How We Estimate Website Worth: Our Formula, Data Sources and Limits

Last reviewed: October 27, 2026. This page describes the model that is live on the site today.

People ask “how much is my website worth?” every day. Site Worth Checker answers with a fast, automated estimate, plus the performance, SEO and security signals behind it. This page explains exactly how that estimate is calculated: the data we pull, how often it updates, the formula step by step, a worked example, and where the model falls short.

Estimates are models, not sale prices

A real sale price depends on revenue quality, churn, brand, legal risk and negotiation. Public tools cannot see your analytics or your bank balance. So our headline number is a model output: useful for research, comparison and education, but not a certified appraisal, a loan document or proof of AdSense income.

Data sources and how often they update

Each full report is cached for 24 hours, so repeat visits are fast. The “Re-check with fresh data” button runs every check again. We use Tranco instead of the retired Alexa rank; Alexa Rank Is Gone: How Website Worth Is Estimated in 2026 explains why.

The formula, step by step

Every domain goes through the same steps. Nothing is adjusted by hand for individual sites.

1. Look up the Tranco rank. Rank 1 is the most popular domain on the web. A subdomain with no rank of its own is credited with 2% of its parent domain’s estimated traffic.
2. Turn rank into visits. Monthly visits = 5,000,000,000 × rank^−0.9. For a sense of scale: rank 1,000 works out to about 10 million visits a month, and rank 1,000,000 to about 20,000. Daily visits = monthly ÷ 30.
3. Derive an authority score from the rank. A 1–100 score that falls as the rank number grows (the top 100 domains get 95). It becomes a multiplier: 0.5 + 2.5 × score ÷ 100, so between 0.5× and 3×.
4. Pick a revenue per visit. The baseline is $0.03 per visit, adjusted by site type: e-commerce and marketplaces about $0.05, social and video $0.036, news $0.024, government and education $0.018. CDNs, API hosts and other infrastructure domains get much lower rates, because their rank comes from machine traffic rather than people. The type comes from our list of known domains first, then from clues such as a detected store platform.
5. Estimate annual revenue. Annual revenue = monthly visits × 12 × revenue per visit × authority multiplier.
6. Apply a valuation multiple. Traffic value = annual revenue × 3.5. Sites above 1 million visits a month get 4×, and above 10 million get 5×.
7. Add domain and technical value. A small amount for things a buyer would also get: the domain name (shorter names, .com, and names without digits or hyphens score higher), domain age (from $100 for under a year up to $10,000 for 20+ years), SSL, PageSpeed performance and SEO scores, a fast server response, detected technologies, and basic content (title, meta description, mobile viewport, word count).
8. Combine and round. Estimated worth = traffic value + domain and technical value, rounded to a readable figure (for example, to the nearest $100 million above $1 billion). A reachable site with no traffic rank is valued at twice its domain and technical value, with a $500 floor.
9. Report daily revenue. Estimated daily revenue = annual revenue ÷ 365.

Two safeguards protect against bad data. If the rank lookup fails, we use the last good rank and mark the report as degraded. If a degraded check comes out below 10% of the last good value for that domain, we keep the last good value instead of publishing a broken number. Checks with no usable data are shown as temporarily unavailable and are never stored.

Worked example: wikipedia.org

Here is the wikipedia.org report as of October 27, 2026, worked through the steps above:

That example also shows the model’s main blind spot. Wikipedia is run by a non-profit and carries no ads, so the “revenue” is what a site with that traffic could earn, not what Wikipedia earns. Every estimate is a statement about potential, based on public signals.

Limits you should know about

• Small sites. Tranco only ranks the top one million domains. Sites outside it get no traffic value, so a small but profitable site will look cheap. If you know the profit, use the website value calculator, which applies profit multiples instead.
• Rank is not traffic. Tranco blends DNS, browser and link-based sources. Some domains rank high because of machine traffic, and deep in the list the sources are bucketed, so small rank differences mean little.
• Revenue is assumed. Revenue per visit and multiples are averages. Real earnings depend on niche, ad setup, products and audience location, and real multiples depend on risk and growth.
• Brand and company value are not included. We do not add market capitalisation or private financials, so the value of the business behind a big brand is outside the model.
• Blocked or slow sites. If a site blocks our requests or PageSpeed times out, technical value is lower and the report is marked as partial.

How to use the number wisely

• Compare similar sites directionally, not to the dollar.
• Re-check after major redesigns or traffic changes.
• Pair our estimate with first-party analytics and financials before buying or selling.
• Treat missing data or blocked crawlers as a reason for lower confidence.

Try it yourself

Enter any domain on the Site Worth Checker homepage and open the Worth Breakdown to see these steps applied to that site. For common questions about accuracy and ads, see the FAQ. If you are comparing tools, our SiteWorthTraffic alternative page explains how this model differs from Alexa-era calculators. Values remain automated estimates for informational and entertainment use.

Related reading: What Is Site Worth? and How Much Is My Website Worth? A Practical Guide.