At some point, everyone doing business online has to answer the same question about somebody else’s website: what is this thing actually making? You’re considering buying a site, or paying for a sponsored post on it, or partnering with its owner — and the number they quote you is, let’s be honest, marketing. The good news is that you can build a defensible estimate yourself, from public signals, in under an hour. Here’s the framework.
Step 1: Identify the model before the math
Revenue estimation fails when people jump straight to traffic numbers. Traffic means nothing until you know what it’s being converted into — and websites only have a handful of engines: display advertising, affiliate commissions, sponsored placements, products or subscriptions, and lead generation. Each converts a visitor into wildly different money: a display pageview might be worth a fraction of a cent, while a single insurance lead can be worth $50. Spend your first ten minutes just cataloguing which engines are visibly running — ad units, “affiliate disclosure” pages, a “write for us” or “advertise” page, a checkout, a quote form. If you want the full taxonomy of these engines and what each typically pays, this breakdown of how websites actually make money is the deepest public reference I know of.
Step 2: Get a traffic figure you can defend
Third-party tools (Similarweb, Ahrefs, Semrush) will all give you a number, and they’ll all disagree — sometimes by 5x. The workable approach: pull two or three, treat the lowest as your base case, and sanity-check it against reality. Does the site’s content volume support that traffic? Do its social profiles and comment sections show a human audience of matching size? A site claiming 100,000 monthly visitors with three comments and 200 followers has a story to explain. And in an era of bot inflation and manufactured metrics, you should actively verify that traffic is real before a single dollar moves — inflated numbers aren’t an edge case anymore; they’re an industry.
Step 3: Apply honest unit economics
Now multiply, conservatively. Display: monthly pageviews × RPM, where RPM runs roughly $2–$10 for general content and $15–$40+ for finance, legal, and B2B niches. Affiliate: estimate clicks to merchant (1–3% of relevant pageviews is typical), then apply the niche’s conversion and commission rates. Sponsored posts: the site’s own rate card, or comparable marketplace listings, times a realistic placement count. Products and leads: visible pricing times plausible conversion (0.5–2% of targeted traffic is honest for most niches). Build a range, not a number — a low case, a mid case, a high case — and be suspicious of your own high case.
Step 4: Subtract what sellers never mention
Gross revenue isn’t the story. Content costs (someone writes those articles), paid traffic (check whether rankings or ads drive the visits — bought traffic dies when the buying stops), platform dependency (a site earning 90% from one ad network or one affiliate program is one policy email from zero), and concentration (three articles generating 80% of traffic is fragility, not strength). A site grossing $3,000 a month on $2,200 of content and link costs is not a $3,000 business.
Step 5: Reconcile against the asking price
Content sites typically trade around 30–40× monthly profit — so run the seller’s price backwards. A $50,000 asking price implies $1,250–$1,700 in monthly profit; if your framework can’t get there on conservative assumptions, the gap is your negotiation, or your exit. The same logic works for advertising decisions: if a $400 sponsored post sits on a site your framework values at $300 a month of total revenue, you now know who the product really is.
The one-line version
Model first, verified traffic second, conservative multiplication third, costs fourth — and never accept a number you didn’t build yourself. The whole exercise costs you an hour. Skipping it has cost other people the entire purchase price.
