SimilarWeb API (Scraper)

What does SimilarWeb API do?

SimilarWeb API extracts website traffic analytics for any domain - visits, rankings, engagement, traffic sources, top countries, and keywords. Paste a list of domains (or full URLs), run the Actor, and download clean JSON / CSV / Excel from the dataset.

Built for bulk runs (up to 100,000 domains) with a fast default path and an automatic reliability fallback when needed. Large lists are auto-split into ~3,000-domain child runs on Apify, with auto-heal under blocks.

This Actor is not affiliated with SimilarWeb. Metrics are third-party estimates - use them as directional intelligence, not as audited truth.

Key features

• Bulk domain analysis - up to 100,000 domains per run (invalid / duplicate inputs filtered before work starts)
• Auto-sharding - runs above ~3,000 domains split into child runs; one parent dataset
• Auto-heal - on blocks: backoff, lower concurrency, refresh sticky extension session once per wave
• Traffic & engagement - monthly visits, bounce rate, pages per visit, average visit duration
• Rankings - global rank, country rank, and category rank when available
• Traffic sources - direct, organic search, paid search, social, referrals, mail, display
• Geography - top countries with traffic share
• Keywords - top keywords when SimilarWeb exposes them (toggleable)
• Success-only billing - failed domains stay in the dataset with error, but are not charged
• Fast default mode - lightweight path tuned for speed and low compute (512 MB)
• Automatic fallback - optional browser path when the fast path is blocked (fetchMode: auto)
• Export anywhere - JSON, CSV, Excel, HTML from the Apify dataset

Why Clovys

Clovys is a nonprofit organization driven by a single mission: to make the internet safer, faster, more reliable, and universally accessible.

We believe that data should not be locked behind exorbitant paywalls or complex technical barriers. By building lightweight, highly optimized Apify Actors, we help developers, researchers, and organizations access the information they need to innovate without friction.

As a nonprofit initiative, we do not optimize for margins. Our pricing is kept as low as possible and is strictly calculated to cover the raw infrastructure costs of data extraction - such as Apify compute units and proxy bandwidth.

For SimilarWeb specifically, Clovys focuses on delivering the most robust, reliable, and fast extraction path we can maintain: a production-grade default mode, retries, auto-heal, and a fallback strategy so bulk runs stay usable at scale.

Use cases

• Competitive analysis - track competitors' traffic, rankings, and channel mix
• Media & marketplace research - score publishers, affiliates, or inventory by audience size
• Marketing ROI - see where traffic comes from (search, social, direct, paid) before allocating budget
• CRM / lead enrichment - attach traffic and rank signals to accounts
• Investment & due diligence - directional digital footprint for a brand or portfolio of sites
• Content & SEO strategy - combine visits, geo share, and top keywords for prioritization

How to use

1. Open SimilarWeb API in Apify Console
2. Paste domains or URLs in Domains (e.g. apify.com, https://www.lemonde.fr)
3. Keep defaults for production (fetchMode: extension, 512 MB)
4. Click Start
5. Download results from the Dataset tab (JSON, CSV, Excel, ...)

Input

Field	Type	Default	Description
domains	string[]	-	Domains or URLs to analyze (required, up to 100,000)
fetchMode	string	extension	extension (fast), auto (fast then browser), browser, http
maxConcurrency	integer	5	Parallel workers (extension/http up to 12; browser keep 1-3)
cacheTtlHours	integer	72	Reuse successful results within the run (0 = off)
proxyConfiguration	object	-	Optional. Datacenter often enough for extension; use RESIDENTIAL if identity is blocked or for auto/browser
includeKeywords	boolean	true	Include top keywords when present
includeCountries	boolean	true	Include geo traffic share
maxRetries	integer	1	Retries per domain on block/extract errors

Output example

You can download the dataset in JSON, HTML, CSV, or Excel.

{
 "domain": "apify.com",
 "title": "apify.com Traffic Analytics, Ranking & Audience",
 "globalRank": 9646,
 "visits": 4201537,
 "bounceRate": 0.37,
 "pagesPerVisit": 7.43,
 "avgVisitDuration": 300,
 "trafficSources": {
 "direct": 0.44,
 "searchOrganic": 0.35
 },
 "topCountries": [
 { "country": "United States", "value": 0.21 }
 ],
 "scrapedAt": "2026-09-24T12:00:00.000Z",
 "error": null
}

Failed domains still appear with error set and metrics null - they are not billed.

Data fields

Field	Description
domain	Normalized hostname
title	Site title / SimilarWeb label
globalRank	Worldwide traffic rank
visits	Estimated monthly visits
bounceRate	Bounce rate
pagesPerVisit	Pages per visit
avgVisitDuration	Average visit duration (seconds)
trafficSources	Channel mix (direct, search, social, ...)
topCountries	Top geos with share
topKeywords	Top keywords when enabled / available
error	Error message if the domain failed

Pricing

Pay-per-event pricing (platform usage is billed separately by Apify):

Event	When charged	Price
result	Each successful domain written to the dataset	$0.005 ($5 / 1,000)
Actor Start	Once when the run starts (scaled by memory GB)	$0.001

Example: 100 successful domains ~ $0.50 in PPE result (+ Actor Start + Apify platform usage).

Only successful results are charged as result. Failed domains are free on the PPE event.

Recommended settings

Goal	Settings
Production / Store	fetchMode: extension, 512 MB, concurrency 5-10
Max reliability	fetchMode: auto, RESIDENTIAL proxy if needed, 2048 MB
Browser-only	fetchMode: browser, concurrency 1-2, 2048 MB
Huge lists	up to 100,000 domains - auto-shard handles the rest

FAQ

How many domains per run? Up to 100,000. Lists above ~3,000 are auto-split into child runs; invalid / duplicate inputs are dropped before sharding. Parallelism is controlled by maxConcurrency.

What if a domain fails? It is stored with error. Other domains continue. Failed items are not billed as result.

Do I need expensive proxies? Often no for fetchMode: extension with Apify datacenter / automatic proxy. If extension identity returns 403, switch to RESIDENTIAL and/or fetchMode: auto.

What does auto-heal do? On block waves it slows down, lowers concurrency, and refreshes the sticky extension session once per wave - so the run keeps going instead of burning proxy on blind retries.

Is this official SimilarWeb data? No. This Actor is independent and unaffiliated. Numbers are estimates from public SimilarWeb surfaces.

Support

Questions, bugs, or feature requests: use the Actor Issues tab on Apify, or contact Clovys through the Store profile.