I freelanced full time for four years before I started taking on studio work, and almost everything I got wrong in year one was a boundary problem rather than a skill problem. The code was fine. The invoicing, the scoping and the saying-no were not.

This is the guide I would hand to someone about to send their first proposal.

Decide what you sell before you look for buyers

“Front-end developer” is a job title, not an offer. Nobody wakes up wanting a front-end developer; they wake up wanting a slow checkout to stop losing them money.

Write your offer down as a sentence with a subject and an outcome:

I rebuild marketing sites for small SaaS companies so their pages load in under a second and the team can edit copy without a developer.

That sentence does three things at once. It tells a stranger whether they are your client, it tells you which projects to decline, and it gives you something to write about that is not “I am available for work.”

The first three clients come from people who already know you

Cold outreach works eventually. Warm outreach works this month. In order of hit rate:

• Former colleagues who moved somewhere new. They already trust your work and they now have a budget you never had access to.
• Agencies with more sold work than staff. One good agency relationship can fill half a year.
• People whose work you admire publicly and specifically. Not “love your site” — “your pricing page does the comparison table better than anyone and here is what I would try next.”

Send five of these a week. Not fifty. Five that are actually about the recipient.

Price the outcome, then sanity-check it by the hour

Hourly billing punishes you for getting faster, which is the one thing you are guaranteed to do. Quote the project.

To get to a number I still use the same three-step arithmetic:

1. Estimate the working days honestly, then add 40%. The 40% is not padding, it is revisions, meetings and the week the client goes quiet.
2. Multiply by the day rate you would need to hit your yearly target working 130 billable days — not 250. The rest of the year is holidays, admin, sales and sick weeks.
3. Compare the result to what the outcome is worth to the client. If your number is a rounding error next to their revenue, you are probably still too cheap.

Then put the number in the proposal without an apology and without three alternative tiers designed to make you look flexible. Two options at most: the scope you recommend, and a smaller scope.

Write the scope like you expect to be held to it

The contract is not there for the court case. It is there so that in week six you and the client are reading the same sentence.

Mine always names four things:

• Deliverables — page by page, feature by feature, with the things that are explicitly not included listed underneath.
• Rounds of revision — two is normal. Say what a third costs.
• What you need from them, and by when — copy, brand assets, access. Late inputs move the delivery date; write that down.
• Payment — 50% up front, 50% on delivery, net 14, and a line about what happens if an invoice ages past 30 days.

A project without a deposit is a hobby with paperwork.

Protect the calendar harder than the inbox

The burnout does not come from the work. It comes from the work arriving at every hour of every day.

What actually helped me:

• One client per morning. Context switching between three codebases before lunch costs more than an hour of overtime.
• A stated response window. “I reply to messages once in the morning and once at the end of the day” is a professional statement, not a rude one. Nobody has ever objected.
• A weekly written update. Five lines every Friday: what shipped, what is next, what I am blocked on. This one habit removed roughly every “just checking in” message I used to get.
• Friday afternoons blocked. Invoicing, bookkeeping, and the next five outreach messages. If sales only happens when work runs out, you will always be one month from panic.

Raise your rate on new clients, not on old ones

Raising a rate mid-relationship feels like a betrayal and is a hard conversation. Raising it for the next person who asks costs nothing and feels like nothing.

Quote the new number once. If it lands, that is your rate now. If two in a row decline for price, you found the ceiling for that kind of work and you learned it in a fortnight instead of a year.

Keep a “no” list

Mine, after four years:

• No projects where I cannot talk to whoever makes the final call.
• No fixed price on work that starts with “we are not sure what we want yet.” That is a discovery engagement, billed separately.
• No equity instead of money from someone I have not worked with.
• No new client in the two weeks before a launch.

Every line on that list came from an expensive month. Yours will look different, and it will be worth more than any of this.

The part nobody says out loud

Freelancing rewards mediocre code and excellent communication far more than it rewards the reverse. The best technical decision I made in four years was writing a weekly update. Everything else was details.