What Is Header Bidding?
Header bidding is a programmatic advertising technique that allows publishers to offer their ad inventory to multiple demand-side platforms and ad exchanges simultaneously, before the ad server makes its final decision. The highest bid wins the impression. This approach replaced the older “waterfall” method, where exchanges were called one at a time in priority order, often leaving revenue on the table.
The name comes from the JavaScript wrapper code placed in the <head> section of a webpage, which runs the auction in the browser before the page fully loads.
How Header Bidding Works
When a user loads a page, the header bidding wrapper fires simultaneously to all connected demand partners. Each partner submits a bid in real time. Once the timeout window closes (typically 300ms to 1,000ms), the highest bid is passed to the publisher’s ad server, such as Google Ad Manager, as a key-value pair. The ad server then compares that bid against its own direct-sold inventory and selects the winning line item.
- User visits page; header tag fires
- Wrapper sends bid requests to all configured demand partners simultaneously
- Partners return bids within the timeout window
- Wrapper selects the highest bid and passes it to the ad server
- Ad server compares the external bid against internal campaigns
- Winning creative renders on the page
Header Bidding vs. the Waterfall Method
Before header bidding, publishers used a daisy-chain approach called the waterfall. In a waterfall setup, the ad server called demand sources in a fixed priority order based on historical averages. If the first network passed, the call went to the second, then the third, and so on. This introduced latency and systematically undervalued inventory because a lower-priority network willing to pay more never got a chance to bid.
| Factor | Waterfall | Header Bidding |
|---|---|---|
| Auction structure | Sequential | Simultaneous |
| Revenue potential | Lower (missed bids) | Higher (full competition) |
| Latency | Accumulates per network call | One parallel timeout window |
| Transparency | Low | Higher |
Revenue Impact
Publishers that adopted header bidding typically report revenue lifts between 20% and 50%, depending on their traffic profile, niche, and number of demand partners integrated. The Washington Post and The Guardian were among the early large-scale adopters, and both publicly credited header bidding with significantly improving their programmatic revenue.
A simple way to estimate incremental revenue from header bidding:
Incremental Revenue = (New Average CPM – Previous Average CPM) x Total Impressions / 1,000
For example: if a publisher previously averaged a $3.00 CPM across 100 million monthly impressions and header bidding raises that to $4.20, the incremental monthly revenue is:
($4.20 – $3.00) x 100,000,000 / 1,000 = $120,000/month
Client-Side vs. Server-Side Header Bidding
The original implementation runs entirely in the user’s browser (client-side), which introduces latency as the browser waits for bid responses. Server-side header bidding moves the auction to an external server, reducing the load on the browser and cutting timeout risk, though it can reduce bid accuracy because server-side calls may lose user-level cookie data.
- Client-side: Higher accuracy, more latency, better cookie matching
- Server-side: Faster page load, reduced user data accuracy, scalable for video and mobile
Many publishers run a hybrid approach, routing display inventory through client-side and video or mobile app inventory through server-side.
Prebid.js: The Open-Source Standard
Prebid.js, maintained by the open-source consortium Prebid.org, became the dominant wrapper technology for web-based header bidding. As of 2024, Prebid.js is integrated across thousands of publishers globally and supports more than 300 bidder adapters, meaning publishers can plug in nearly any demand-side platform without writing custom code. Publishers largely chose it for its vendor-neutral governance model, which contrasts with proprietary wrappers offered by individual ad tech vendors.
For mobile apps, Prebid Mobile SDK fills a parallel role, enabling header bidding on iOS and Android where browser-based JavaScript does not apply.
Demand Partners and Bid Density
The value of header bidding scales with the number of active demand partners, but only up to a point. Adding more bidders increases competition and raises CPM, but it also increases timeout risk and latency. Publishers typically run between 5 and 15 demand partners, tuning their timeout window to balance revenue against page performance.
Platforms such as The Trade Desk, Index Exchange, Magnite, and OpenX are common demand partners in publisher header bidding stacks. Google participates through its Open Bidding product, a server-side alternative that competes with traditional client-side header bidding setups.
Header Bidding and Google’s Role
Google’s ad server, Google Ad Manager (formerly DoubleClick for Publishers), holds a dominant position in the publisher market, which created tension as header bidding emerged. Google initially competed through its own exchange, AdX, which had first-look advantages in the waterfall. Header bidding eroded that structural advantage by letting other exchanges compete on equal footing.
The U.S. Department of Justice filed an antitrust case against Google in 2023. It specifically cited Google’s practices around header bidding, arguing that Google manipulated auction dynamics to protect AdX. This context matters for publishers evaluating the long-term trajectory of header bidding regulation and platform policy.
Key Metrics to Monitor
- Bid rate: Percentage of auctions where a demand partner submits a bid
- Win rate: Percentage of submitted bids that win the impression
- Timeout rate: Percentage of bid requests that exceed the timeout window
- Average CPM per partner: Used to prioritize and prune low-performing demand sources
Monitoring these metrics within a supply-side platform or analytics dashboard helps publishers continuously optimize their wrapper configuration without guessing.
Relationship to Programmatic Advertising
Header bidding is one mechanism within the broader programmatic advertising ecosystem. It sits at the supply side of the transaction, giving publishers more control over how their inventory is auctioned. On the demand side, advertisers interact with header bidding indirectly through their demand-side platforms, which submit bids through the same unified auction. Understanding header bidding is essential for anyone managing real-time bidding strategy on either side of the marketplace.
Frequently Asked Questions
What is header bidding in simple terms?
Header bidding is a technology that lets website publishers hold an open auction for their ad space before the page loads, so every ad network competes at the same time instead of one at a time. The highest bidder wins and their ad appears on the page.
Does header bidding increase publisher revenue?
Yes. Publishers that implement header bidding typically see revenue increases of 20% to 50% compared to the older waterfall method. The gain comes from full market competition: every demand partner sees the same impression at the same time and bids accordingly.
What is the difference between header bidding and Google’s Open Bidding?
Header bidding (specifically client-side via Prebid.js) runs the auction in the user’s browser and gives publishers direct access to demand partners. Google’s Open Bidding runs the auction on Google’s servers, which is faster but routes bid requests through Google’s infrastructure, giving Google visibility into the competing bids.
What is Prebid.js?
Prebid.js is the leading open-source header bidding wrapper, maintained by the nonprofit consortium Prebid.org. It supports more than 300 demand partner integrations and is free to use. Publishers implement it to run header bidding without relying on any single vendor’s proprietary technology.
Is header bidding still relevant with server-side options available?
Yes. Client-side header bidding via Prebid.js remains the standard for display inventory on the web because it preserves cookie-based user data that improves bid accuracy. Server-side solutions are more common for video and mobile app inventory, where browser-based JavaScript cannot run. Many large publishers use both.