What is Header Bidding?

Also known as: Prebid, Server-side header bidding

Quick definition

Header bidding is a technique where a publisher offers each impression to multiple SSPs and exchanges simultaneously before calling its ad server, so demand competes on price at once instead of in a sequential waterfall.

How it works

Page or app loads a wrapper (e.g. Prebid.js or an SDK)
Wrapper requests bids from several demand partners in parallel
Bids are passed as key-values to the publisher ad server
Ad server compares header bids with direct deals and picks the winner

Key points

Client-side: Auction runs in the browser; transparent but adds page latency and limits partner count.
Server-side: Auction runs on a server (e.g. Prebid Server); scales to more partners but reduces cookie match rates.

Benefits

Higher yield through simultaneous competition
Less dependence on a single ad server's exchange

Trade-offs

Latency and complexity
Duplicate bid requests inflate infrastructure cost

FAQ

How is header bidding different from a waterfall?

A waterfall calls partners one after another by historical priority; header bidding asks them at the same time and lets the best real-time price win.

Explore on AdTech Radar

Official sources

Prebid.org

Last reviewed 2026-10-27. Explanations are written by AdTech Radar; consult the official specification for implementation.

Related terms

OpenRTB

OpenRTB is the IAB Tech Lab specification that standardizes how an exchange or SSP sends a bid request describing an impression to DSPs and how DSPs answer with a bid response, so thousands of buyers and sellers can trade in real time through one common JSON message format.

Supply Path Optimization (SPO)

Supply path optimization is the practice of buyers choosing the fewest, cheapest and most transparent routes to the same publisher inventory, cutting redundant resellers so more of the media budget reaches the publisher.

Bid Shading

Bid shading is a DSP algorithm that lowers a bid below the buyer's maximum value in first-price auctions, estimating the lowest price likely to still win so advertisers do not overpay.