The AdSense playbook
How to get the most out of AdSense
Most advice for raising AdSense revenue tells you to add more ads. That is the one move that reliably backfires. Here is the opposite: a ranked playbook for earning more from the same traffic, in the order that actually moves the number, with the data behind every step linked so you can check it.
Start here
What you can change, and what you cannot
Three things set most of your RPM and you cannot tweak them. Who is viewing: advertisers bid in local markets, and on our network 97% of revenue is US traffic, the high-value end. Your topic's demand: it varies by vertical, though less than people think, our two niches pulled nearly the same cost per click ($0.16 and $0.19). And how much premium demand you can reach: at small scale you cannot reach much, 88% of our AdSense revenue comes from Google's own advertiser pool.
You cannot manufacture better geography or more demand by fiddling with your ad setup. What you can change is whether your ads are seen, and whether your layout costs you the rankings that bring the traffic. Every step below is one of those two. The receipts are in the full teardown.
The playbook
In order of leverage
- 1. Put ads where they are seen, not everywhere. Favor a dismissible anchor and in-content placements; stop adding units in dead spots like the page bottom or a deep sidebar. Across our network the anchor is viewable 88% of the time and a page-bottom unit just 28%, a 60-point spread that has nothing to do with how many ads you run. Click-through follows position too: 4.8% at the anchor against 0.5% at the bottom. The viewability data →
- 2. Hold every unit to a 60% viewability floor. In AdSense, check each unit's Active View viewability. Anything under 60%, move it into the content or a dismissible anchor, or cut it. On our sites revenue is almost entirely click-driven, so a unit that is not seen cannot pay, it just adds page weight and clutter. Three of seven of our placements failed this floor, and the fix was never "add more." The net-positive test →
- 3. Run responsive sizes; never force an odd one. Use responsive ad units and let the auction pick the size. The highest-earning size on our network was no fixed size at all, the responsive creative at $2.54 RPM, beating every pinned size. The only real size mistake is forcing a strange one: the 250x250 square earned $0.32. The sizes study →
- 4. Treat auto ads as a scalpel, not a switch. If you run them, keep the anchor, turn vignettes off, cap the ad-load slider, and exclude your header and the area above the fold. The vignette earns about four times a normal unit ($9.06 RPM against $2.06 for in-page on our network), which is exactly why it is a trap: the per-impression money is real, and so is the cost to your rankings and your reader. Auto ads vs manual →
- 5. Stop optimizing fill rate. Ignore the fill number; watch RPM and viewability. Our site that filled 41% of its ad requests out-earned the one that filled 83%, on page RPM ($4.11 against $1.54). A higher fill rate is not more money. It often just means a price floor set too low. The fill-rate study →
- 6. Protect your Core Web Vitals. Reserve each ad slot's space so the page does not shift as ads load. This one is mechanism, not our first-party data: Core Web Vitals are a Google ranking signal, and the most common way ads harm them is layout shift from slots that were never given reserved height. The traffic your ads feed on sits downstream of your rankings, so a few percent more ad revenue is a bad trade for a ranking slip. The free scan reads your live CLS →
- 7. Spend your effort where you have leverage. Do not burn weeks on what you cannot move. Our two niches pulled nearly the same cost per click, yet one site earned far more per impression, and the gap was viewability and click-through, not the topic. "My niche does not pay" is usually a viewability problem wearing a costume. The niche study →
- 8. Know when AdSense stops being the answer. When you clear a managed network's traffic bar, move up for the demand. On our network 88% of AdSense revenue comes from Google's own advertiser pool, with only a thin tail of other buyers. A header-bidding network puts many more bidders on every impression, which is the real ceiling-raise. AdSense is the floor you upgrade from. Which network, and when →
Myths
What you will be told that our data contradicts
| The common advice | What we measured |
|---|---|
| "Add more ad units." | Placement, not count, drives viewability and clicks. see |
| "Get your fill rate up." | Our lower-fill site earned more per page. see |
| "Pick the best-converting size." | Responsive beat every fixed size we served. see |
| "Turn on everything in auto ads." | The vignette pays most and costs most. see |
| "Your niche caps your earnings." | Same cost per click; the gap was viewability. see |
Then measure
How to know it worked
Do not guess. The free scan reads your live Core Web Vitals and ad stack and returns a Calm score in seconds. The test scores each unit against the two bars. The audit runs the whole playbook on your real numbers, with your viewability reporting attached.
One honest note on the numbers above. They are measured on the small network we run, two young sites, almost entirely US, earning modest sums. Your absolute figures will differ. We publish the method, not a promise, and we link the data behind every claim so you can check it yourself. The levers are the same at every size. What changes is how much each one is worth.