Similarweb Scraper: Website Traffic Checker & Competitors

For marketers, SEO agencies and analysts checking lists of domains. Get estimated monthly visits, category rank and up to 20 similar websites from Similarweb-powered SimilarSites, without an account.

About $0.0018 per useful domain. Checking 100 websites costs about $0.18, plus a $0.00005 start charge at the default memory. The three-domain example costs about $0.00545 at list price. Eligible plan discounts reduce the domain price. Competitor suggestions are included in that same price.

This is a partial-coverage scraper. It uses Similarweb's public SimilarSites surface. It does not currently retrieve the full Similarweb profile, global or country rank, engagement statistics, visit history, traffic-channel shares, country audience shares or keyword metrics. Those output fields are null. The source labels its traffic figure Monthly Visits but does not supply a reporting month, so dataMonth is also null. Read the source coverage before using this for a dated report.

What you get

Field	Meaning
verdict	A readable traffic summary, rounded to two significant figures
recordType	DOMAIN, a free NOTE, or the free SUMMARY
domain	The source's canonical website domain when it supplies a redirect
requestedDomain	The cleaned input domain that produced the result
title, description	The website metadata supplied by the public source
category	The source's category identifier
monthlyVisits	Estimated monthly visits, rounded to the nearest whole visit
categoryRank	The category-rank number shown by SimilarSites
similarSites	Up to 20 alternative domains, when requested
similarSites[].monthlyVisits	Each alternative's estimated monthly visits
similarSites[].similarityRank	Its order in the source's similar-site suggestions
similarSites[].similarityScore	The raw source Grade value, not the page's rescaled similarity percentage
similarSites[].category	The alternative's source category
similarSites[].topCountryRank	A country-rank number when supplied; the country is not identified
dataMonth	null: the reporting month is not supplied
scrapedAt	The UTC time this result was observed, not its reporting period
coverage	Always partial for this route
source, sourceUrl	The named source and its live SimilarSites page
url	The corresponding Similarweb website profile link
notes	Coverage information, including unavailable requested features

The following fields are retained for a consistent website-traffic schema but are unavailable on this source: globalRank, countryRank, country, visitsHistory, visitsChangePercent, bounceRate, pagesPerVisit, avgVisitDurationSeconds, topCountries and topKeywords. trafficSources contains direct, search, social, referrals, mail and paid, all with null values. A missing value is never replaced with zero or estimated from another metric.

Use with AI agents (MCP)

Connect an MCP client to https://mcp.apify.com and request this Actor by its full name, vant47/similarweb-traffic-scraper. A suitable request is: "Check the monthly visit estimates, category rank and similar websites for github.com, shopify.com and apify.com. Tell me which requested metrics are unavailable."

Send the input below and read the DOMAIN rows and SUMMARY. Preserve the source and coverage notes; do not infer a reporting month or missing engagement statistics.

{
  "domains": ["github.com", "shopify.com", "apify.com"],
  "includeCompetitors": true,
  "includeKeywords": false,
  "maxItems": 3
}

How to use it

1. Add the domains you want to monitor. A full website URL or Similarweb website profile also works; the Actor cleans it to a domain.
2. Choose whether to include similar sites and set the maximum number of useful domain rows. Set an Apify spending limit if you want a firm run budget.
3. Start the run, open the Website traffic dataset view and export JSON, CSV, Excel or another Apify-supported format. Nested competitor lists are easiest to use in JSON.

The default example checks three domains. Larger lists run serially, with about three seconds between source requests. Source delays and retries can increase the duration.

Inputs

Input	Default	How to use it
domains	Empty in an API call; three domains in the Console example	Array of domains, website URLs or Similarweb profile URLs
includeCompetitors	true	Return up to 20 similar sites in each useful domain row
includeKeywords	false	Include keywords only when exposed by the free source; currently remains null
maxItems	100	Maximum distinct useful domains, from 1 to 10,000
proxyConfiguration	Automatic	Datacenter public page first, JSON as an alternative, then residential only if needed

github.com, https://www.github.com/features and its Similarweb profile are equivalent inputs. Paths and query strings do not select page-level traffic.

The Console uses a string-list editor for domains. Use arrays in API requests. The optional urls and startUrls array aliases are accepted when domains is absent. These aliases do not add separate output units. Duplicate normalized domains are delivered and charged once in a run. Canonical redirects returned by the source are also deduplicated.

{
  "domains": [
    "https://www.amazon.co.uk/",
    "amazon.de",
    "lemonde.fr",
    "abc.net.au",
    "takealot.com"
  ],
  "includeCompetitors": false,
  "maxItems": 5
}

Regional domains are separate websites. Estimates are website-wide; this is not a country audience filter.

Example output from the live source

This shortened example uses values read on 29 October 2026. The live GitHub page displayed 513M monthly visits and category rank 4. The source JSON provided the more precise estimate shown below. Similar-site suggestions may change between runs.

{
  "verdict": "About 510M monthly visits; category rank #4; 20 similar sites. Reporting month unavailable.",
  "recordType": "DOMAIN",
  "domain": "github.com",
  "requestedDomain": "github.com",
  "title": "GitHub · Build and ship software on a single, collaborative platform",
  "category": "Computers_Electronics_and_Technology/Programming_and_Developer_Software",
  "monthlyVisits": 513197610,
  "categoryRank": 4,
  "globalRank": null,
  "countryRank": null,
  "country": null,
  "dataMonth": null,
  "visitsHistory": null,
  "bounceRate": null,
  "pagesPerVisit": null,
  "avgVisitDurationSeconds": null,
  "topCountries": null,
  "topKeywords": null,
  "similarSites": [
    {
      "domain": "stackoverflow.com",
      "monthlyVisits": 69124392,
      "similarityRank": 1,
      "similarityScore": 0.9368279293465819,
      "topCountryRank": 968
    }
  ],
  "coverage": "partial",
  "source": "Similarweb-powered SimilarSites",
  "sourceUrl": "https://www.similarsites.com/site/github.com",
  "url": "https://www.similarweb.com/website/github.com/"
}

Use monthlyVisits for calculations and retain its undated, estimated status. A competitor's topCountryRank has no country code and is not necessarily for your country.

