Five ways to find work if you hate selling

There's a spreadsheet you built in March. Forty target companies, a contact name for maybe twelve of them, a column headed "status" that is empty all the way down. You've opened it twice since and closed it both times.

That spreadsheet is the through-line here. It isn't a research problem — you know who to contact and roughly what to say. The blocker is the ask itself, and no amount of further advice about subject lines fixes an aversion to sending them.

So this post skips that channel entirely. Cold outreach has its own mechanics and they're covered in cold email that gets a reply, line by line. What follows is five ways of getting work where the selling is built into the structure of the thing rather than performed by you in a message.

They are slower. That is the honest trade. Each one below says what it actually returns.

1. Publish the work itself, on something you own

Not a portfolio grid. A written account of one project: what the client wanted, what constraint made it hard, what you tried, what you'd do differently. The buying signal is the reasoning, not the screenshot — anyone can show a finished interface, almost nobody shows the decision behind it.

This channel sells for you because the reader arrives already convinced or already gone. There is no ask in it at all.

Two practical points. First, put it somewhere you control. Working Not Working ran for thirteen years as a creative-talent platform, was acquired by Fiverr in February 2021, and closed on 30 June 2025 with no shutdown notice published (Fiverr's own acquisition announcement is still up; the closure was not announced). Portfolios that lived only there went with it. Second, a profile on a distribution platform is fine as a mirror, not as the original — Contra, for instance, charges freelancers no commission on project earnings and monetises client-side instead, with per-payment fees of $2 to $29 and a $29/month Pro tier (Contra pricing, client fees).

This week: one project, 600 words, published on your own domain. Not five. One, finished.

What it returns: nothing at all for a long stretch, then occasional inbound of unusually high quality, because someone who read 600 words about your process has pre-qualified themselves. You cannot forecast from it, so don't. Treat it as an asset you're compounding while other channels pay the rent.

2. Answer the question your buyer is already typing

Somebody in your market is right now writing a question in a forum, a Slack, a subreddit, a GitHub discussion, a LinkedIn comment thread. Answering it well is the entire pitch. You are demonstrating the thing rather than describing it, and the ask is zero — you never make one.

The discipline is choosing venues where buyers are, not where peers are. A subreddit full of other freelancers is a nice place and a bad channel. A community of ecommerce operators, or of finance leads at mid-size companies, is where the person with a budget goes when something breaks.

Answer the question and stop. No signature block, no "happy to help further, DM me." The follow-up is the buyer's move to make, and it converts at a much higher rate when they make it.

This week: three answers, twenty minutes each. Pick questions where your answer is genuinely better than the ones already there.

What it returns: low volume, high intent, lumpy timing. Weeks of nothing, then a question you answer well and the person who asked it turns out to have a budget. It compounds sideways too: people remember who answers, and that memory is exactly the mechanism behind the good work that never gets posted.

3. Take overflow from freelancers who are full

The best-converting lead in the market is a freelancer at capacity who has just been asked to do something they can't fit in or don't want to do. The client is qualified, the budget exists, the timeline is agreed, and someone with credibility has already vouched for the standard of work.

Your ask here goes to a peer, not to a buyer. That's the whole point. "I've got three weeks free from the 6th, send anything you can't take" is a sentence most people can say without discomfort, and it is a complete sales process.

The money question is what the arrangement costs you. There are two structures. In a referral, the client contracts with you directly and the freelancer who passed it on either takes a fee or takes nothing. In a subcontract, they hold the contract, you invoice them, and they keep a margin for carrying the client relationship and the payment risk. I can't give you a market-standard margin figure, because no one publishes one that survives scrutiny — so use the mechanism instead: the margin should reflect who absorbs the loss if the end client doesn't pay, and who is on the hook if the work is late. If they carry both, a margin is fair. If you're doing client-facing work and chasing your own approvals, it isn't.

Worth comparing against the alternative. A Fiverr seller keeps 80% of the order value (Fiverr); on Malt the freelancer fee runs 5% to 10% excluding tax, dropping to 5% after six months on the same project (Malt). A prime who takes a slice of your rate is charging you for the same thing a marketplace charges for — demand and payment intermediation. The difference is the prime also handles the client, which the marketplace does not.

Who holds the contract, who chases the invoice, and what the prime's margin buys are worked through properly in why other freelancers are your best referral source. For now: get on three overflow lists.

