Steam Review Sampler

Private-first Actor for researching what players like and dislike about competitor games. Input app IDs from Steam Market Scout, select positive/negative reviews, language, recent/helpful/funny ordering, and cap output at 100 reviews per game per run.

Useful workflow: first discover comparable titles, then sample both helpful negative and positive reviews; group recurring complaints into onboarding, balance, performance, content depth, monetization and UX. Review text is user-generated and non-representative. This Actor emits no profile URL or SteamID, uses public endpoints without login, and has a hard per-game sample limit with delays.

Example: appIds=["588650"], reviewType="negative", sortBy="helpful", maxReviewsPerGame=40, language="english". The schema pre-fills Dead Cells (588650) with a small 10-review, one-page sample so Apify Store automated QA can verify a non-empty dataset within its five-minute limit; replace the app ID for normal use.

Filters and limitations

Optional filters include review date range, minimum/maximum playtime at review, minimum helpful votes, minimum text length, and required/excluded phrases. Set maxPagesPerGame and maxReviewsPerGame to control the workload. Results are sampled from public endpoints and are not a representative survey. Uses Steam’s current official IUserReviewsService/GetAppReviews endpoint. The old Store appreviews route was retired in October 2026, so this Actor deliberately does not use that legacy endpoint. Steam may rate-limit anonymous calls; the Actor applies delays and reports partial failures rather than treating them as zero reviews. This Actor does not require Steam login and does not output profile URLs or Steam IDs. \nBy default the current Steam API excludes reviews marked as off-topic/review-bomb activity; set includeOfftopicReviews=true to include them for a separate research pass. Playtime filters are applied locally in minutes so they are not rounded to the API's whole-hour filters.\n