Realistic ad revenue, and tracking where traffic came from

How the numbers work

RPM is revenue per thousand pageviews, and it is the only figure worth tracking because it folds together everything else — fill rate, click-through, cost per click and the number of ad units.

The arithmetic is simply pageviews divided by a thousand, multiplied by RPM. A site with 50,000 monthly pageviews at a $5 RPM earns around $250 a month.

That last row surprises people who expect scale to improve rates. It does not, materially — a site ten times larger earns roughly ten times as much at a similar RPM. What raises RPM is who is reading and what they are reading about.

Why geography dominates

Advertisers bid for audiences they can sell to, and the value of a reader tracks the purchasing power and advertising market of the country they are in.

The practical result is a wide spread. Readers in the United States, United Kingdom, Canada, Australia and parts of Western Europe command several times the rates of the global average, and the gap against the lowest-value markets can be an order of magnitude.

This is worth knowing before drawing conclusions from a revenue report. A month where traffic rose and revenue fell usually means the traffic mix shifted, not that anything broke. Look at revenue by country before looking for a technical explanation.

Subject matter, and the seasonal pattern

Advertiser competition varies enormously by topic. Insurance, legal services, finance, business software and health carry high rates because the value of a converted customer is high. General interest, entertainment and hobby content sit at the bottom.

Seasonality then moves everything together: rates climb through the final quarter as retail budgets are spent, peak in the run-up to Christmas, and fall sharply in January when those budgets reset.

The January drop catches every new publisher. Revenue can halve with no change in traffic and no fault of yours. It recovers through the year, and reading a January report as evidence of a problem leads to changes that fix nothing.

Being realistic

Most sites earn less from display advertising than the case studies suggest, and the reason is selection: people write up the outcomes worth writing up.

A useful sanity check is that a site needs substantial traffic before display income is meaningful. At a $5 RPM, 100,000 monthly pageviews is around $500 a month — a decent supplement, and not a salary.

AdSense Earnings Calculator lets you run the arithmetic for your own numbers rather than someone else's. Enter a realistic RPM for your audience rather than an aspirational one, and the answer is usually sobering and useful.

Tracking where traffic came from

Ad revenue by source is only knowable if the sources are tagged, which is what UTM parameters do — they attach campaign information to a link that analytics reads on arrival.

Five parameters exist and three matter.

• utm_source — where it came from: newsletter, twitter, partner-site.
• utm_medium — the channel type: email, social, cpc, referral.
• utm_campaign — which campaign: spring-launch, weekly-2026-08.
• utm_term — paid search keyword. Rarely needed otherwise.
• utm_content — distinguishes variants: header-link against footer-link.

Source and medium are the pair that analytics groups by, so getting those consistent matters more than anything clever in the campaign name.

The mistakes that corrupt the data

1. Inconsistent case. UTM values are case-sensitive: Facebook and facebook become two separate rows that never recombine. Use lowercase for everything, always.
2. Tagging internal links. A UTM on a link between your own pages starts a new session and overwrites the original source, so the visitor who arrived from a newsletter is recorded as arriving from your own footer.
3. Inconsistent medium values. email, e-mail and Email fragment the same channel into three.
4. Spaces in values, which get encoded as %20 and appear as such in reports. Use hyphens.
5. Personal data in parameters. Email addresses or names in a URL end up in analytics, server logs and referrer headers. Never do this.

The second is the most damaging and the least obvious, because everything appears to work. Attribution quietly collapses onto your own site, and the channel that actually brought people in shows as producing nothing. UTM Builder assembles and encodes the parameters correctly; the discipline of keeping values lowercase and consistent is the part that has to come from you.

Frequently asked questions