firm-fee-structure
Compares fixed fee, time and materials, retainer, value-based and fixed-term handover for one piece of work, with what the client buys written as results, what each model rewards, who carries an overr
SKILL.mdpacks/consultants-pack/skills/firm-fee-structure/SKILL.md
GitHubFee Structure Choice
When To Use
Your day rate was easy to defend when a market review took weeks. Now parts of it take days, the client knows it, and a fee built on hours shrinks every time you get better at your job. Run this before the proposal, to pick the charging model that fits this piece of work and to say why in terms of what the client receives.
When Not To Use
If you still need to shape what is on offer, start with Good-Better-Best Options. If a fee is already quoted and the client is pushing on it, use Negotiation Plan. This skill never sets or lowers a number.
Inputs
- The options or scope being priced, with deliverables and rough phases
- How you charge today, your rates and past fees
- Any value the client put on the problem, in their figures, with who said it If you have none of this, I start from one line describing the work and mark the comparison as a first draft.
Approach
Comparing fee models is a practitioner convention. The writing rule is borrowed from the US federal acquisition rule FAR 37.602, which asks for work to be described by required results and measurable standards and not by hours. That is also the straight answer to the AI question: when AI speeds you up, an hourly fee falls while the result is worth the same to the client. The failure it prevents is announcing "value pricing" with no value figure from the client, which is a guess under a better name.
Workflow
- Ask, in one go: what exactly does the client hold at the end? How much of this work does AI already make faster for you? Has the client said, in their own figures, what the problem costs or what fixing it is worth?
- Write what the client buys as a result with a standard they can check, in the manner of FAR 37.602: "an operating model the leadership team has approved", not "[n] days of analysis". Each model is judged against that line.
- Compare the five models on four counts: what the client pays for, what it rewards, who carries an overrun, what AI speed does to it. Time and materials falls as you get faster. A fixed fee leaves you the gain and the overrun. A retainer pays for access, which AI does not shrink. Value-based ties to their result. A fixed-term handover pays for building their capability, including their own use of AI.
- Test value-based honestly. With the client's own value figures, show the link. Without them, mark it "not possible yet: ask for [figure]" and add the question to your next call. What the work is worth to them is not mine to estimate.
- For the model you lean towards, name the risk it puts on you with this client and the protection it needs: written assumptions, change control, a review point, a cap.
- You choose the model, or a mix by phase, and I write the reason in two sentences the client could read, plus your answer to "why pay for hours?". Tax, contract and payment terms: check with a qualified adviser.
Output Format
# Fee Model Comparison
What the client buys (result and standard): [one line]
| Model | Client pays for | Rewards | Overrun carried by | What AI speed does to it | Fit here |
|---|---|---|---|---|---|
| Fixed fee | [ ] | [ ] | [ ] | [ ] | [fits / partly / does not] |
| Time and materials | [ ] | [ ] | [ ] | [ ] | [ ] |
| Retainer | [ ] | [ ] | [ ] | [ ] | [ ] |
| Value-based | [ ] | [ ] | [ ] | [ ] | [possible / not possible yet: ask for [figure]] |
| Fixed-term handover | [ ] | [ ] | [ ] | [ ] | [ ] |
## Protections for the chosen model
| Model | Risk to you | Protection |
|---|---|---|
| [model] | [risk] | [assumption / change control / review point / cap] |
## Your answer to "why pay for hours?"
[Two or three sentences in your voice, built on the result line above]
## Decision
[Your name] chooses the model and sets every figure by [date]. Tax, contract and payment terms: check with a qualified adviser.
Done When
- What the client buys is written as a result with a checkable standard
- The five models share the same four columns of comparison
- Value-based shows the client's own figures or reads "not possible yet"
- The model you chose is written down with its reason and its protections
Quality Bar
- The comparison is about what the client receives, never about how busy you are
- AI speed is stated plainly, where it helps you and where it hurts
- No market rate, benchmark or "typical" fee appears
- Your own use of AI is never hidden in the answer on hours
- No price recommendation; every rate and fee is yours, I invent no figure for value, and you choose the model.
Next
Run firm-moscow-scope (MoSCoW Scope) to fix exactly what the fee covers.
About the makers
This pack is made by Polar Bear, a consultancy built by ex-McKinsey founders with a dream to make AI work for People, not instead of them. We help our clients build people systems and AI-first ways of working, and we run our own company on Claude. If your team has outgrown the self-serve version, message Pauline (linkedin.com/in/paulinebertry).