Bidding sites solve exactly one problem well: they put work in front of you when nobody knows your name. The price of that convenience is that you compete on price, against strangers, in front of a buyer who has never met you. It is a reasonable starting point and a terrible destination.
Every alternative channel below has the same underlying mechanic: instead of competing for attention at the moment of purchase, you accumulate trust before the purchase exists. That is slower for the first three months and dramatically better for the next three years.
1. Agency subcontracting — the fastest channel nobody uses
Agencies have a permanent structural problem: their sales capacity and their delivery capacity never match. They win a project their team cannot absorb, or a specialist leaves mid-engagement, and suddenly they need someone competent this week.
You will not get your full retail rate — expect 60–75% of it, since the agency carries the client relationship, the sales cost and the risk. In exchange you get something extremely valuable early on: work that arrives without you selling for it, briefs written by people who understand the craft, and payment terms enforced by an accounts department rather than a founder’s mood.
How to actually do it: list twenty agencies within a two-hour timezone of you that sell what you deliver. Email the production or delivery lead, not the general inbox. One paragraph on the specific thing you do, one link to relevant proof, one line saying you have capacity from a specific date. Follow up once, six weeks later. This unglamorous list has kept more freelancers solvent than any funnel.
2. Past colleagues, revisited on purpose
People you have worked with are the only audience that has directly observed your work. That is a conversion advantage no portfolio can replicate. Yet most freelancers announce their new business once, on the day they start, and never mention it again.
The better pattern is a rolling, low-pressure check-in. Twice a year, message twenty former colleagues individually — not a broadcast — with something genuinely useful attached: an observation about their industry, a tool that solved a problem you know they had. Mention your availability in a single closing sentence. Careers move; the colleague who was a junior when you left is a head of engineering four years later with a budget and a vendor problem.
3. Communities where your buyers already talk
This is the highest-yield long-term channel and the one most often executed badly. Executed badly, it looks like joining a Slack group and posting “Hi everyone, I do X, DM me!” Executed well, it looks like eight weeks of answering other people’s questions thoroughly, in public, with no ask attached.
What makes it work is that public helpfulness is unfakeable evidence. Someone reading your fourth detailed answer about database migrations does not need your portfolio; they have already watched you think. When they need that work done, you are not a candidate among ten — you are the person they already trust.
Choose venues by who is in them, not by size. A 400-member forum full of agency owners and technical founders will outperform a 90,000-member general freelancing group every time, because the small one contains buyers and the big one contains competitors. This is also the argument for private, niche communities over open marketplaces in general — a distinction worth reading about in marketplaces vs private communities.
4. Specific, researched outreach
Cold email has a bad reputation because most of it is mail-merged noise. Volume outreach converts at a fraction of a percent; researched outreach to twenty carefully chosen companies routinely converts at 5–10%.
The difference is a first sentence that could only have been written by someone who looked. Not “I love your brand” — something concrete: their checkout drops mobile users at the address step, their careers page has been hiring for the same role for five months, their docs site takes eleven seconds to load on a phone.
- Line 1: the specific thing you noticed
- Line 2: why it costs them something
- Line 3: one sentence of proof you have fixed it before
- Line 4: a small ask — a 15-minute call, not a project
Four lines. No attachments, no deck, no rate card. The goal of the first email is a conversation, not a contract.
5. Publishing that answers buying questions
Content marketing for freelancers does not mean a blog with daily posts. It means writing down the answers to the questions clients ask you during sales calls — once, properly — and letting search engines and colleagues distribute them.
Twelve genuinely useful articles compound for years. “How much should a WooCommerce migration cost?” is a page that quietly qualifies leads, sets your price anchor and demonstrates expertise while you sleep. And when a prospect arrives having already read three of your pieces, the sales conversation starts from a completely different place: they are deciding when, not whether.
6. Productised referral partnerships
Find the people who sell to your clients immediately before or after you would. A WordPress developer’s natural partners are hosting providers, SEO consultants, brand designers and bookkeepers who serve the same size of business. None of you compete; all of you meet the same buyer.
Make it concrete rather than a vague “let’s refer each other”. Agree a specific trigger (“when your client asks about site speed, send them to me”), a specific handoff (a three-line intro email), and reciprocate deliberately. Two or three of these relationships can supply a meaningful share of a freelance pipeline with almost no ongoing effort.
7. The clients you already have
The cheapest client to win is one who has already paid you. Freelancers chase strangers while a finished project sits three months old and unmentioned, and the client who was delighted with it has since hired someone else for the follow-up work — because you never told them you did that too.
Two habits capture most of this lost revenue:
- The 30-day check-in. A month after delivery, ask how the thing is performing. Not a pitch — a genuine question. It surfaces problems while goodwill is high, and problems are projects.
- The specific referral ask. “Do you know anyone else?” produces nothing. “Do you know anyone running a Shopify store who’s frustrated with their checkout?” produces names, because you have given their memory something to search for.
Ask at the moment of maximum goodwill — right after you deliver something that worked, not months later when the glow has faded.
What to do when a channel is not working
Before abandoning a channel, work out which stage is actually broken. Almost every “this doesn’t work” verdict is really one of four different problems, and they have different fixes.
| Symptom | Real problem | Fix |
|---|---|---|
| Nobody replies | Message or targeting | Get more specific about who and why |
| Replies, no calls | Credibility gap | Add proof — case studies, public work |
| Calls, no proposals | Qualification | Ask about budget and timeline sooner |
| Proposals, no wins | Price or scope framing | Offer tiered options instead of one number |
Note that only the last row is about price, yet price is where nearly everyone starts cutting. If people are not replying at all, a lower rate will not help — nobody is reading far enough to see it.
Effort vs payback
| Channel | Time to first client | Rate quality | Compounds? |
|---|---|---|---|
| Agency subcontracting | 2–6 weeks | Medium | Yes — repeat work |
| Former colleagues | 1–8 weeks | High | Slowly |
| Community participation | 2–4 months | High | Strongly |
| Researched outreach | 3–8 weeks | High | No — always manual |
| Publishing | 4–12 months | Highest | Strongly |
| Referral partners | 1–3 months | High | Yes |
| Bidding sites | Days | Low | No |
The sensible strategy is not to abandon platforms on day one. It is to use the fast, low-rate channels to cover expenses while deliberately building the slow, high-rate ones — and to notice the moment the slow ones can carry you, rather than staying on the treadmill out of habit.
Run two channels at a time, for at least ninety days each. One channel is fragile; six half-started channels produce nothing at all.
Whichever you pick, the requirement is patience measured in months rather than days. Most freelancers abandon a channel at week three — right at the point where the compounding ones are about to start returning anything. If you are still weighing whether the platforms are worth their cut in the meantime, the platform comparison lays out what each one actually costs.
Leave a Reply