The most basic and common terminology that every Adtech guy should know about is an "Auction". What is it in terms of our business, how it works and manipulates the things in real time.
In simple words, it's an algorithm which considers all of the values from the partners (demand) who are willing to serve their ads on specific set of inventory of publisher.
For example: ebay.com/furniture has 2 ad placement of ad size 728X90 and 300X250 respectively. Now in order to fill this space in real time there will be number of bidders who would compete and best fit will win over the rest.
How will it win? What's the best fit? Why only 1?
Well, that's the logic when we say an auction. Everyone will be having specific set of instructions with them which have encoded values like, identity, category, price, serving technology etc. Auction engine will receive the all inputs, process them and find the best fit out of all depending upon the inventory type.
let's consider (assume) an expmple what an auction engine can do.
DBM (DSP), Nestle(Advertiser), Brazil(Targeting region), 27-40 (Age), Male(Demographic), $5.00 (Cost)
MM (DSP), BMW(Advertiser), New York(Targeting region), 27-40 (Age), Male(Demographic), $5.00 (Cost)
Ad.com (DSP), Ford(Advertiser), US(Targeting region), All (Age), Male & Female(Demographic), $2.50 (Cost)
Turn (DSP), Budwiser(Advertiser), US&CA(Targeting region), 21+ (Age), Male(Demographic), $3.00 (Cost)
.
.
.
.
AOL (DSP), Loreal(Advertiser), CA(Targeting region), 27-40 (Age), Female(Demographic), $7.00 (Cost)
Among all the bidders who've entered the auction, engine (program) will consider each and every parameter and compare with the available inventory. This comparison will start with the price and can scale upto technology, ad type, browsers, and many more granular data etc.
Once, everything has been matched the one with most relevance and highest price will get the advantage and win the auction to serve the ad (creative).
This overall phenomenon looks pretty simple and straightforwrad at glance, but if you go to understand the exact expectation and layers of auctioning you'll gauge the complexity of the concept and find it more interesting.
Feel free to get in touch for further discussion :)
In simple words, it's an algorithm which considers all of the values from the partners (demand) who are willing to serve their ads on specific set of inventory of publisher.
For example: ebay.com/furniture has 2 ad placement of ad size 728X90 and 300X250 respectively. Now in order to fill this space in real time there will be number of bidders who would compete and best fit will win over the rest.
How will it win? What's the best fit? Why only 1?
Well, that's the logic when we say an auction. Everyone will be having specific set of instructions with them which have encoded values like, identity, category, price, serving technology etc. Auction engine will receive the all inputs, process them and find the best fit out of all depending upon the inventory type.
let's consider (assume) an expmple what an auction engine can do.
DBM (DSP), Nestle(Advertiser), Brazil(Targeting region), 27-40 (Age), Male(Demographic), $5.00 (Cost)
MM (DSP), BMW(Advertiser), New York(Targeting region), 27-40 (Age), Male(Demographic), $5.00 (Cost)
Ad.com (DSP), Ford(Advertiser), US(Targeting region), All (Age), Male & Female(Demographic), $2.50 (Cost)
Turn (DSP), Budwiser(Advertiser), US&CA(Targeting region), 21+ (Age), Male(Demographic), $3.00 (Cost)
.
.
.
.
AOL (DSP), Loreal(Advertiser), CA(Targeting region), 27-40 (Age), Female(Demographic), $7.00 (Cost)
Among all the bidders who've entered the auction, engine (program) will consider each and every parameter and compare with the available inventory. This comparison will start with the price and can scale upto technology, ad type, browsers, and many more granular data etc.
Once, everything has been matched the one with most relevance and highest price will get the advantage and win the auction to serve the ad (creative).
This overall phenomenon looks pretty simple and straightforwrad at glance, but if you go to understand the exact expectation and layers of auctioning you'll gauge the complexity of the concept and find it more interesting.
Feel free to get in touch for further discussion :)



