Contents
Key points
- Billed by the hour, faster work lowers the fee. Under a flat fee, it lowers the effort while the fee stays the same.
- Whether AI lowers your fees or raises your margin depends on where your hours sit today: billed by effort, inside a flat fee, or not billed at all.
- A flat fee only protects the margin when its scope is clearly defined and extra work is recorded separately.
- Blended models, value pricing for advisory work and passing savings on transparently are all options. None of them fits every service.
- Every new fee model starts with a measurement: how many hours a process takes today, and how many it takes with an agent.
The objection is arithmetically correct
Billing by effort means selling hours. When an agent does part of the work faster, there are fewer hours to sell. The arithmetic is short.
A worked example with round assumptions: a bookkeeping client is billed 40 hours a year by effort. 15 of those go into recording and reconciling receipts. If that work takes 5 hours with an agent and approval, you bill 30 hours instead of 40. At the same rate, the fee for this client falls by 25%.
The Swiss fiduciary association TREUHAND|SUISSE has published a post on alternative pricing models in the fiduciary sector on its blog, in German. It describes billing by the hour as dangerous above all for routine work that can be automated. So the objection has a real point. It applies, however, to only part of your hours.
Where your hours sit today decides the answer
Most fiduciary firms have three kinds of hours. Faster work affects each kind differently, in part in opposite directions.
| Where the hour sits today | What faster work does |
|---|---|
| Billed by effort | The fee falls, as long as rate and model stay the same |
| Inside a flat fee or under a fee cap | The fee stays, the effort falls, the margin rises |
| Not billed at all: internal work, rework, extra work that disappears into the flat fee | Costs fall away without any loss of fees |
One pilot shows how lopsided the split can be. At a fiduciary and property management firm, 96.2% of property management hours sat inside the flat fee; about 3% were billed separately. A single pilot, not an average and not a promise. For a firm like that, faster property management work would have had almost one consequence only: a higher margin.
You can sort a year of your own time-tracking export into these three buckets: hours per service type and, beside them, how they were billed. The third bucket is the hardest to find, because it appears on no invoice. How to make it visible for one client is shown in Flat fee or extra work.
Four fee models for when work gets faster
No formula from outside fits your prices. There are, however, four models, each with a different answer to the question of who earns from the time saved.
| Model | Who earns from the time saved | What it requires |
|---|---|---|
| Flat fee with a defined scope | The firm, as long as the scope holds | A list of services in the contract and time tracking that separates extra work from flat-fee work |
| Blended model | The firm for routine work, both sides for advice | A clear line between routine and advice, written into the contract |
| Value pricing for advice | The firm, if the benefit to the client is agreed in advance | A defined result and experience with similar cases |
| Transparent pass-through | The client directly, the firm through freed-up capacity | A recalculated rate and demand for the hours that become free |
Flat fee. It suits routine work such as bookkeeping, payroll or VAT returns. Its weak point is scope: if the contract only says “bookkeeping”, all extra work lands in the flat fee and the time saved is lost there again. It protects the margin only once the contract lists the services individually and names extra work with its own rate.
Blended model. A base flat fee covers the routine; advice and anything out of the ordinary is billed by effort or as a package. The association's post gives bookkeeping priced by the number of entries as an example, kept separate from personal advice based on the digitised figures.
Value pricing for advice. For a succession plan, a restructuring or a company valuation, the client cares about the result rather than the hours. Pricing by significance requires the result and the scope to be described in writing beforehand. For routine work it hardly fits.
Transparent pass-through. You keep the hourly model, bill fewer hours and tell the client so. The fee per client falls, the benefit is visible on the invoice, and your gain lies in capacity for more clients. The association's post notes that an hourly rate then has to cover the “machine hours” as well as staff costs.
What the discussion in the association shows
The TREUHAND|SUISSE post is based on a bachelor's thesis at the FHNW School of Business, with a survey of member firms and expert interviews. The response rate was low, and the authors themselves state that it allows no statistically reliable conclusions. Three observations are still useful for the question of AI and fees.
