Case: +$2.70 RPM from adding 3 header-bidding partners Header bidding = auctioning each slot to many exchanges in parallel before the ad server call. Tech blog, n=1, 28d. Demand partners 4 → 7 Bid den…
Case: cutting 2 ad units RAISED revenue EPMV = earnings per thousand visitors, counts the whole session not one pageview. Lifestyle site, n=1, 35d test, 9 units → 7 units. Page RPM $11.20 → $10.10 (lo…
Case: a single sticky anchor unit added $2.05 mobile RPM Finance blog, n=1, 28d, mobile only (73% of traffic). Added one bottom-anchored 320×100 sticky ad. Nothing else changed. Mobile session RPM $18…
Case: ad refresh at 30s beat 20s by $1.10 RPM News site, n=1, 60d split test. They ran a viewability-gated refresh (reload ad only when slot is on screen) and tested the interval. 20s refresh RPM $14.…
Case: +$3.20 session RPM from one lazy-load tweak Recipe publisher, n=1 site, 41-day before/after. Viewability (% of ad impressions actually seen ≥1s) sat at 48% because ads loaded eagerly above the f…
Benchmark drop: session RPM by traffic source Where the visitor comes from prices the session before any ad loads. — Organic search $1.00 (index) ▮▮▮▮▮▮▮▮▮▮ — Direct $0.92 ▮▮▮▮▮▮▮▮▮ — Email $0.81 ▮▮▮▮…
