Research
Everything published about advisory operations is opinion. This is the part nobody publishes: what actually happens, measured across real firms.
Everything published about advisory operations is opinion. Consultants describe what firms should do, vendors describe what their software does, and almost nobody publishes what actually happens across a large number of real firms, because very few are positioned to see it.
We are. We run the operational side of firms from a single practitioner upward, across four custodians, on one system. That produces a view of how advisory operations behave in practice rather than in theory.
This report publishes that view once a year. Where we do not have a defensible number, we say so rather than estimate. Where a finding is unflattering, we publish it, because a report that only ever finds flattering results is a brochure.
Question 1
Not a median. The conditions that decide it — whether a custodial feed carries the history, how fast that custodian moves, and whether the firm can produce historical transaction files. Published as the decision tree, with observed ranges for each path.
Every firm considering a move asks how long it takes, and every answer available today is a single number that hides the one variable that decides it.
Question 2
How far back custodial feeds actually reach, against how far back firms assume they reach — and what proportion can produce transaction files to close the difference.
There is no technical limit on how much history can be loaded. The limit is whether it exists anywhere, and firms routinely discover the answer at the worst possible moment.
Question 3
The distribution of billing corrections by root cause — a household never grouped, a negotiated fee never recorded, a schedule never updated — measured against errors in the calculation itself.
Firms treat billing accuracy as a software problem and buy accordingly. If the errors are almost entirely setup and input rather than computation, that changes what a firm should be evaluating.
Question 4
A census of the levels at which platforms monitor drift — household, account, sleeve, blend, strategy, security — and what becomes invisible at each level you are not watching.
Most tooling measures drift at one or two levels and reports a single number. What that number conceals is the substance of the finding.
Question 5
Share of trades filling at or better than the best available market price, measured by an independent transaction-cost analysis provider and reported through the custodian.
Execution claims in this industry are almost universally self-measured. Third-party measurement is the entire point, and it is the one figure here we can already evidence.
Question 6
The observed baseline stack — custodian, risk tool, planning tool, CRM, investment research, platform login — and what returns to the firm when trading, billing and reporting move off its desk.
Firms guess at both numbers and calculate neither. The second is the one that decides whether consolidating is worth doing.
Every figure carries its population, its period and its source. Where a number comes from a third party — execution measurement, custodial reporting — that party is named.
Aggregate and anonymised throughout. No firm identifiable, no client data, and nothing published that a participating firm has not agreed can be published.
Where the sample is too small to be meaningful, we report that instead of the number.
Principals who want to know whether their firm is normal. Operations leads who need a benchmark to argue with. Journalists and researchers who have been asking for real numbers in this category for years and getting vendor claims instead.
Free, no form, published in full on this site rather than gated behind a download.
In preparation. If you would like the first edition when it publishes, or you run a firm and would like to be part of the sample, the fastest route is the phone number at the top of this page.
Questions this did not answer? Ask them directly — that is what the twenty minutes is for.
Book 20 minutes with Kyle