- Flat fees often bill planned rather than actual hours, so they are not a fundamentally different model. As we read it, a firm that measures its flat fee against actual hours at every renewal passes efficiency on to the client over time after all.
- Respondents name two problems with flat fees: defining the services clearly, and uncertainty when estimating the effort. Both shrink once you know your own hours per service.
- Standardisable services such as payroll, property management and bookkeeping are seen as suitable for flat fees and packages. For business and wealth advisory work most respondents see it differently, although the authors consider packages possible there too.
The post also addresses culture: an industry that bills hours and pays for hours gets the work done in whatever time the budget allows. A firm that changes its fee model has to talk about incentives in the team too, for example whether targets still reward billed hours.
Where the time saved can go
Time saved without a plan spreads across the rest of the week, and by year end it shows up in no figure. Three uses are obvious for fiduciary firms.
- Advisory work. Conversations about succession, tax planning or liquidity for which day-to-day work leaves no time today. This work can be billed by effort or by value, and no agent takes it off your hands.
- More clients with the same team. When positions are hard to fill, freed-up capacity is often a bigger lever than a higher rate.
- Seasonal peaks. Tax returns, year-end accounts and VAT returns bunch into a few months. Every hour of routine an agent prepares in that period relieves the team exactly when overtime would otherwise be the norm.
Which use fits is the owner's decision. It gets easier once it is clear how many hours actually become free, and in which process. Where an agent comes in on top of your accounting software, without a change of system, is described in AI in Bexio and Abacus.
Measure first, then choose the model
Building a new fee model on estimated savings is riskier than keeping the old one. That is why the measurement comes first.
- 01Sort the hoursSort a year of time-tracking exports into the three buckets: billed by effort, inside the flat fee, not billed.
- 02Choose one processThe routine process with the most hours, for example recording receipts or reconciling VAT.
- 03MeasureAn agent handles this process for a limited number of clients, with approval and a replay log. Hours are counted before and after, in your own hours.
- 04DecideYou decide for each service which fee model applies. The basis is measured hours, not an estimate from a presentation.
- 05AgreeThe model is discussed with the client, ideally at the next contract renewal and with a list of services in the annex.
That way you go into the conversation with clients and team with figures. At Grow Amplified this first step is called the audit: one week, a process map and the cost of each process in your own hours at your internal rate. What follows is described under Approach. If you would like to check for yourself first how much extra work sits inside one flat fee, use the free Flat-Fee Leak Check.
Frequently asked questions
Will I earn less with AI if I bill by the hour?
For the hours you bill by effort today, yes, as long as rate and model stay the same. For hours inside flat fees and for work you do not bill at all, only the costs fall. Your own time-tracking export shows which share is larger.
Can I raise my hourly rate if AI makes me faster?
In Switzerland, fiduciary fees are a matter of negotiation, as the association's post also points out. A higher rate can be justified if it includes tools and review and the client does not pay more overall for the same service. What matters is that you can show the calculation behind it.
Do I have to tell clients that I use AI?
As we understand it, that depends on how you use it. A TREUHAND|SUISSE post distinguishes between simple support and AI that influences a professional service or decision, and recommends a tiered approach, for example in the terms of engagement. Clarify the details with your legal adviser.
Which services suit a flat fee?
Services that recur and whose scope can be described, such as bookkeeping, payroll, VAT returns or property management. The precondition is a list of services in the contract and time tracking that records extra work separately.
When should I change my fee model?
Only once you have measured. A sensible moment is the next contract renewal, service by service, with a list of services in the annex. Switching every client at once is rarely necessary.
Does Grow Amplified recommend particular prices?
No. Grow Amplified measures how many hours a process takes today and how many it takes with an agent. Which fee model you derive from that is for you and your clients to decide.
15 years in banking and consulting, in transformation and finance operations. Today he builds review lanes with Grow Amplified: agents read and reconcile, a person approves, and a replay log records who decided what.