SimilarWeb API Alternative: Traffic and Rank Data Cheaply

SimilarWeb has the data — global rank, estimated monthly visits, traffic-source breakdowns, top referrers, audience geography — and it has an API to deliver it. The catch is that the API sits behind an enterprise sales motion. There’s no self-serve developer key, no metered pay-as-you-go tier, and pricing lands in the territory of four- to five-figure annual contracts negotiated with a sales rep. For a team that just wants to enrich a few thousand domains with a rank and a visits estimate, that’s wildly out of proportion. The useful thing most people miss: SimilarWeb already shows much of that data on its public website overview pages, free to anyone with a browser. This guide covers the SimilarWeb API alternative built on that public surface.

Why the official API is the wrong tool for most teams

To be fair about it: if you’re a market-intelligence firm ingesting millions of domains with SLAs and historical depth, the SimilarWeb enterprise API is the right purchase. The mismatch is for everyone else:

• No self-serve tier. You can’t sign up and get a key. You talk to sales, and the conversation starts in the low-to-mid five figures per year for meaningful volume.
• Annual commitment. It’s a contract, not a metered call. A one-off enrichment project doesn’t justify it.
• Opaque entitlements. Which metrics and how much history you get depends on your plan tier, negotiated case by case.

If your actual need is “give me rank and a rough visits estimate for this list of competitors,” paying enterprise pricing for it is absurd — and you don’t have to.

What the public overview page exposes

Every domain SimilarWeb tracks has a free public page:

https://www.similarweb.com/website/{domain}/

That page is React-hydrated, and the data it renders is delivered as JSON to the front-end (visible in the page’s embedded state and the XHR calls the app makes on load). What you can read without any login or contract:

• Global rank, country rank, and category rank.
• Total visits for the most recent month, plus a short trailing trend.
• Engagement metrics — visit duration, pages per visit, bounce rate.
• Traffic-source split — direct, referrals, search (organic vs paid), social, mail, display.
• Top referring sites and top destination sites.
• Top countries by share of traffic.
• Top search keywords (organic and paid, where surfaced) and top social networks.
• Category and a “similar sites” list.

It’s the same shape of data the API returns, just rendered for the free web experience rather than served under contract.

How the public surface actually works

The overview page hydrates from internal endpoints under the site’s own API path (the calls the front-end fires, returning the JSON that populates each widget). Practically, you have two ways in:

1. Parse the embedded state in the server-rendered HTML — the initial data blob the React app boots from carries the headline metrics.
2. Read the XHR responses the page requests on load for the deeper widgets (sources, geography, keywords).

Either way, no authentication is involved for the free public metrics — it’s the data SimilarWeb chooses to show window-shoppers to upsell them.

▶ Try the SimilarWeb Scraper on Apify — rank, visits, and traffic-source data from public overview pages, no enterprise contract required. No auth required.

Rate limits, accuracy, and how to live with both

Two constraints to be honest about.

Rate limiting. SimilarWeb protects its public pages against bulk access. A single IP requesting hundreds of domains a minute will get challenged or rate-limited. Keep throughput modest — on the order of a request every few seconds per IP — and rotate IPs for larger lists.

Accuracy. SimilarWeb’s numbers are modeled estimates, not measured truth, on both the API and the public page. The free public figures are sometimes rounded or bucketed more coarsely than the paid API’s. Treat visits and shares as directional — excellent for ranking domains relative to each other, less so for an exact headcount. For competitive comparison (which is most use cases), relative accuracy is what matters and the public data delivers it.

A clean output schema

One row per domain:

{
  "domain": "example.com",
  "global_rank": 4821,
  "country_rank": { "country": "US", "rank": 1320 },
  "category": "Computers Electronics and Technology > Programming and Developer Software",
  "category_rank": 64,
  "total_visits": 18400000,
  "visits_trend": [16200000, 17100000, 18400000],
  "avg_visit_duration_sec": 412,
  "pages_per_visit": 5.1,
  "bounce_rate": 0.38,
  "traffic_sources": {
    "direct": 0.41,
    "referrals": 0.07,
    "search_organic": 0.39,
    "search_paid": 0.02,
    "social": 0.06,
    "mail": 0.03,
    "display": 0.02
  },
  "top_countries": [
    { "country": "US", "share": 0.34 },
    { "country": "IN", "share": 0.11 }
  ],
  "top_referrers": ["news.ycombinator.com", "reddit.com"],
  "top_keywords": ["example tool", "example api"],
  "scraped_at": "2026-06-05T12:00:00Z"
}

Use domain as the natural key, normalized to the registrable domain (strip www., lowercase) so your dataset dedupes cleanly.

Use cases

• Competitive benchmarking — rank a set of competitors by estimated visits and watch the trend month over month.
• Lead qualification — enrich inbound or prospect domains with a traffic estimate to prioritize the ones with real reach.
• Channel analysis — the traffic-source split tells you whether a competitor lives on SEO, paid, or social, which shapes your own strategy.
• Market mapping — pull the “similar sites” and category rank to chart who actually competes in a niche.
• Acquisition / investment screening — a quick public traffic read is a cheap first filter before deeper diligence.

Build it yourself vs. a managed actor

The headline metrics are easy to grab from the embedded HTML state. The deeper widgets (sources, geography, keywords) come from separate XHR calls whose paths and payloads SimilarWeb changes as it refactors the front-end, and the whole thing sits behind bot protection that escalates with volume. Maintaining a working SimilarWeb API alternative means keeping up with those front-end shifts and managing IP rotation — exactly the upkeep a managed actor absorbs so your pipeline doesn’t break every time SimilarWeb ships a redesign.

Common pitfalls

• Treating estimates as measurements. Use the numbers comparatively. Don’t report “exactly 18.4M visits” to a stakeholder as fact.
• Mixing rounded public figures with paid figures. If you ever blend public-page data with someone’s API export, document which is which — they round differently.
• Domain vs. subdomain. SimilarWeb tracks blog.example.com separately from example.com. Decide which you mean and be consistent.
• Thin-data domains. Low-traffic sites may have no estimate at all (SimilarWeb hides figures below a confidence threshold). Handle empty gracefully.
• Over-requesting from one IP. The fastest way to get blocked. Throttle and rotate.

Wrapping up

SimilarWeb’s traffic data is genuinely valuable, and its API is genuinely enterprise-priced — a four- to five-figure annual contract you negotiate with sales, with no self-serve tier. For the common need of enriching a domain list with rank, visits, and traffic sources, that’s overkill, and the data you want is already on the free public overview pages. A SimilarWeb API alternative built on that surface gets you directional rank and traffic figures without the contract. If you’d rather not chase the front-end’s shifting XHR endpoints, a managed actor returns it already normalized.

▶ Open the SimilarWeb Scraper on Apify — global rank, monthly visits, and traffic-source breakdowns from public pages. Enrich a domain list without an enterprise plan. Pay per domain.