Growth Marketing Glossary

Fill Rate

fill ratenoun

How much of the shelf actually sold — the publisher metric where 100% can mean priced too cheap.

filled 72%unfilled 28%ad requests ÷ impressions servedhow much of the inventory actually sold
Schematic — requests filled versus unfilled
Term
Fill Rate
Formula
Impressions served ÷ ad requests
Low fill
Demand, floors, or plumbing problems
100% fill
Often means underpriced inventory

Forms & parts of speech

fill rate · noun
The sold share of requests.
"Fill rate hit 99% the quarter the floors dropped - revenue per session fell with it. Full isn't the goal."

Definition in plain terms

Fill rate is the share of a publisher's ad requests that came back with a paying ad: requests filled divided by requests made. It is sell-side yield's basic gauge — the counterpart of the buy side's metrics this glossary documents — and its reading requires the trader's instinct: low fill leaves shelf space empty, but 100% fill usually means the inventory was priced too cheap, because someone would always have paid the floor you didn't set.

The mechanics

The diagnostic tree for low fill: demand-side gaps (too few bidders connected — the SSP and header-bidding configuration), FLOOR-PRICE settings above what the auction will clear (the deliberate trade: emptier but dearer), targeting-unfriendly inventory (geos, formats, or contexts thin on demand), and plumbing failures (timeouts, malformed requests, the ads.txt authorization gaps that silently drop bids — the DOMAIN-SPOOFING entry's machinery seen from the sell side). The optimization is a price-times-volume problem, not a fill-maximization one: revenue per thousand REQUESTS (not per filled impression — the eCPM entry's denominator discipline) is the honest objective, and it often peaks well below 100% fill, because floors that hold price discard the bids that would have dragged the average down. The craft variables: floors tuned per segment with the data (the dynamic-floor systems running exactly this trade), passback and mediation chains ordered by realized eCPM, latency budgets (every extra auction hop trades fill for user experience), and unfilled-inventory strategy — house ads, content recirculation, or deliberate blanks, each better than backfilling with junk demand that poisons the user experience the inventory's value rests on. App-side mediation runs the same logic with the waterfall ordering as the lever.

When it matters

Fill rate matters to anyone selling inventory — publishers, apps, retail-media networks — as the yield dashboard's volume gauge, read always beside price. It matters diagnostically at integration changes (new SSPs, header-bidding rewires, ads.txt edits) where plumbing failures masquerade as demand softness. The discipline is the paired reading: revenue per request as the objective, floors as the price-volume dial, plumbing eliminated before demand gets blamed, and 100% fill treated as the warning it usually is.

Worked example. A recipe publisher celebrates 98% fill after an ad-ops 'optimization' - floors zeroed across the board - until the quarter's revenue per session reads 14% DOWN: the inventory filled completely with bids that cleared at whatever they pleased. The yield rebuild reframes the objective as revenue per thousand requests: floors re-tune per segment from auction data (recipe-card placements hold a $2.80 floor and drop to 81% fill - at 22% more revenue; the long-tail article slots run near-zero floors where demand is thin), a timeout audit recovers 4 points of fill that was plumbing rather than pricing (one SSP's adapter timing out before bidding), and unfilled impressions route to house promos for the publisher's own cookbook - worth more than the junk backfill they replace. Full-shelf vanity retires; the dashboard's new pair - fill rate beside revenue per request - keeps the trade visible.
Failure modes to watch. Fill maximized while revenue per request falls; floors zeroed to chase the vanity of full; plumbing timeouts and ads.txt gaps read as soft demand; junk backfill poisoning the experience the inventory's value rests on; and fill compared across segments whose demand profiles share nothing.

Synonyms & antonyms

Synonyms

fill ratead fill rateinventory fill

Antonyms

unfilled impressionssell-through rate (the cousin)

Origin & history

Fill rate entered publisher vocabulary with ad networks' earliest yield reports — the percentage of the shelf that sold — and kept its seat through the programmatic and header-bidding eras, where the floors-versus-fill trade became the daily mechanics of sell-side price discovery.

Etymology: source.

Usage trends

Search interest for this term over the last five years:

View interest-over-time on Google Trends →

Common questions

What is fill rate?
The percentage of ad requests that returned a paid impression — the sell side's volume gauge, read beside price because revenue per request is the real objective.
Why is 100% fill rate a warning?
It usually means floors were too low — the inventory filled with bids that would have paid more; revenue per request often peaks well below full fill.
What causes low fill rate?
Thin demand connections, floors above clearing prices (sometimes deliberately), unfriendly inventory segments, and plumbing — timeouts and authorization gaps that drop bids silently.

Related tools & calculators

Resources & people to follow

Curated, non-competitor resources verified per term.

Related training

Disciplines

Areas of marketing where fill rate is a core concern:

Sources

  1. trendsGoogle Trends — "ad fill rate"