Similarweb Scraper: Traffic & Competitor Data in 2026

Traffic is the number competitors never publish

Companies disclose revenue when they must and headcount when it flatters them. Website traffic they almost never share, yet it is the clearest outside signal of how a digital business is doing. Similarweb fills that gap with monthly estimates of visits, rankings and traffic sources, the figures quoted in marketing decks and investor memos.

The catch: similarweb.com shows one profile at a time, and the official API is a paid product. For a benchmark of 50 competitors or a map of a whole category, neither is practical.

What you can extract

One flat row per website, 76 fields:

• Traffic: monthly visits, month-over-month change, 3 months of history
• Rankings: global, country and category rank, each with 3 months of history
• Engagement: bounce rate, pages per visit, visit duration
• Traffic mix: leading source and its share, organic vs paid search split
• Audience: top countries, age bands, gender split, sites and topics visitors also browse
• Sources: top keywords with CPC, referrals, social networks, ad partners
• Similar sites: the 10 closest websites, each with an affinity score
• Company: name, year founded, headquarters, employee and revenue ranges

The technical challenge

Similarweb profiles are built for people browsing, not for datasets. Bulk extraction runs into three problems:

• Normalization: “4.2M” visits and percentage bounce rates must become numbers before anyone can sort or join them, and a missing value must stay empty, not become a guess.
• Partial visibility: Similarweb keeps part of its referral and advertising lists for paying users. A clean extraction returns the visible entries and keeps the complete counts.
• Speed: a headless browser per domain costs several seconds each; reading the data the pages themselves load is far faster.

DIY or a ready-made Actor?

Our Similarweb Scraper on Apify takes domains, keywords (10 matching sites each), or a country and category from the Top Websites ranking (67 countries, 211 categories, movers of the top 100 included). It can also read the 10 similar sites of every row as full rows. No Similarweb account or API key is needed, and 10 domains come back in about 2 seconds. Pay-per-result at about $0.003 per domain.

It is not a drop-in replacement for the paid API, since data reserved for subscribers stays out; our API vs web scraping comparison covers that trade-off.

A practical workflow

1. Define the set: competitors, a lead list of domains, or a ranking when you don’t know the players yet.
2. Expand it: switch on similar websites to surface the competitors you missed; each row records its origin.
3. Schedule monthly: Similarweb refreshes once a month, so a run early each month builds a clean time series.
4. Join and act: merge on domain with your CRM or portfolio data, sort by visits, and flag the sites marked rising.

Use cases that pay off

• SEO and marketing benchmarks: visits, bounce rate, traffic mix and organic vs paid share across your competitors, refreshed monthly.
• Market mapping: the top sites of a category in a given country, with this month’s movers flagged.
• Lead scoring: domains ranked by traffic, geography, employee range and revenue range.
• Investment screening: traffic and rank history for portfolio companies and targets.

The legal boundaries

• Public profiles only: the scraper reads what similarweb.com shows any visitor and does not bypass logins, paywalls or CAPTCHAs.
• Estimates, not audits: every figure is Similarweb’s model unless the site shares its own analytics, which the row flags. Label numbers as estimates and verify them before regulated use.
• Terms still apply: you remain responsible for complying with Similarweb’s terms of service. The data describes websites and companies, and demographics are aggregate shares.