Every channel for finding clients works for someone. The arguments about which is best are usually arguments between people at different stages of their business who have forgotten that they are not describing the same situation.
So rather than ranking them, here is what each one actually costs you, what it gives back, and when it starts making sense.
Referrals convert better than anything else. The client arrives already trusting you, the price conversation is shorter, and the work usually resembles what you did for whoever sent them.
The catch is that you cannot turn them on. Referral volume is a function of how many happy clients you have, which is a function of how much work you have already done. For anyone past a few years in, it can carry the whole business. For anyone in their first eighteen months, planning around referrals is planning around something you do not control yet.
Upwork and its neighbours solve the immediate problem. There is demand, it is searchable, and you can be earning inside a week.
What you give up is the price. You are competing in a window where the buyer sees fifteen profiles side by side, and the only difference they can evaluate quickly is the number. Some people run profitable businesses there by specialising hard enough that the comparison stops working. Most end up doing more work for less, and the clients do not carry over when they leave.
Ads can work, but the arithmetic is brutal for a small shop. A click on anything resembling "web design" costs real money, a tiny fraction of clicks become enquiries, and a fraction of those become projects. Work backwards from a project worth a few thousand and you find you need a budget large enough that a bad month genuinely hurts.
Agencies that make ads work usually have something the rest do not: a narrow niche, a landing page built for that niche, and enough historical data to know what a lead is worth. Buying traffic before you have those is paying to learn what you could have learned for free.
The key idea: the right channel is the one that matches what you currently have most of. Early on that is time. Later it is reputation and money. Picking a channel that needs the thing you have least is the usual mistake.
Chambers of commerce, trade associations, the local business breakfast — these have a reputation for being a waste of a morning, and for a lot of people they are. But the businesses in the room are exactly the ones buying five-thousand-euro websites, and almost nobody in your trade turns up.
It is slow. You will go several times before anything happens. The return comes from being the person they already know when the website finally becomes urgent.
Contacting businesses directly is the channel with the worst reputation and the most control. You decide how many today. You decide which trade, which town, which kind of business. Nothing else on this list lets you do that.
The reason it has a bad name is that most of it is done badly — untargeted, generic, sent in volume. Done narrowly, with a specific reason for contacting that specific business, it behaves like a different channel entirely.
Notice outreach never leaves. The agencies that keep a small amount running permanently are the ones that do not panic when a big client ends.
Whichever channel you choose, the limiting factor is the same: knowing who to approach. A list of businesses is not a list of prospects. The useful question is which of them has a website problem worth paying to fix, and answering that means opening each site, judging it, and checking whether anyone is maintaining it.
That is the slow part, and it is the part LeadCrawlr does. It takes a trade and an area, finds the businesses on Google Maps, looks at each website and ranks it from 1 to 10 — with the ones that have no website at all kept separately, because they are a different conversation. The output is the shortlist you would have built by hand over a weekend.
Pick the channel that fits what you have now, and give it long enough to produce evidence. The common failure is not choosing wrong — it is switching after three weeks, repeatedly, and never finding out whether any of them would have worked.
LeadCrawlr scans Google Maps, scores every business website 1–10, and flags the ones that need you most. $49/year, unlimited scans.
Get LeadCrawlr — $49 / year