Disclosure: This article may contain affiliate links. We earn a commission at no extra cost to you if you purchase through our links. We only recommend tools we genuinely believe in.
A client once asked me for a price on a project, and I gave it to him in the same message, in about ninety seconds. I had not checked my calendar, I had not written down what was included, and I had not asked what the work was actually worth to him. He said yes immediately, which felt like a win, and then I spent the next six weeks doing the work for roughly half of what it should have paid, because the price I had blurted out became the price, and the scope I had not written down became whatever he asked for. That was not a pricing mistake. That was a pricing system that did not exist.
This post is the sequel to How I Handle Scope Creep Without Losing the Client (or the Margin). Scope creep is what happens after a bad price. This is what happens before one: how I price a project now, with AI doing the arithmetic and me doing the judgment. It also builds on How I Run a Monthly Retainer Review With AI (and Keep Clients Renewing), because the retainer review is where a bad project price shows up as a margin problem, one month later.
This is a playbook, not a results report. The figures below are a starting template to tune after your first few projects, not measured outcomes. Your rates will differ because your costs, your market, and your clients differ.
The Problem: The Price Gets Set Before the Work Is Understood
Most bad prices are set in the same ninety seconds mine was. Three habits cause almost all of them:
- Pricing from the gut. You pick a number that feels right, which usually means a number that feels safe, which usually means a number that is too low. The gut knows what the client will accept. It does not know what the work costs.
- Pricing from the last quote. The last project was smaller, or simpler, or for a friend, and its price becomes the anchor for this one, even though nothing about the two projects matches. A price inherited from a different job is a guess with a history.
- Pricing from the fear of losing the client. You shade the number down because you want the yes, and you tell yourself you will make it up on the next project. You will not. The next project gets priced from this one, and the discount compounds.
The fix is not confidence. The fix is a worksheet: five numbers, written down, before any quote goes out. A price you can show your working for is a price you can defend, and a price you can defend is a price the client respects, because it reads as a calculation, not a mood.
The System in One Pass
The stack is deliberately boring, and it reuses the tools from the scope creep and retainer review posts. No new platform is required to start, and no part of this system talks to the client on its own.
| Stage | Tool | What happens |
|---|---|---|
| Capture | Your inbox or call notes, plus the pricing worksheet | The scope gets written down as a list of deliverables, with what is excluded |
| Estimate | Claude or ChatGPT, with the scope list and your rate card attached | A first-pass estimate: hours per deliverable, plus the lines you forgot |
| Compare | The worksheet, with the three models below | One model picked per project, with the reason written next to it |
| Check | You, with your calendar and your costs open | You verify the hours, the rate, and the risk margin, and you rewrite the quote in your own words |
| Send | Email, in writing, with the scope attached | The client gets one price, one scope, and one date, before any work starts |
Notice what automation does not do in this table: it does not set the price, it does not choose the model, and it does not send the quote. It does the arithmetic from the scope you wrote. You decide, and you send, in your own words.

Three Pricing Models, Compared
Every project gets one of three models, picked before the estimate, not after. The model is part of the quote, because each one changes what the client is buying.
| Model | When it fits | The risk if you pick it wrong |
|---|---|---|
| Hourly | The scope is genuinely unknown, or the client wants open-ended help | The client watches the clock, and every hour needs justifying |
| Fixed project fee | The scope is written down and the deliverables are clear | You absorb the overrun if the scope was wrong, so the worksheet has to be honest |
| Value-based | The work has a clear, measurable payoff for the client, and you can name it | You need the client's numbers to price it, and a wrong value guess becomes an argument |
Most of my projects land on the fixed fee, because a written scope plus a fixed price is the easiest quote for a client to say yes to, and the easiest one for me to defend. Hourly is for discovery and small, undefined work. Value-based is for the rare project where the payoff is obvious and the client will share the numbers that prove it. If you cannot name the value, do not price on value. Price on the work.
The Pricing Worksheet: Five Numbers Before Any Quote
The worksheet is one page, and it takes about twenty minutes once it exists, as a planning figure. Five lines, in this order, every time.
| Line | What goes in it | Example, as a planning figure |
|---|---|---|
| Hours, honestly | Your best estimate per deliverable, then add 20 percent for the parts you always underestimate | 40 hours of build and review, plus the buffer |
| Rate floor | The lowest hourly rate the work is worth to you, from your real costs, not from the last quote | A $150 per hour floor, from costs and target margin |
| Scope list | Every deliverable, and the two or three things that are explicitly not included | Three landing pages, two rounds of revision, no ongoing ad management |
| Risk margin | A percentage for the unknowns: a new client, a new channel, a tight deadline | 10 percent for a familiar client, 20 percent for a new one |
| Value check | One line on what the work is worth to the client, if you know it | A launch that has to land before a seasonal deadline |
The rate floor is the line people skip, and it is the line that matters most. Work out what an hour has to earn for the business to work: your costs, your tax, the time you do not bill, and the margin you need to keep going. That number is the floor. Any price below it is a donation, however good the client is, and donations belong in a different budget from the one that pays the rent.
Draft the estimate with AI from the scope list if you like. Give the model the deliverables, the rate floor, and the last three similar quotes, and ask it to flag the line items you forgot. It is good at that. It will find the revision round you left out, the reporting you assumed was free, and the handover call that always takes an hour. What it cannot do is decide whether the price is right for this client, this month, at this risk. That is your line.
What Breaks, and the Guardrail for Each
| Failure | Warning sign | Guardrail |
|---|---|---|
| The quote goes out in ninety seconds | The price was set before the scope was written | No quote without the worksheet, no exceptions, even for a friendly client |
| The scope is never written down | The client asks for one more page, and you cannot say whether it is included | The scope list travels with the quote, including what is not included |
| The rate floor is a feeling | Two similar projects get two different prices, and you cannot explain why | The floor comes from costs, written on the worksheet, and it only changes quarterly |
| The risk margin is always zero | Every project runs over, and every overrun comes out of your margin | A new client or a tight deadline always carries a margin, even a small one |
| The quote is negotiated down on the call | You agree a lower price with nothing removed, because the silence felt awkward | A lower price removes a deliverable. The trade is written down, before you say yes |
| The price is never reviewed | You are busy, the clients are happy, and the margin is quietly shrinking | The retainer review checks project margins too, one line per project, every month |
Rule: a price you cannot explain in one sentence is a price you guessed. Write the scope, price the scope, and send the quote in writing, before you start.
Lessons at the End
- The price is set before the work is understood, or it is not set at all. Ninety seconds and a gut feeling is not a pricing method. Five numbers on a worksheet is.
- The rate floor comes from costs, not from the last quote. A price inherited from a different job is a guess with a history.
- Fixed fee fits most projects, hourly fits the unknown, value-based fits the provable. If you cannot name the value, price on the work.
- The scope list travels with the quote. Including what is not included is what stops the next scope creep post from being about you.
- A lower price removes a deliverable. Never discount the number without changing the scope, and write the trade down before you say yes.
- AI does the arithmetic, you do the judgment. It finds the forgotten line items. It cannot decide what an hour of your work is worth.
If you build only one thing from this post, build the rate floor and the worksheet: the lowest hourly rate the work is worth to you, from your real costs, and five lines filled in before any quote goes out. The worksheet takes about twenty minutes once it exists, as a planning figure. It does not stop clients from negotiating. It stops you from negotiating against yourself.
