Most bookings never touch it

You do not need a social account to earn money doing this, and most people who tell you otherwise are describing a different trade. The overwhelming majority of bookings come from someone searching inside a marketplace, comparing a handful of listings, and picking one - a path that never touches an external profile.

That does not make audience-building worthless. It makes it a second business with its own skills, its own timeline and its own failure modes, and the honest question is whether you want to run two.

Where bookings actually come from

Marketplace search is the dominant channel, repeat buyers are the second, and platform-internal browsing is a distant third. Off-platform traffic is a rounding error for almost everyone who does not already have an audience.

The mechanism is worth understanding rather than just the ranking. A buyer with a specific intent goes where the supply is aggregated, because that is where comparison is cheap. Nobody with a task to complete goes looking for individual practitioners on a social platform first; they do that only after the marketplace has failed them, or after a name has been recommended to them.

The piece on where demand comes from has the full breakdown, and the short version is that most earners badly overestimate browsing and badly underestimate search.

What the second job actually costs

Assume, conservatively, three to five hours a week of posting, replying and thinking about what to post. Assume six months before anything measurable happens, and assume the first three of those produce nothing at all.

If your effective rate is £25 an hour, that is £300 to £500 a month of time, spent against a return that arrives, if it arrives, in the second half of the year. Against the same hours spent taking more jobs, being reachable in the busy windows and getting your turnaround down, audience-building loses badly for the first two quarters and only sometimes wins after that.

There is also an attrition cost. Audience work has a much higher quit rate than client work because the feedback loop is slow and noisy, and quitting halfway leaves a stale account that reads worse than no account. The time commitment that nobody mentions is already larger than new earners expect before you add this on top.

When it does pay

Three situations, and they are narrow.

You already have an audience. If you have a following from something else and it plausibly overlaps with this demand, you are not building, you are converting, and that is a completely different economics.

Your format is inherently demonstrable. If what you sell can be shown in public without violating anyone's privacy - a rubric, a comparison method, an explainer - then public work doubles as portfolio and the hours are not purely promotional. The worked-example piece covers the version of this you can do without any audience at all, which captures most of the benefit.

You are trying to leave the platform's rails. If your medium-term goal is direct buyers rather than marketplace buyers, an audience is the only route there. Be clear that this is a strategic bet with a two-year horizon, and check the terms first, because most platforms prohibit diverting buyers off-platform and enforce it - the mailing list piece covers where the line usually sits.

Outside those three, the honest answer is that you are better off improving the listing you already have.

The minimum viable version

If you want a presence without a second job, build a landing point rather than a channel.

One account, on the platform where your buyers already are. A bio that names the format and links to your marketplace listing. Four or five posts that establish the account is real, and then nothing on a schedule. Where it is not obvious that a promoted service is your own, the FTC's Endorsement Guides FAQ expects the connection to be made clear; where it is obvious, no disclosure is needed.

That is enough to serve the actual function of an off-platform account, which is not acquisition but verification: a buyer who found you in the marketplace, half-decided, and went looking for evidence that you exist as a person. It costs an afternoon and no ongoing hours, and it captures most of the realistic upside.

The failure mode to avoid is the account that posts weekly for two months and then stops. A dormant account is worse than no account, for the same reason a dormant listing is - recency is read as reliability everywhere a buyer looks, not just on the listing.

The privacy arithmetic

An off-platform presence is the largest single hole in a working alias.

The exposures are mundane rather than dramatic: a reused profile image, a handle that also appears on a personal account, a photograph with location metadata, a follower overlap that identifies you by association. The UK Information Commissioner's Office suggests using "the highest privacy settings when you first create your profile", then loosening them gradually. Each exposure is small and they compound, and anonymity is almost impossible to recover once any one of them lands - the earner's version of that argument is worth reading before you open the account rather than after.

The general treatment of how identifying information travels between services is set out here, and the image-specific hazards - reverse search, reuse, metadata - are covered here. Both are more careful than the advice a marketplace will give you, because a marketplace has no particular interest in your being hard to find.

What to do instead, if you want a growth lever

Referrals and repeat buyers are the channel that actually compounds in this trade, they cost no additional hours, and they are invisible from outside - which is why nobody writes posts recommending them. A buyer who returns twice is worth more than several hundred followers, and the thing that produces one is noticing something specific in the work rather than being fast.

Buyers describe finding people by search and by recommendation, in that order, and almost never by following them - which is visible in their own account of how a commission starts. The one thing a following cannot do is win you the cheap generic end of the market, because that end already belongs to instant automated tools rather than to any person, however visible.

If your aim is simply to be more findable within a platform rather than outside it, the listing and visibility mechanics on Rate Cock's judges page are the cheaper lever, and they respond in weeks rather than quarters.