Prebid auctions operate on a strict time limit. Bidders are typically given 1-2.5 seconds to submit their offers. Exceeding this brief window results in a timeout, effectively disqualifying the bidder from the auction.
While the default timeout is set at 1 second, industry best practices suggest extending it to 1-2.5 seconds for optimal performance.
Despite the technical nuances and different timeout types mentioned in the official documentation, the core concept remains simple for publishers: it's the crucial timeframe within which ad bidders must respond to ad requests.
Think of it like an auction house setting a deadline for bids. Too short, and you risk missing out on high-rollers. Too long, and everyone's stuck waiting, which frustrates both bidders (who lose out on potential wins) and your visitors (who just want to see your awesome content, not a loading screen). So, how do you find the sweet spot for maximum ad revenue and a happy audience? Let's dig in!
So, what is a Prebid timeout?
When a user loads a page, Prebid sends bid requests to connected bidders. Bidders have a limited time to respond; those that don't meet the deadline time out. The timeout value is set in the pbjs.requestBids() call:
pbjs.requestBids({
bidsBackHandler: sendAdserverRequest,
timeout: 1000,
labels: ["custom1"]
});
A timeout situation doesn’t include the case where a bidder decides not to bid. Even if a bidder isn’t interested in the auction, it’ll send a ‘no bid’ response. So, a bidder is considered as ‘timed out’ only when it cannot send a bid or ‘no bid’ within the predefined time frame.
The Prebid timeout sequence:
Bid Request: A user loads a webpage, triggering a bid request to multiple demand partners (SSPs, Ad Exchanges, etc.) via Prebid.
Timeout Period: Demand partners have a set time (the Prebid timeout) to respond with their bids.
Timeout Expiration: If a demand partner doesn't respond within the set time, their bid is excluded from the auction.
Auction Completion: The highest bid from the responding demand partners wins the ad impression, and the page load is finalized.
What is the timeout rate in Prebid?
Publishers are interested in the timeout rate of their bidders because it shows how frequently their bidders fail to respond. For example, if a bidder received 100 requests and couldn’t respond to 20 of them because it timed out, it has a 20% timeout rate. If the timeout rate is unsatisfactory, you can allot more time to your bidders to improve it.
Two methods of bid timeout: Auction timeout and Failsafe timeout
Prebid Timeouts are crucial for balancing revenue and user experience. There are two primary types:
Auction Timeout: This sets the maximum time for Prebid to coordinate with demand partners. A short timeout might limit revenue by cutting off potential bids, while a long timeout could negatively impact user experience due to delayed page load times.
Failsafe Timeout: This is a backup timer set by the publisher. If the auction timeout expires before the failsafe timeout, the auction is canceled, and the ad request is sent directly to the ad server. A failsafe timeout that's too short can result in lost revenue, as the auction might end prematurely.
For publishers using Prebid Server, the s2sConfig timeout should be set lower than the auction timeout. This ensures that bids from Prebid Server are received in time for the ad server call.
By carefully configuring these timeouts, publishers can optimize their header bidding setups to maximize revenue while maintaining a seamless user experience.
Why Shouldn’t Your Bidders Time Out Frequently?
Although it’s impossible to achieve, your bidders should ideally have a 0% time-out rate. It means they should never fail to respond to a bid request. When your bidder fails, it reduces the competition for your impressions and brings your revenue down. For example, bidder A could’ve bid $2 for an ad, but it timed out, so bidder B won the auction with just $1.5 because it was the next best bid. Here, you lost $0.5 because bidder A couldn’t respond in time.
While the problem looks minuscule when we look at a single auction, the losses add up when too many bidders are timing out frequently.
Why Do Bidders Time Out?
Bidders can time out due to hundreds of reasons. For example:
An insufficient time limit for the bidder to respond.
Faulty integration with the wrapper.
Slow internet connection of the user.
Browser’s lack of power to parse your site quickly.
Improper CMP configuration.
Technical glitches on the bidder’s side.
Heavy load on the site.
Too many bidders in the auction, and many more.
Why Prebid Timeouts Matter?
Revenue Implications
Missed Bids = Lost Revenue: Short timeouts might cut off slow-responding but potentially high-paying demand partners, reducing overall revenue.
Optimal Timeout = Maximized Revenue: Finding the sweet spot can increase competition, driving up bids and revenue.
User Experience (UX) Considerations
Latency vs. Load Time: Longer timeouts can delay page loading, negatively impacting UX and potentially leading to:some text
Increased bounce rates
Decreased engagement
Lower SEO rankings (due to page speed being a ranking factor)
Balancing Act: Premium publishers must weigh the need for speedy page loads against the potential revenue benefits of slightly longer timeouts.
How Do You Optimize Your Timeout?
Analytics are your weapon.
Prebid's Analytics Module: Gain insights into bidder performance and identify slow-performing bidders.
Third-Party Tools: Get a granular view of bidder behavior across different devices and regions.
Dynamic Timeouts: Dynamic timeouts adjust based on real-time factors like device type and network speed. Think of it as personalizing the response window for each bidder, ensuring they have the time they need to compete effectively.
A/B Testing: Don't be afraid to experiment! A/B test different timeout settings to see what drives the most revenue without compromising user experience. Remember, a happy user is a loyal user, and loyal users translate to long-term revenue growth.
Collaborate with Your Wrapper Provider: These individuals have a deep understanding of Prebid technology. They also have insights into bidder performance across their entire network. Leverage their expertise to optimize your timeouts for maximum impact.
Vet your bidders
Regularly review bidder performance: Identify inactive or low-performing bidders.
Prioritize high-quality bidders: Focus on bidders with strong performance and fast response times.
Configure failsafe timeouts carefully
Set a higher failsafe timeout: Prevent premature auction termination.
Avoid overlapping timeouts: Ensure the failsafe timeout is significantly higher than the Prebid timeout.
Leverage automation tools
Dynamic timeout adjustments: Use AI-powered tools to adjust timeouts based on real-time factors.
Automated bidder selection: Optimize bidder selection to improve performance.
Accelerated dynamic data fetching: Reduce latency and improve auction efficiency.
Remember, Publishers:
Device & Geography Matter: Account for variations in user devices and network speeds across different regions.A bidder's performance can vary significantly across different regions.
User Experience is King: Don't sacrifice user experience for a few extra milliseconds. A slow-loading site is the enemy of engagement and revenue.
Balance is Key: The goal is to maximize your ad revenue without driving users away. It's a delicate dance, but one that can supercharge your ad strategy.
By optimizing Prebid timeouts, publishers can significantly enhance their ad revenue.
What’s Next?
Many publishers don’t bother about their time-out because it takes meticulous observation and testing to perfect it.
Many publishers set their time out between 1000ms to 2000ms, and it still works out for them. But, if you spend some time figuring out the best duration, it can noticeably improve your revenue and user experience.
Even if you’re using a third-party header bidding wrapper, make sure it gives equal importance to Prebid time out.