1:03Ep 9: Read your Prebid analytics
Ranking bidders by revenue alone hides why they underperform.
Lesson notes
Ranking bidders by revenue alone hides why they underperform.
A Prebid analytics adapter records each bidder’s requests, bids, timeouts and wins.
Prebid.js fires an event at each step of the auction. An analytics adapter, from a vendor or in-house, sends those events to a dashboard. Say Bidder C got ten thousand requests on recipes.example. It bid four thousand times, a forty percent bid rate. It won four hundred, a ten percent win rate. Fifteen hundred answers came after the timeout, a fifteen percent timeout rate. A low bid rate points to setup, consent or demand. A low win rate means it bids, but low. A high timeout rate means its bids arrived too late, so check speed before you drop the partner.
Today, confirm your build has an analytics adapter, then find your highest timeout rate.
Next, dynamic floors in Prebid.
Go deeper · 3 min readHeader Bidding Optimization – A 101 Guide for PublishersUp next in this series
1:02Ep 10: Dynamic floors in PrebidEp 10: Dynamic floors in Prebid
1:07Ep 1: How Prebid.js runs an auctionEp 1: How Prebid.js runs an auction
1:02Ep 2: Bidder adapters and your Prebid buildEp 2: Bidder adapters and your Prebid build
1:06Ep 3: Price granularity, the buckets that decide your payEp 3: Price granularity, the buckets that decide your pay
1:00Ep 4: One line item per price bucketEp 4: One line item per price bucket
1:04Ep 5: Send-all-bids vs top priceEp 5: Send-all-bids vs top price
1:01Ep 6: The Prebid Universal CreativeEp 6: The Prebid Universal Creative
1:06Ep 7: Managed wrapper vs self-hosted PrebidEp 7: Managed wrapper vs self-hosted Prebid
1:17Ep 8: Prebid, Open Bidding and Amazon in one setupEp 8: Prebid, Open Bidding and Amazon in one setup