Fill Rate

Fill rate is the ratio of filled impressions (ads actually served) to total ad requests sent by a publisher’s inventory. It answers the question: Of all the opportunities I offered to the market, how many actually got a creative served?

In Google Ad Manager, the metric is typically computed as Served impressions / Eligible requests (programmatic) or Total impressions / Total ad requests (all channels). The exact denominator depends on whether you exclude non-eligible requests (e.g., due to frequency capping or roadblocks).

What fill rate does NOT tell you

• It does not measure revenue per request — that’s RPM.
• It does not measure bid density — that’s match rate.
• It does not measure viewability or quality of the fill.

So what: Fill rate is a necessary but insufficient metric. Use it as a signal of technical delivery health, not as a revenue optimization target.

Fill Rate = Served impressions / Total ad requests

Example: If a publisher sends 1,000,000 ad requests and 920,000 ads are served, the fill rate is 920,000 / 1,000,000 = 92%.

Important caveats

1. Denominator definition varies by platform. Google Ad Manager’s Programmatic Fill Rate uses Programmatic eligible requests as the denominator — requests that passed targeting, frequency capping, and roadblock checks. Other platforms may use all requests including those filtered out.
2. Passback loops can inflate fill rate. If a waterfall chain eventually serves a house ad, that counts as fill even if no paying advertiser bought it.
3. Header bidding wrappers may report fill differently. Some SSPs count a bid response as fill only if the creative renders; others count any response with a price.
4. Timeouts matter. A request that times out before any response is counted as unfilled, which can depress fill rate on slow connections or complex auctions.

A fill rate of 95%+ generally indicates healthy delivery — nearly every request is monetized. Below 80% you likely have a systemic issue: low demand, misconfigured price floors, or technical failures.

But never read fill rate in isolation. Always pair it with:

• eCPM — High fill + low eCPM often means you are filling with low-paying remnant or house ads.
• Match rate — If match rate is high but fill rate is low, the issue is likely after the bid (e.g., creative rendering failure).
• Viewability — Fill rate can be high while viewability is low if you are serving ads below the fold.

Common dashboard misread

A publisher sees fill rate jump from 88% to 96% after adding a new mediation partner and celebrates. But eCPM dropped 30%. The new partner filled cheap inventory that previously went unfilled — net revenue may be flat or down. Fill rate alone is a vanity metric without revenue context.

Levers that increase fill rate

• Lower price floors — More demand passes, but at lower CPMs. Use floor optimization tools to find the sweet spot.
• Add demand sources — More SSPs, exchanges, or mediation partners increase competition and fill.
• Enable passback / house ads — Ensures every slot gets something, even if unpaid.
• Increase timeout windows — Header bidding timeouts that are too short cause unfilled requests. Balance against page load latency.
• Relax targeting constraints — Narrow geo or device targeting reduces eligible demand.

Levers that decrease fill rate

• Raise price floors — Filters out low bids, improving eCPM but potentially reducing fill.
• Add frequency caps — Reduces eligible impressions for repeat users.
• Stricter creative requirements — E.g., requiring MRAID or specific sizes may cause mismatches.

Tradeoffs

Fill rate vs. eCPM is the classic publisher tension. Optimizing fill alone can flood your inventory with $0.10 CPM ads. Optimizing eCPM alone can leave 30%+ of requests unfilled. The correct north star is RPM (Revenue Per Mille requests), which combines both: RPM = Fill Rate × eCPM.

Fill rate vs. latency — Longer timeouts increase fill but degrade user experience. A common mistake is setting a 2-second header bidding timeout to chase 99% fill, then watching bounce rates climb.

Fill rate is a health check, not a revenue target. It tells you if your inventory is being monetized, but not how well.

• Always pair fill rate with eCPM and RPM to understand revenue impact.
• Filter out house ads to get a true paid fill rate.
• Optimize for RPM, not fill rate alone.