# Changelog of Websites by Traffic Volume 2026 | $5/1K | Filter 40M by Visits (`apivault_labs/website-traffic-database`) Actor

- **URL**: https://apify.com/apivault\_labs/website-traffic-database/changelog.md
- **Full Actor documentation**: https://apify.com/apivault\_labs/website-traffic-database.md

## Changelog

### \[2.2] — 2026-09-19

- Added explicit auto, traffic, global-rank, domain-lookup, and similar-site workflows.
- Added a safe `maxResults=25` Console/MCP default and compact, outreach, and full output presets while preserving legacy API behavior.
- Added eight numbered form sections, typed Dataset fields, official results/summary/errors links, and stable SUMMARY/ERRORS records.
- Preserved the production 2.1 grouped-shard OOM fix and successful no-result behavior for unknown similar-site seeds.

### \[2.1.2] — 2026-08-09

#### Fixed

- **Runs no longer crash on the default input.** Rows from all 67 shards were
  collected before merging, which peaked near 1 GB and got the container killed
  (exit 137) about 40 seconds in — the default settings failed every time. Shards
  are now folded into the running top-N in groups and trimmed as they arrive, so
  peak memory is flat: the same run now finishes in ~42 s using 133 MB instead of
  977 MB, with identical results.
- **Looking up competitors of an untracked domain no longer fails the run.** The
  database covers a slice of the web, so many well-known domains simply aren't in
  it. That now ends the run successfully with an explanatory message instead of
  reporting an error.

#### Changed

- **Max results now defaults to 1,000** instead of 10,000. Since you pay per
  result, the old default meant a single click could bill far more than intended.

### \[2.1]

#### Changed

- **Clearer guidance for the Top-N (Global Rank) mode.** The input now states
  plainly that to get the top sites (top 100, top 1000, …) you only fill the
  Global Rank field — the min/max monthly-visits fields do not need to be
  emptied, because they are ignored automatically in rank mode.

### \[2.0]

#### Fixed

- **Global Rank mode now returns the pure worldwide top N.** Previously a value
  left in an optional filter (e.g. a country) still applied in rank mode and
  could shrink the top-N list to zero. Selecting by Global Rank now ignores the
  traffic range and all optional filters (category, country, channel,
  engagement, growth, keyword), so "top 100" always means the global top 100.

### \[1.9]

#### Changed

- **Clearer, self-explanatory input layout.** Two ways to choose sites are now
  their own numbered sections: **① Select by monthly traffic** and
  **② Or: get the world's top sites by Global Rank**. Optional filters,
  results/sorting, and the advanced domain-lookup / competitor tools follow as
  sections ③–⑥.
- **No mode dropdown to set.** Just fill the section you want: enter a Global
  Rank number to rank the top sites (visits then ignored), or leave it empty to
  select by traffic.

### \[1.8]

#### Added

- **Top Global Rank input:** enter `100` to get sites ranked **1–100 worldwide**,
  automatically ordered from rank 1 to 100. Filling it ignores the visits range.

### \[1.7]

#### Added

- **7 new filters** to narrow the database like a market-research tool:
  - **Ranks for keyword** — reverse keyword lookup (sites that have a keyword
    among their top search keywords).
  - **Traffic channel + minimum channel share** — filter by any channel
    (search, direct, social, referral, email, paid ads, AI/LLM) and a min %.
  - **Minimum growth %** — fast-growing / trending sites by month-over-month change.
  - **Engagement:** maximum bounce rate, minimum pages per visit, minimum time on site.

#### Changed

- Default `maxResults` raised from 1,000 to 10,000.
- **Bulk export up to 1,000,000 results.** Pulls over 50,000 stream results
  incrementally (memory stays flat), so paying users can export full lists. The
  export runs shards in parallel over server-side cursors (sorted once, streamed
  in chunks) and writes results in batches, so it stays fast and light on the
  database. Up to 50,000 stays an accurate global top-N.

### \[1.6]

#### Added

- **Pay-per-result billing.** Each delivered website record is now charged once,
  before it is written, so you never pay for anything you don't receive.

### \[1.5]

#### Changed

- **Clearer input layout.** The three ways to use the actor are now numbered and
  explained: ① filter the database (main mode) first, then ② results/sorting,
  then the optional ③ domain lookup and ④ competitor search at the bottom. Each
  section says when to use it and what it overrides.

### \[1.4]

#### Added

- **Find similar sites / competitors** — pass one domain via `similarTo` to get
  competitors: sites in the same category with comparable monthly traffic.

#### Fixed

- Pruned data sources that returned no results, making queries cleaner and
  slightly faster.

### \[1.3]

#### Changed

- Repositioned as a **filter/discovery tool over the 40M-site dataset** (not a
  per-URL live checker). Updated title, description and docs for clarity.
- Domain lookup is now clearly documented as working only for domains already
  tracked in the database.

### \[1.2]

#### Fixed

- **"Monthly Visits" now really means monthly.** The min/max visits filter and
  the "Monthly Visits" sort previously used the all-time total (a multi-month
  sum), which returned confusing results. They now filter and sort on the most
  recent month's visits, matching the `latestMonthVisits` field.

### \[1.1]

#### Added

- **Domain lookup** — pass one or more domains to get full traffic data for
  specific sites (all filters are ignored in this mode).
- **Top search keywords** with search volume per site (`topKeywords`).
- **Top countries** visitor breakdown per site (`topCountries`).
- **Monthly visits history** (`monthlyVisits`) and **month-over-month growth**
  (`monthlyGrowthPercent`).
- **AI/LLM traffic** — AI referral share is now included in the traffic mix, plus
  a new "Only sites with AI traffic" filter.
- **New filters:** global-rank range, minimum search-traffic share, and
  "only growing sites".
- **New output fields:** category rank, country rank, and data date.
- **New sort options:** category rank, country rank, AI visits.

#### Fixed

- Traffic mix now includes the **AI** source (previously dropped from output).

### \[1.0] — 2026-08-03

#### Added

- Initial release with 40M+ websites database
- Filter by traffic range (minVisits, maxVisits)
- Filter by category (partial match)
- Filter by country code
- Multiple sort options (visits, rank, bounce, pages, time)
- Traffic source breakdown (direct, search, social, referral, mail, ads)
- Fast parallel queries for performance