RPM Benchmarks: What a Domain Visitor Is Worth
RPM, revenue per thousand visits, is the honest way to talk about what domain traffic earns, because it survives comparisons across names, networks and months. It is also where domain reality diverges hardest from content-site intuition: a blog and a parked domain can receive the same thousand visitors and earn wildly different amounts. This piece anchors on the few numbers the industry actually agrees on, builds clearly labeled illustrative ranges on top of them, and shows you how to compute the only benchmark that finally matters: your own.
What numbers can we actually anchor on?
Public, agreed-on figures in domain monetization are scarce, so it pays to be explicit about the ones that exist. The NamePros consensus for parking ordinary names without genuine traffic is $3-5 per month per domain as of mid-2026. Parked type-in traffic typically monetizes at a few cents per visitor at best, with most visits earning nothing because no ad gets clicked. Sedo, the largest parking operation, pays owners an 80-90% revenue share, so the split is not where the money leaks. And industry consensus puts portfolio sell-through around 1-2% a year, which means for most holders the occasional sale, not the click stream, is the payday, and parking merely offsets renewals. Everything else you will read, including the table below, is synthesis on top of those anchors.
Illustrative benchmarks by traffic type and geography
The ranges below are illustrative planning figures derived from the anchors above and from how parking economics work; they are not measurements, and any individual domain can land outside them. They assume human visitors after bot filtering.
| Traffic segment | Illustrative value per 1,000 visits | Why |
|---|---|---|
| Bot and monitor hits | $0 | Filtered by every serious network; clawed back when not |
| Mixed type-in, low-bid geos | A few dollars | Low click rates on landers times low advertiser bids |
| Generic type-in, tier-one geos | Roughly $10-50, taking a few cents per visitor as the ceiling | Real buyers behind the clicks, but landers convert weakly |
| Commercial-intent type-in, tier-one geos | Above the generic range when keyword bids are strong | Insurance, hosting or finance names attract genuine ad competition |
| Live referral clicks | Depends on destination; often beats parking when redirected to a matching offer | Intent borrowed from the citing page monetizes poorly on a generic lander |
| Residual search visits | Rarely material | The slice decays toward zero on any dropped domain |
Why geography multiplies the number
Advertisers bid on visitors they can actually sell to, so the same click is worth different amounts depending on who is clicking and from where. A parked name whose visitors sit in the United States, United Kingdom, Canada, Australia or the richer EU markets earns a multiple of an identical name pulling equivalent volume from low-bid regions, because the auction behind the lander clears at higher prices. This is why a raw "traffic" figure without a geography breakdown is close to meaningless when pricing a domain, and why parking dashboards report earnings by country. When you evaluate a listing, ask where the visitors are, not just how many there are; two names with identical counts can differ several-fold in revenue on geography alone.
Redirects reset the ceiling
Parking RPM is a floor, not a verdict, because a lander is the least persuasive page a visitor can land on. The same measured visitors pointed at a genuinely matching offer or rebuilt resource can earn an order of magnitude more, the arbitrage we detailed in redirecting domain traffic to offers: as an illustrative example there, 300 matched visits a month converting at 1% into a $20 lead payout is $60 against roughly $5 parked. The catch is match quality, which most domains lack, and program rules, which forbid lazy setups. So read the table above as what traffic earns passively; active monetization is a different ceiling with different work attached.
How to benchmark your own portfolio
Compute RPM per domain, not per portfolio: revenue times 1,000 divided by visits, over at least 90 days to smooth the lumpiness of small numbers. Use the visit count from your own logs or the parking dashboard's filtered human count, never raw hits, or bots will flatter the denominator into nonsense. Then act on the spread. Names earning below the $3-5 monthly floor with no traffic story are renewal-fee liabilities and sale candidates, the arithmetic covered in is domain parking dead. Names beating the floor deserve a network split test, per our ad network comparison, and the outliers deserve investigation: where do the visitors come from, and what would they be worth redirected or rebuilt? When shopping for new names, treat every estimate column, including the traffic estimates on research platforms such as DomCop, as a sorting device, and assume your measured RPM will be decided by type, geography and match, exactly as above.
Reading the table against a real listing
An illustrative walkthrough shows how the ranges earn their keep. Suppose a listing claims 2,000 visits a month. Apply the standard bot discount for long-expired names, 30-70% of raw hits, and call it 1,000 human visits; the seller's own screenshot shows most of them outside the top bid tiers, and the name is generic rather than commercial. The low rows of the table put that inventory at a few dollars per 1,000 visits parked, so perhaps $5-15 a month. If the seller wants a $500 traffic premium, the payback period at the parked floor runs to years before renewals are even counted; the bid only makes sense if verification upgrades the story, tier-one geography or a genuine offer match, in which case you redo the arithmetic with the higher rows. That is the entire use of a benchmark table: it converts a claimed number into a payback period you can argue about. Run it on every listing before bidding, and write the assumptions next to the bid so future you knows which guess was wrong.
The walkthrough's arithmetic in one picture: strip the bots, then price the remainder at parked rates (illustrative)
Frequently asked questions
What is a good RPM for a parked domain?
Anything holding double digits per 1,000 human visits is strong for pure parking, since that implies a few cents per visitor, the practical ceiling for lander clicks. Most parked names sit far below that, which is why the $3-5 monthly consensus exists.
How much is a single domain visitor worth?
Parked: usually a fraction of a cent to a few cents, depending on geography and keyword intent. Redirected to a genuinely matching offer, or converted into a customer on a rebuilt site, the same visitor can be worth dollars. The visitor did not change; the destination did.
Why do my numbers differ from every published benchmark?
Because portfolios differ in the three inputs that matter: traffic type after bot filtering, geography mix and keyword commerciality. Published figures, including the illustrative ranges here, only bound the plausible; your 90-day measured RPM is the real number.
Is buying domain traffic by RPM arbitrage profitable?
Only when the purchase price respects the floor. Paying for claimed traffic at optimistic RPMs is the classic loss; the discipline is to price on verified human visits at parked rates, then treat any redirect or rebuild upside as margin rather than as the plan.