The traffic graph is cheering. The ad dashboard isn't.
Your old blog posts are finally attracting readers. Traffic climbs, and you open the advertising dashboard expecting a little celebration. Instead: 85 ad impressions, $0.15 earned, and an effective CPM of $1.765.
The content is apparently at work. The ads seem to have forgotten their shift.
Pageviews are not ad impressions
A pageview measures a viewed page. An ad impression is counted under the ad network's own delivery and measurement rules. Adsterra says an ad must successfully load and meet its counting conditions. One page can generate multiple ad impressions—or none.
Suppose a properly matched period shows 890 pageviews and 74 recorded ad impressions. That's 74 ÷ 890, or about 8.3 ad impressions per 100 pageviews. It does not prove that exactly 8.3% of pages displayed an ad: multiple ads can appear on a single page.
First match the date range, time zone, domain, placement and included page types. Comparing unrelated date windows turns a measurement problem into a guessing game.
How many views would generate a meaningful monthly income?
At an illustrative eCPM of $1.765, 85 impressions earn about $0.15. For a monthly target of ¥200,000—roughly US$1,266 if we assume ¥158 per dollar—the site would need about 717,000 ad impressions per month, with that eCPM held constant.
But pageviews depend on how often ads are actually counted:
| Ad impressions per 100 pageviews | Monthly pageviews required |
|---|---|
| 100 | About 720,000 |
| 50 | About 1.43 million |
| 20 | About 3.59 million |
| 8.3 | About 8.6 million |
These are illustrative, pre-tax figures. Exchange rates, ad rates and traffic quality change. Eighty ad impressions from a thousand pageviews can be a far bigger revenue problem than a modest eCPM.
Find the missing impressions before chasing more visitors
Check which page templates actually contain approved ad placements. Compare article pages with the homepage or app views that may intentionally carry no ads. Then test an eligible page in a real browser: is the ad script loading, or is it blocked, delayed, hidden by consent rules, or failing before the ad appears? Geography and ad format can matter too.
Don't assume every pageview was eligible to show an ad. And don't respond by plastering every screen with intrusive units. Driving away readers is a rather expensive way to improve a spreadsheet.
The practical goal is simple: compare pageviews and counted ads for the same pages and dates, then track impression yield and eCPM separately. The visitors found the front door. Now figure out why so few ads are making it onto the floor.
