Metadata
- Author:
- Full Title::
- Category:: 📚Books
- Finished date::
New highlights added 2023-04-09
Traditionally, big publishers sold their digital inventory through an in-house direct sales team, which negotiated inventory deals with advertisers or their agencies. With the advent of ad networks and ad exchanges, the non-premium and remnant inventory was gradually offered via non-direct channels. (Location 468)
even with a greater shift to programmatic, the in-house direct sales are not likely to disappear completely. Their focus will probably move to the most premium inventory, custom campaigns, and new trading models – such as programmatic guaranteed and programmatic direct in general. (Location 475)
Current growth of programmatic digital video along with programmatic audio is a good example, and publishers who embrace these trends will reap the rewards. (Location 481)
SSPs have the benefit of connecting publisher inventory to several demand sources – such as ad exchanges, networks and DSPs. (Location 496)
Just like many DSPs, some SSPs are gradually expanding their offering to include bidding functionality and other technologies. The line between SSPs, DSPs, DMPs and other ad tech vendor types is becoming blurrier. (Location 508)
Magnite (formerly Rubicon Project), Xandr Monetize (formerly AppNexus), PubMatic, Index Exchange or Google’s Ad Manager. (Location 510)
Specialized brand safety, fraud prevention and viewability measurement (jointly referred to as ad verification) vendors are now an integral part of the programmatic ecosystem. (Location 521)
Advertisers use ad servers (typically integrated with their DSPs of choice) (Location 538)
publishers integrate their ad servers with selected SSPs (Location 539)
the well-known ad servers include Google Ad Manager, Sizmek, Equativ (formerly Smart) or Adform. (Location 540)
A common misconception equates auction-based transaction types (commonly referred to as RTB, or real-time bidding) with programmatic advertising, ignoring fixed-price transaction types. (Location 588)
Open auction (Location 596)
tends to have low priority in publisher monetization setups, behind PMP auction, (Location 597)
To further protect direct sales and other more premium channels, publishers sometimes mask URLs in an open auction (multiple domains are grouped under the same URL), or the auction is completely blind (URL is not disclosed at all). (Location 600)
A bid response contains one or several bids from various seats (i.e. buyers, usually different advertisers or agencies) under the same DSP, or even from the same seat (one advertiser/agency with several campaigns). A bid response usually includes Ad Markup (Ad Tag or Ad serving code, so the ad can be rendered). (Location 684)
In case the DSP didn’t already send Ad Markup in a bid response, it must be sent at this point. Ad Markup is normally sent in a bid response though. (Location 695)
New highlights added 2023-04-10
Both ads.txt (by the publisher) and sellers.josn (by SSPs) are a way for buyers in exchanges to verify whether they are really buying an adunit at a publisher. See https://www.inmobi.com/blog/understanding-the-differences-between-ads.txt-and-sellers.json
Sellers.json (Location 712)
New highlights added 2023-04-10
Data in digital advertising is an incredibly complex topic, (Location 954)
Readers who would like to know a bit more will find an in-depth exploration in a complementary book, “Data in Digital Advertising.” (Location 955)
Access to mobile app data is becoming more restricted recently due to privacy measures, both on iOS (AppTracking Transparency Framework) and Android (Privacy Sandbox). Mobile websites offer similar data to desktop websites but are often more difficult to track due to high market share of iOS devices in many countries. (Location 991)
First-party data is collected directly from owned or controlled sources, typically by advertisers or publishers[xxxiv]. This is the most valuable, exclusive and accurate data marketers have (and it’s free, apart from the technology costs). (Location 1070)
Zero-party data, a recently prominent category of data based on ownership, is data that is intentionally shared by the users themselves. (Location 1100)
new business models could become viable, based on total user control and monetization of their own data. (Location 1104)
Advertisers and agencies create most of the data demand, (Location 1109)
while publishers, closed advertising platforms and third-party vendors/data exchanges provide data supply. (Location 1109)
With the demise of third-party cookies, publishers (with their own first-party data) are poised to replace third-party vendors as primary data providers within the open programmatic ecosystem. (Location 1110)
Contextual targeting (Location 1165)
This method of targeting has been gaining renewed popularity in the recent years, as it doesn’t require either cookies or user consent. There’s also been a significant progress in terms of targeting accuracy, particularly due to the rise of semantic targeting. (Location 1169)
Semantic targeting is a more sophisticated form of contextual targeting, utilizing semantic techniques and natural language processing to determine the overall context (Location 1179)
New highlights added 2023-04-14
It is important to note that not every web page is suitable for contextual targeting. Many web pages don’t offer a single and clearly defined context (e.g. the main page of a news site). Here, behavioral targeting could be more appropriate. (Location 1183)
Behavioral targeting is ideal for prospecting campaigns. Advertisers often purchase third-party audience data for this purpose, either from specialized data vendors, or directly from publishers (here, advertisers are essentially purchasing publishers’ first-party data, immune to third-party cookie deprecation). (Location 1205)