3 ms·
I also work in ad tech (lead the algo team in job ads). Second price auctions online are great, but they're not strategyproof. You still need a bidding strateg
by VHRanger 3y ago
I also work in ad tech (lead the algo team in job ads).
Second price auctions online are great, but they're not strategyproof. You still need a bidding strategy.
For thin/small/sparse markets they simply don't work. You need a first price auction.
Google definitely game their own 2nd price auction, but this is just one reason amon many they're not strategyproof.
- whimsicalism 3y agoDo advertisers even control their bid in the auction? In my experience, not fully.
- VHRanger 3y agoMy job is to write algos that optimize the bid on behalf of clients. Most pay per click ad platforms have an API that let you set the bid.
- whimsicalism 3y agoI also work in this field, > Most pay per click ad platforms have an API that let you set the bid. even when you are let to set the bid, usually the auction is not conducted just based on your bid but also adjustments from the probability that you will actually pay out
- soared 3y agoI’ve been a buyer for 8 years are so on ttd and another dsp but haven’t heard of the probably of paying out thing - can you share more info on that?
- whimsicalism 3y agoIf you are bidding on a pay per click basis, typically your bid will be adjusted to expected revenue based on the probability someone actually clicks on your ad. For CPM ads this probably doesn't apply. This is not for open web exchanges, referring more to stuff like adwords, facebook, etc.
- soared 3y agoAh shoot that makes sense I was so confused. I was thinking for programmatic the ssp was reducing a bid based on whether they expected the advertiser to actually pay haha
- whimsicalism 3y agoi phrased it poorly
- VHRanger 3y agoRight, this is why I said it's not strategy proof. The bid is just an input you control into an opaque function where the output is some cost and conversions