# Changelog of Google Trends Scraper: Compare 100 Keywords (`aaronoptimus/google-trends-scraper`) Actor

- **URL**: https://apify.com/aaronoptimus/google-trends-scraper/changelog.md
- **Full Actor documentation**: https://apify.com/aaronoptimus/google-trends-scraper.md

## Changelog

### Next build (unreleased)

- **Trending Now, rebuilt:** reads the same public data as trends.google.com/trending. New filters: window (4, 24, 48 hours or 7 days), category, active only, max topics per location (default 50), and supported regions such as `US-CA`. Each topic now has search volume, % increase, start and end time, active flag, categories, related searches (breakdown) and up to 10 news articles. Trending Now URLs with `hours`, `category` and `status` are understood. The RSS feed stays as a fallback.
- **Monitor mode (`onlyNew`):** scheduled runs return only Trending Now topics and rising related queries that earlier runs didn't, remembered in a named key-value store (`monitorStoreName`). A run with nothing new ends normally.
- **Input names from Apify's Google Trends Scraper** are accepted: `isMultiple` (a comma line becomes one comparison), `timeRange: ""`, `category: ""`, `startUrls`, `customTimeRange`, `maxItems` (max search terms) and `maxRequestRetries`.
- **New prices with Apify tier discounts:** search term from $0.003 (Free) down to $0.0012 (Diamond), Trending Now topic from $0.001 down to $0.0004, run start $0.00005.
- Flat rows get Trending Now columns (`searchVolume`, `increasePercentage`, `startedAt`, `endedAt`, `isActive`, `trendingCategories`, `breakdownTerms`, `source`), added after `warnings` so existing columns keep their positions.
- The Output tab links the Overview and Trending Now views as well as all results and the run summary.

### 1.1 — October 2026

- **Faster recovery from throttling:** when Google throttles the default proxy, all parallel sessions now switch to the residential fallback together (before, each session had to be blocked 3 times on its own). A resumed or migrated run starts on the fallback if the previous part had switched.
- **No more timed-out runs:** shortly before the run's timeout the scraper stops starting new terms, saves its results and ends normally. Unstarted terms are listed in the run summary (`unfinishedTerms`) and not charged.
- **Faster runs:** Google withholds related topics from automated requests. After 3 terms come back without topics, the rest of the run skips that request (a note on each item says so). Related queries are unaffected.
- **Correct zeros:** points Google reports as zero volume are now `0` (as on the website and in Google's CSV), not `null`. This fixes averages, % change and trend labels for low-volume terms. Terms with only zeros are still marked `hasData: false` and not charged.
- **Input:** a line like `pickleball, padel` now counts as two terms, as on Google Trends. Invalid location, language and date formats are rejected before the run starts, so a typo no longer costs the run-start fee.
- **Output:** flat rows carry the item's notes in a new `warnings` column (last column, on the summary row). The Overview table shows whether each term had data. The run summary adds proxy traffic figures (`network.downloadedKb`, `fallbackRequests`, `fallbackDownloadedKb`).
- Error messages say that only the run-start fee applies when nothing could be fetched.
- Builds install the exact locked dependency versions (`npm ci`).

### 1.0.1 — October 2026

- Add a warning when Google returns no related topics or queries for a term, and an FAQ entry explaining it.
- Fill in the Apify username (AaronOptimus) in the README examples.

### 1.0 — October 2026

First public release.

- Interest over time, interest by region (country / state / DMA / city), related queries and related topics.
- Compare mode, including **up to 100 terms on one scale** (anchor normalization) with precision warnings; terms that dwarf the anchor are flagged `notComparable` and not charged.
- **Google Trends URLs** accepted as input (location, dates, category and search type taken from the URL).
- Several locations per run (`US, GB, DE`).
- **Flat rows** output format for CSV / Excel / Google Sheets.
- Trending Now by country, with approximate search volume and news links.
- Reliability: per-session cookies and consistent browser fingerprint, session/IP rotation, capped backoff, per-term time limit, residential fallback, parallel sessions, resume after restarts.
- Pay only for terms that return data; failed, no-data and all-zero terms are free; *Maximum cost per run* is respected across all event types and parallel sessions.
- Crash- and migration-safe billing: progress is saved before every charge, so a restarted or migrated run never charges the same term twice.
- Clear errors for invalid locations and unsupported requests; duplicate lookups are skipped.