4. Tell adjacent suppliers what you take, and what you don't

A designer needs a developer. A developer needs a copywriter. A brand consultant needs someone to build the thing. Every one of those people is asked "do you know anyone who does X" several times a year, and each holds exactly one name per category, because holding two is effort.

This is the same recall mechanism as a client shortlist, except the gatekeeper is a supplier rather than a buyer, and suppliers are far easier to approach because you are not asking them for money.

The message is not a pitch. It's a capacity notice with a boundary in it, which is what makes it useful to them:

Hi [Name] — we crossed over on [project / thread / event].

I do [one specific thing] for [one specific kind of buyer]. Recent example:
[one line, with the outcome].

I don't do [the adjacent thing you don't do] — I usually point people at
[name or type of person] for that.

I've got capacity from [month]. If something lands on your desk that isn't
yours, send it over and I'll do the same.

"One specific thing for one specific kind of buyer." They can only repeat a sentence they can hold in their head. "Full-stack marketing support" is unrepeatable. "Shopify checkout work for DTC brands doing over £1m" gets repeated verbatim.

The recent example. One line with an outcome in it. This is the evidence that stops the sentence being a claim.

"I don't do X." Counter-intuitive and the highest-value line in the message. Naming what you decline makes the rest credible and tells them precisely which requests to forward. It also stops you being sent work you'll hate.

"Send it over and I'll do the same." Reciprocity stated once, plainly. Don't oversell it and don't promise a volume you can't deliver.

Capacity from [month]. A real month. Vagueness here reads as "not really available."

This week: five people, five messages, all sent in one sitting.

What it returns: the highest hit rate of anything in this post, and the longest lag. Referrals arrive when the other person's pipeline overflows, which is not on your schedule.

5. Make the first purchase small enough to be a shrug

Most of the discomfort in selling is proportional to the size of the ask. A €12,000 project needs a decision, a stakeholder, probably a procurement step. A €900 fixed-scope audit delivered in a week needs one person's approval and a card.

So productise a small first engagement: fixed scope, fixed price, fixed duration, a named deliverable. An audit, a two-day sprint, a diagnostic with a written recommendation. You stop selling yourself and start selling a defined thing, which is a conversation almost everybody finds easier.

Price it as an entry point, not as a discount. Two rules. It must be profitable on its own — if it loses money and the follow-on doesn't materialise, you've bought a client at a price you didn't calculate. And it must end in a written recommendation that names the next piece of work, because that document is what converts.

Put a validity date on the quote. It creates a decision point without you having to create one in a follow-up message, which is precisely the interaction this reader avoids — the quote generator has a valid-until field and produces the PDF in your browser, no signup and no watermark.

The overhead objection is real: a productised offer means a proposal, a contract and an invoice for a small fee, repeatedly. That's why setup cost matters more here than on big projects. It's the comparison the Dubsado alternative page makes directly — Dubsado starts at $335 a year with no month-to-month price published (Dubsado pricing), which is a meaningful outlay if your entry offer is €900. Worklyn's proposals convert to a contract or an invoice in one click, and proposals, contracts and invoices are unlimited on the free plan.

What it returns: the shortest path from first contact to paid of anything here, and the clearest signal. A buyer who won't spend €900 on a diagnostic was never going to spend €12,000 on the project.

Which of these to actually pick

Not all five. Pick two: one that pays this quarter (3 or 5) and one that compounds (1 or 2). Running five channels badly is how the March spreadsheet happened in the first place.

If you want the full ranking of channels by hours invested per client won, including the ones this post deliberately skips, that's where freelance work actually comes from in 2026.

Five messages today, one offer this month

In the next thirty minutes: write the five adjacent-supplier messages from section 4 and send them. Five people you've already met, one sitting, no research phase. If you can only manage three, send three.

This month: define one productised first engagement (scope, price, duration, deliverable) and quote it to two people. Then delete the March spreadsheet, or archive it honestly as a channel you've decided not to run. An unworked list isn't a pipeline; it's a monthly reminder that you're avoiding something, and it costs you more in morale than it would ever have returned in leads.

Worklyn's client portal gives a small first engagement its own magic link where the client approves, signs and pays in one place, so a €900 audit doesn't need five emails to close.