What is charged

One Domain charge covers a distinct useful website result. A result is useful when the source supplies a monthly visit estimate, category rank or requested competitor data. Website metadata alone does not trigger a domain fee. Competitor rows are nested in the same result and have no separate charge. Turning competitors off does not change the domain price.

Job	Domain charges at list price
Three-domain example	$0.0054
100 useful domains	$0.18
1,000 useful domains	$1.80
Repeated copies of one normalized domain	One domain charge
A domain with no useful public data	No domain charge

Add the small platform start charge to those examples. At default memory it is $0.00005. Bronze has a 5% domain discount, Silver 10%, and Gold and higher 20%. The user does not separately pay the Actor's platform usage. No outside API subscription, API key, account cookies or paid unblocking service is required.

maxItems limits useful domain rows. The Apify maximum charge is a separate money limit. The Actor checks the remaining allowance before fetching another domain and stops cleanly when there is no room for another domain charge. A source failure, invalid input or missing record produces a free note. A run that cannot afford a domain may still have the platform start charge.

Monitor website traffic over time

Save a task with the domain list and create a weekly or monthly Apify schedule from it. Keep the scrapedAt, domain, monthlyVisits, categoryRank and sourceUrl fields in your destination table. One result per domain per scheduled run gives you a history of observations.

Because the source does not reveal the underlying traffic month, two runs on different dates may contain the same monthly estimate. Do not call collection dates traffic months, sum repeated observations as separate months, or calculate a month-over-month growth rate from them. This Actor returns no historical visit series. It is useful for checking the currently displayed estimate and discovering related websites.

Use a webhook to send the dataset to a spreadsheet, warehouse, Make, Zapier or n8n. Filter numerical reports to DOMAIN rows. Keep notes and the SUMMARY, which reports delivered domains, duplicates, missing data, invalid inputs, failures and the stopping reason.

API examples

Keep your Apify API token in an environment variable. The source requires no credentials.

curl --request POST \
  --header "Authorization: Bearer $APIFY_TOKEN" \
  --header 'Content-Type: application/json' \
  --data '{"domains":["github.com","shopify.com"],"maxItems":2}' \
  'https://api.apify.com/v2/acts/vant47~similarweb-traffic-scraper/run-sync-get-dataset-items'

import os
from apify_client import ApifyClient

client = ApifyClient(os.environ["APIFY_TOKEN"])
run = client.actor("vant47/similarweb-traffic-scraper").call(
    run_input={"domains": ["github.com", "shopify.com"], "maxItems": 2}
)
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    if row.get("recordType") == "DOMAIN":
        print(row["domain"], row["monthlyVisits"], row["dataMonth"])

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('vant47/similarweb-traffic-scraper').call({
  domains: ['github.com', 'shopify.com'], maxItems: 2,
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
const domains = items.filter(row => row.recordType === 'DOMAIN');
console.log(domains);

For large lists, start an asynchronous run and read its dataset when finished. Rebooted runs recover completed rows and acknowledged charges without delivering or charging those rows again.

Reliability and limits

The route is public and signed out. It does not visit the supplied business websites, use account sessions, solve CAPTCHAs or collect individual visitors. A plain HTTPS call reads the JSON embedded in the source's public page. Its own public JSON endpoint is an alternative surface, with fresh-session retries and request pacing. No full browser is needed for the delivered product.

Cloud acceptance on 29 October 2026 returned useful traffic for 35 of 35 domains, across retail, news, software, travel and regional websites. Ten live-page checks matched visits, category rank and competitor order. Money-limit checks delivered one paid domain under a one-domain cap and none below it. A real reboot recovered three unique domains without a second charge. These checks cover the available subset.

Small or new domains may have no data. Metadata can lag website changes. Check canonical redirects using requestedDomain and domain. Similarity does not verify a business relationship. Country-rank numbers without country labels are not a common-country leaderboard.

FAQ

Is this the full Similarweb profile?

No. It is a working partial route through Similarweb-powered SimilarSites. The output contains monthly visit estimates, category ranks, website metadata and similar-site suggestions when available. The missing profile fields stay null.

Why is the reporting month null?

The source does not publish it in this payload or label it on the profile. The observation date is recorded separately. A precise data month must not be inferred from today's date.

Does includeKeywords unlock keyword data?

No. It requests keywords only when exposed by the free source. On the current route, topKeywords remains null, a coverage note explains why, and there is no extra charge.

Are monthly visits unique people or page views?

The source calls the metric Monthly Visits. It is not a unique-visitor count or page-view count. Do not rename the field to either of those measures.

Can I request country-specific traffic?

No country filter or audience breakdown is available here. A regional domain such as amazon.co.uk is a separate website input, not a country filter for amazon.com.

What happens if the source blocks a request?

The Actor tries the alternative public surface and fresh sessions with ordinary pacing. If no useful record is returned, it writes an uncharged note and continues. Invalid input also becomes a note rather than a failed run.

Are duplicates or failed domains charged?

There is no domain fee for either. Duplicate normalized or canonical domains count once in a run. The small platform start charge remains. Separate scheduled runs each deliver a new observation and can each incur a domain fee.

Similarweb scraper, website traffic checker, monthly traffic estimates and competitor websites from a public, signed-out source.