reveals the same pattern across the firm. Data exists, but it is spread across systems. Skilled people spend time assembling it. Reviews happen late because the preparation work takes too long.

The Meeting Prep Agent solves a related issue by pulling portfolio data, recent communications, and goal progress into a one-page meeting brief. Advisers can spend less time hunting for context before a review and more time with the client.

When these workflows are connected, the firm has a cleaner operating rhythm:

• Onboarding captures the right household and fee data
• Client records are maintained through structured workflows
• Billing reconciliation checks the records before charges go out
• Advisers receive clearer information before client conversations
• Documentation and approvals are stored in a usable audit trail

That doesn’t require replacing your CRM, custodian, or billing platform. It requires deciding which system owns which data, then building a reliable workflow around the handoffs.

For more examples of where firms are applying this approach, browse the Omni operations resources. The useful ideas are rarely the flashy ones. They are the repeatable processes that give experienced staff fewer files to chase and fewer exceptions to explain.

Turn billing reconciliation into a controlled workflow

If fee reconciliation still depends on a spreadsheet, email threads, and one person who knows where the exceptions live, you have a clear automation candidate.

Start by measuring three things in your next billing cycle: the hours spent preparing files, the number of exceptions requiring review, and the value of corrections or missed charges. Those numbers give you a realistic view of the opportunity.

An Omni Audit takes 60 minutes and produces three practical outputs: a map of the current workflow, a shortlist of high-value automation opportunities, and a recommended first build. No slide deck. No generic transformation plan.

See the AI audit for financial advisory firms for the wider framework, or Book a call with Sam if you want to work through your billing process directly.