What a marketplace is actually selling

The argument, stated fairly

Freelance marketplaces attract two kinds of advice, both useless. One says they are a race to the bottom where nobody makes a living. The other sells a course on ranking first in search results. The truth is that they are a specific kind of market with knowable mechanics, and whether they work for you depends on facts you can check before you spend three months there.

What the platform actually takes

Start with the arithmetic, because it changes the decision.

Commission is typically in the range of 10% to 20% of what the client pays, and on some platforms it is higher on small jobs. Historically several platforms used sliding scales that reduced the cut on long-running clients; these terms change, sometimes annually, so read the current fee page rather than a blog post about it.

Payment processing and withdrawal take another slice. Moving money to a bank account in another country typically costs a fixed fee plus an exchange-rate margin of 2% to 4%, and the margin is the part people miss because it is not shown as a fee.

Bidding credits. On several platforms you pay (in money or in a limited monthly allowance) for the right to submit a proposal. This is the mechanic that most changes behaviour: it converts applying into a cost, so the rational response is fewer, better-targeted proposals, which is the opposite of what most newcomers do.

Add it up honestly. A quoted $500 job can arrive as roughly $400 in your bank account. If your rate arithmetic says you need a certain number, the platform number has to be about 25% higher to reach it.

Why prices are low, mechanically

Marketplace prices are low for a reason that is not about the people on them.

The buyer sees a list of interchangeable-looking profiles with star ratings and prices. They have no way to assess craft, because they are not designers, and the platform’s interface deliberately makes comparison easy. In a market where the buyer cannot distinguish quality, the only visible axes are price and review count. So price falls until it hits the floor set by whoever has the lowest cost of living and the fewest alternatives.

This is not a moral failure of anyone involved; it is what happens to any market where quality is unobservable at the point of purchase. The practical consequence is the important bit: you cannot win on quality in an interface that does not display quality. Competing on being better, in the generic categories, does not work.

When it does work

Three situations where marketplaces pay.

A narrow, technically-specified service. “Prepare press-ready artwork for offset printing, CMYK, bleeds, trapping, PDF/X-1a” is a category where the buyer knows exactly what they want and can tell whether they got it. Quality becomes observable, so price stops being the only axis. The same is true for retouching to a spec, file conversion, vectorising, format adaptation, subtitling.

Getting the first reviews when you have no history. Nobody argues these platforms are a good long-term home; they are a functioning place to convert zero evidence into ten completed jobs and a rating, which is worth something elsewhere.

Finding repeat clients you then keep. Many marketplace earners make most of their money from a handful of clients found there years ago. Note that taking a client off-platform usually breaches the terms and can get you banned; platforms generally allow it after a period or for a fee, and you should read that clause rather than assume.

The mechanics that decide outcomes

If you use them, four things carry most of the weight.

1. Apply within the first hour. Buyers read the earliest proposals and often hire before the listing closes. A better proposal sent on day two loses to an adequate one sent in forty minutes.
2. Answer the specific job. The first line must prove you read the posting: reference their file format, their deadline, their industry. Template proposals are visible at a glance and are the main reason experienced buyers ignore most of their inbox.
3. Ask one clarifying question. It starts a conversation rather than a comparison, and conversations are where price stops being the only variable.
4. Do not race to the lowest bid. Bidding under everybody attracts precisely the clients who chose on price, who are statistically the most demanding and the most likely to dispute. The bid that wins profitable work is usually mid-range with a concrete reason attached.

Platform risk, which is real

Your account is the asset, and you do not own it. Suspensions happen through automated systems, appeals are often handled by templated responses, and reinstatement can take weeks or never come. People lose an entire income overnight with no recourse, and the reviews and ranking that took two years cannot be transferred anywhere.

Treat the platform as a channel, never as the business. The counter-measure is the same as for social platforms: get a client’s direct contact where the terms permit it, keep your own records of every job, and build at least one channel that is not on somebody else’s servers.

What it costs to try

Nothing but time, which is the honest reason it is worth testing rather than arguing about. Set a bound before you start: a fixed number of proposals (say forty) sent properly, within an hour, to well-matched listings. Then count what happened. Forty targeted proposals is enough to tell you whether your category converts at a rate worth continuing; two hundred sloppy ones tell you nothing.

The limitation

For most creative disciplines where the value is judgement rather than execution (identity, art direction, creative strategy) marketplaces are a poor fit and will stay one, because the interface cannot represent what you are selling. A brand designer competing on a listings page is trying to sell a decision in a shop that only displays prices.

If that is your work, the platform is at best a temporary source of small jobs and reviews while you build the channels that suit it. Knowing that in advance saves the six months people usually spend concluding they are not good enough, when the market was not built to show what they are good at.