Crowdsourcing model

A crowdsourcing model is the named arrangement you choose when you get work from a large, distributed group rather than from staff or a contracted supplier. Contest, microtask, open collaboration and funding are the main types, and each suits a different kind of task.

Think of this entry as a chooser rather than a definition. Our crowdsourcing entry covers the practice itself, and this one sorts that same practice into the arrangements you actually have to pick between.

The choice turns on two questions. Can the work be cut into pieces a stranger can finish alone? And do you want one excellent answer or a very large number of adequate ones?

Get that pair wrong and the crowd behaves badly. A contest run over work that needed ten thousand small hands produces one entry — and then silence.

Motivation is the third question people forget. A crowd works for money, for status, for the cause, or for early access, and an arrangement offering the wrong one attracts the wrong contributors.

Key takeaways

• A crowdsourcing model is an arrangement type, not the practice of crowdsourcing itself.
• Contests buy one best answer — microtasks buy volume at low unit cost.
• Open collaboration suits work the crowd wants to own; funding models buy proof of demand.
• Pick by task shape first, then by how much review capacity you can spare.

How it works

Every model answers three questions differently: how work is divided, how contributors are rewarded, and how quality is checked. Answer those and the model chooses itself. Most failures trace back to picking an arrangement before reading the task shape.

Review capacity is the hidden cost. Microtask models need overlap scoring, because you are trading individual accuracy for volume and only agreement between workers tells you which answers to trust.

Contests invert that trade. You get few entries, so judging is cheap, but you pay for attention up front — and you get nothing at all if the prize is too small for the effort asked.

Cost per usable answer is the number to watch. A cheap unit price means nothing if half the work fails review, and a large prize is cheap if it returns one answer you would otherwise have bought from a consultancy.

Where the crowd is your own user base, community forums and community led growth act as open collaboration models with a commercial purpose attached.

Where you want the work to lift a specific region, impact sourcing is the deliberate opposite of an anonymous crowd — it targets named, trained workers in under served places.

For small, ongoing volumes, smallsourcing often beats a crowd outright. A crowd carries coordination overhead that only pays back above a certain task count.

Examples

Four public examples show the range: citizen science uses open collaboration, federal prize competitions used the contest model, peer edited reference projects run on shared authorship, and product presales run on funding.

Citizen science is the clearest open collaboration case. USAGov’s guide describes citizen science, also known as participatory science, and points volunteers to federal projects that need many pairs of eyes on data.

Government has used contests directly. Challenge.gov ran federal prize competitions for years before the platform was sunset on March 30, 2026, with its content moved elsewhere, which is a useful reminder that platforms outlive neither budgets nor fashions.

Microtask models dominate data labelling. Image tagging, audio transcription and content moderation sampling are all cut into units of a few seconds, priced per unit, and checked by giving the same unit to several workers.

Open collaboration carries a governance cost. Someone has to arbitrate disputes, protect the shared artefact from vandalism and keep a visible history, which is why these projects live or die on their moderators.

Idea platforms sit between the two extremes. The crowd is known and internal, submissions are modest in number, and the real work is the filter that stops a popular idea beating a useful one.

Crowdfunding sits slightly apart. You are not buying work at all, you are buying evidence that demand exists before you commit a production budget, which makes it a market test wearing a sourcing label.

Related terms

These five entries help you place a crowdsourcing model against the alternatives. Two describe crowd shaped arrangements, one is a deliberate counterpoint, and two cover the contracted sourcing structures a crowd is usually compared with.

• Community forums: a standing venue where users answer each other, often the cheapest crowd you already have.
• Community led growth: using that same community as a distribution channel rather than a work source.
• Impact sourcing: hiring named workers from under served areas, the opposite of an anonymous crowd.
• Co-sourcing: sharing one function between your team and a provider under a single plan.
• Multi sourcing model: splitting work across several contracted suppliers instead of a crowd.

FAQ

How do I choose between a contest and a microtask model?

Ask whether you want one excellent answer or a great many adequate ones. Contests buy the first, microtasks buy the second, and mixing them up wastes the budget.

Is a crowdsourcing model the same as crowdsourcing?

No. Crowdsourcing is the practice; a crowdsourcing model is one of the named arrangements that practice can take.

What quality controls does a crowd need?

Overlap scoring for microtasks, expert judging for contests, and peer review with visible history for open collaboration. Every model needs a way to discard bad work cheaply.

When is a crowd the wrong answer?

When the work needs context, confidentiality or continuity. Those three push you toward a contracted team rather than an open call.

Can a crowd model sit alongside outsourcing?

Yes, and it often does. Firms commonly run a contracted team for steady volume and open a crowd only for spikes or for one off creative problems.

If you are weighing a crowd against a contracted delivery team, compare the options on Outsource Accelerator.