# Content Gathering (/docs/products/content-gathering)



Content Gathering points a trained crowd at your search criteria — keywords, platforms, date ranges, languages, and minimum engagement thresholds — and returns matching posts, articles, and multimedia, each with source links and engagement metrics attached.

Unlike Crowdee's other products, gathering doesn't assume a known, finite set of inputs. It's an open-ended search: you don't know in advance how many matching items exist, so the underlying job runs as a `gathering`-category crowd job rather than the fixed-variant model used elsewhere.

## What You Get [#what-you-get]

* **Deduplicated items.** Every submission is hashed per job, so the same post found by multiple crowd workers is only counted — and paid — once.
* **Structured metadata.** Platform, author handle, content type, and engagement metrics (likes, shares, views) travel with every item, not just the raw text or file.
* **Review status.** Gathered items start `pending` and move to `accepted`, `rejected`, or `duplicate` as you (or an automated rule) review them.

## Chaining into Other Products [#chaining-into-other-products]

Accepted items convert directly into:

* A [dataset](/docs/concepts/datasets) version, ready for [Multimedia Verification](/docs/products/multimedia-verification.mdx).
* An [input data set](/docs/crowdsourcing/input-data), ready for [Source Research](/docs/products/source-research.mdx) or a custom crowdsourcing job.

No manual export step is required — both conversions are single API calls against the gathering job.

## Cost [#cost]

There's no fixed platform price — you set `rewardCredits` per gathered item, the same way you would for any crowd job, and credits are blocked upfront for `maxAssignments × rewardCredits`. Typical campaigns pay €0.10–€0.50 per accepted item, depending on platform and content-type complexity.

## Where to Go Next [#where-to-go-next]

See [Content Gathering](/docs/crowdsourcing/content-gathering.mdx) for the full technical reference: job creation, criteria schema, the worker-facing search-assist widget, and the dataset/input-data conversion endpoints.