Freelance pricing: should you charge less now that AI made you faster?

The real question AI puts to a freelance designer is not “will it take my job?”. It is far more uncomfortable than that: if the work you used to do in two days now takes two hours, can you still ask the client for two days’ money? If something in you says no, the problem is not the AI — it is your pricing model.

Hourly pricing eats itself

Here is what happens to everyone who sells hours: the better you get, the faster you are; the faster you are, the less you earn for the same job. Your mastery punishes you. That contradiction existed before AI, it just worked slowly — an experienced designer was maybe twice as fast as a beginner, and that was that.

AI took that multiplier from two to five, sometimes ten. The contradiction is no longer slow; it is in your face. If you bill hourly, every efficiency gain is a direct loss of income.

What is the client actually buying?

When a hairdresser has you build a site, what they are buying is not “sixteen hours of design labour”. It is: being findable in search, getting booking calls, not looking amateur next to the salon down the road. That value does not change according to how many hours it took you. What helps their business is identical; only your cost went down.

You are not obliged to hand the efficiency gain straight to the client. A carpenter does not make the table cheaper because they bought a better saw.

Three models, three outcomes

Hourly

The most common, the most transparent, and the most damaging in an AI era. Its transparency is deceptive too: selling hours gives the client the right to inspect your inefficiency. Every minute you spend in a “why did that take six hours?” conversation is a minute spent discussing cost rather than value.

Fixed project fee

Better. You define the scope, you state the price, and how long it takes is your business. Getting faster increases your profit — the incentive points the right way. The risk is scope creep: “let us just add this” dissolves the margin, so what is included has to be in writing.

Value-based

The hardest and the healthiest. You tie the price to what the work is worth to the client rather than to what it costs you. A site that brings three new customers a month is a meaningful annual revenue line for that business, and the price is discussed in that frame. It requires understanding the client’s business; not everyone can do it, and the few who can earn the best money.

Sell hours and getting faster makes you poorer. Sell outcomes and getting faster sets you free. AI made the difference impossible to ignore.

Price is a function of value, not of your cost

The most common freelance mistake is calculating price from the inside: how long did it take, what is my hourly rate, multiply. By definition that ties your ceiling to your cost — as efficiency rises, income falls.

Calculating from the outside looks like this: what does this site earn the client? If a towing company gains three new customers a month and the average revenue per customer is known, the annual value of the site is a computable number. If your price is a small percentage of that number, the conversation moves off the “expensive” axis and onto the “sensible” one.

To do that you have to ask about their business — and this is exactly the step most designers skip. A conversation that starts with “how many pages?” ends in a price negotiation. One that starts with “how many customers do you get a month, and where do they come from?” goes somewhere else entirely.

When they say “AI does it, why so much?”

You will hear it eventually. Do not get defensive; defensiveness reads as weakness. Say the true thing: AI speeds up production, not judgement. Which section goes at the top, which sentence earns trust, what question the visitor arrived with — that is your work, and the client is coming to you precisely because they do not know those things.

Have the counter-question ready too: “You could use the same tool — there is a free tier. Try it for a week, and if it does not work out we can talk.” Saying that takes nerve, but it ends the argument, because the client is not hiring the tool. They are hiring you.

How to move to fixed pricing

“Quote a fixed price” is easy advice and hard to act on — where does the number come from? The practical route: write down the real duration of your last five projects. Take the average, multiply by your target hourly rate, then add twenty per cent for scope creep. That is your starting price.

Then the important part: when you say that number to the client, never use the word hour. Not “sixteen hours at this rate”, but “a five-page business site, two rounds of revisions, this price”. Same number, completely different conversation.

Your estimate will be wrong on the first few projects. That is normal — but unlike hourly, the mistake teaches you something: next time you estimate better and the margin stays with you. Under hourly billing there was no reward for getting faster. Here there is.

When should you say the price?

A small and expensive detail: quoting in the first message. When someone writes “how much?” and you answer with a number immediately, that number floats without any context and becomes the only thing to compare.

Ask first: how many customers do you get, where do they come from, what do you expect the site to do, when do you need it? Those four questions both get you to the right price and give the client the feeling that you understand their business. A price said after that feeling lands very differently.

Build packages

Offering three options rather than one number turns the discussion from “yes or no” into “which one”. Classic, and it works:

• Basic: one page, content supplied by the client, one round of revisions — for people who decide fast
• Standard: multiple pages, copywriting included, two rounds, domain connection — most clients pick this
• Complete: standard plus business profile setup and three months of maintenance — the most profitable one

Feature the middle package; people avoid the extremes and choose the middle. And keep the top tier listed even if you do not expect to sell it — its presence is what makes the middle look reasonable.

Losing the cheap client is good news

The biggest obstacle to raising prices is fear: what if they leave? Some will — and the ones who leave are usually the segment that takes the most time, negotiates the hardest and pays the latest. Losing them is not damage, it is housekeeping.

Do a simple exercise: list last year’s clients by revenue, then write next to each the hours you spent. Most freelancers seeing that table for the first time notice the same thing — the lowest-paying clients consume the most time. So they are not just earning you little; they are eating the time you would have given to good clients.

What will you do with the freed-up time?

This is where the real opportunity is. If building sites with AI saved you two days on a project, you invest those two days either in a new client or in producing more value for an existing one. The second is usually more profitable: selling to an existing client is cheaper than finding a new one.

• A monthly retainer: content updates, seasonal campaigns, new pages — recurring revenue
• Business profile setup and a review strategy — small work, high perceived value
• Content production: informative writing in the client’s field
• Conversion work: not building the site, but showing that it works

One warning: do not sell these by the hour either. Not “four hours of support a month”, but “a monthly maintenance and updates package”. The moment the client starts counting hours you are back in the same trap.

To bring it together: AI is not lowering your rate, it is killing the model where you sell hours. The distinction matters. For a designer who moves to selling outcomes this is the best news in years — the same value now takes far less of your time, and the difference stays with you. The only thing that has to change is the word at the top of the proposal: not hours, but outcome.