# Changelog of G2 Reviews Scraper — Software Reviews & Pricing Data (`logiover/g2-reviews-scraper`) Actor

- **URL**: https://apify.com/logiover/g2-reviews-scraper/changelog.md
- **Full Actor documentation**: https://apify.com/logiover/g2-reviews-scraper.md

## Changelog

### 2026-10-15

- A run that reaches its maximum cost per run now stops cleanly with its results saved and a status message, instead of ending as aborted.
- A proxy group your account cannot use no longer stops the run at start-up; it falls back to the default Apify Proxy.
- Memory is managed within the run's own limit, which removes a rare out-of-memory crash on large pages.

### 2026-10-10 (speed update)

- Faster: 1,100 rows in 251 s instead of 590 s, the prefilled example in 50 s instead of 112 s. The first wave goes out at once (main page, a twin in another sort order, star × company-size views); the rest of the views are planned in one go from G2's own filter counts instead of level by level; views are chosen so they barely overlap (~7 new reviews per page instead of ~6).
- Default parallel requests 48 (max 50); runs stop the moment the target is reached and drop requests still in flight.
- Failed views get a replacement planned immediately; idle slots near the end go to products that are still short.
- Cell pages are parsed from the relevant fragments only (2–3× less CPU and memory).
- Fixed: concurrent pages could push a few rows past `maxItems`; the row cap is now reserved before writing, so a run never returns more rows than requested.
- The run summary now reports seconds to first rows, total seconds, page-latency percentiles and requests dropped at the finish.

### 2026-10-10

- First release.
- Product reviews with structured answers (likes, dislikes, problems solved), reviewer role, industry and company size, validation, incentive and source flags.
- Reaches past G2's ten-reviews-per-view limit for visitors by splitting reviews into non-overlapping filtered views using G2's own conditional filter counts.
- Product rows: rating, 5→1 star counts, pros/cons themes with mention counts, AI summary, pricing plans, vendor, categories, alternatives, reviewer demographics.
- Category rankings (15 products per page, all pages), optionally with each product's reviews.
- Filters: star rating, company size, keyword; five sort orders.
- Unknown products/categories fail with NOT_FOUND; every run writes a SUMMARY record.
- Default reviews per product is 500 (API calls that omit the field get a useful sample, not a token one).