More on this book
Kindle Notes & Highlights
Read between
January 9 - January 15, 2023
For the purpose of this book, programmatic advertising is defined as the automation of buying and selling of ad inventory, supported by the OpenRTB ecosystem.
In a world where each impression can be individually bought and sold, data comes to the forefront to help make good trading decisions. In fact, automation technology is designed to derive its power from data. In the programmatic revolution, marketers who have access to high-quality data, and know how to use it, win.
This chapter also introduces data management platforms,
set, cookies usually have the same domain attribute as the domain that a user is currently visiting – these are called first-party cookies. However, if a web page contains content from other domains (typically banner ads), cookies with a different domain attribute can be set as well. These are called third-party cookies, and can be read by the domain that set them. Third-party cookies can be used to easily track users across multiple domains, and are controversial with regards to data control and privacy. They can be blocked in most browsers, rendering them unsuitable for common analytics
...more
Tracking is also behind user segmentation – if, for example, the browsing history reveals that a particular user has visited car-related websites multiple times over the last week, they might fall into a “car interest” or even “car purchase intender” behavioral segments.
Cookie misuse, particularly with regards to third-party cookies, has led to increased regulation and concern.
Tracking pixels are often used by third parties in order to track user activity on a particular site.
With the advent of dynamic advertising, content of a banner ad can be generated and personalized on the fly as the webpage loads – based on data about the user who is about to see the ad.
However, possibilities are much broader – including storytelling,
DSPs are increasingly offering ad serving functionality as well.
Ad exchanges also benefit advertisers, letting them decide how much each impression is worth to them (often based on additional data) and bid accordingly.
exchanges can be accessed via DSPs, or other bidding technologies.
do so, they usually either set up their own in-house programmatic team, or employ an agency or agency trading desk to manage programmatic buying on their behalf. Programmatic
Both the sell-side as well as buy-side players can collect and take advantage of user data, which is typically facilitated through a demand-side platform (DSP).
Data levels the playing field, giving smart advertisers – albeit with smaller media budgets – a fair chance to compete and win.
Others work with their agencies and agency trading desks, or DSP managed services teams.
demand side platform (or DSP) is a gateway to the programmatic world for agencies and advertisers.
DSPs are used to purchase advertising in an automated way on a wide range of inventory – including ad exchanges, ad networks or SSPs. Demand side platforms serve as a central point of programmatic campaigns, enabling the design, execution, optimization and subsequent analysis of a media strategy.
Parameters such as budgets, maximum bids, timing, frequency, inventory or geography can be set according to campaign objectives. Furthermore, one can target specific audiences, and adjust bids according to the audience value to the particular advertiser.
DSPs increasingly work directly with advertisers, bypassing traditional media agencies.
There is a large number of DSPs to choose from at the moment, each with specific strengths and weaknesses. Some of the better known players include DataXu, AOL, Turn, The Trade Desk, MediaMath, Google’s Doubleclick Bid Manager (DBM), Rocket Fuel, AudienceScience, Appnexus or Adform.
publishers are key players in the programmatic ecosystem – because they control digital content and related ad inventory. Publishers range from the premium, well-known media brands to long-tail one-person blogs.
Programmatic advertising poses the greatest challenge to traditional direct sales models to date, as barriers to offering even the most premium inventory in an automated fashion gradually disappear. With new technologies such as header bidding, publishers can dynamically allocate inventory between direct and programmatic campaigns, increasing effective scale and attractiveness of biddable premium inventory.
publishers sit on valuable, large scale data that can be collected, utilized and monetized via a data management platform.
of all, publishers can create behavioral and other audience segments which can be sold directly for ad targeting as third-party data.
Last but not least, publishers can use their data collected via a DMP to customize content for individual visitors.
example, as users give more of their attention to mobile devices, publishers need to solve mobile content monetization and cross-device tracking challenges. Adblocking is a serious concern for many publishers, who could potentially lose a significant portion of their revenues. Also, industry concerns about issues such as ad fraud, viewability, or privacy are likely to significantly alter status quo in the marketplace.
key factor is scale, and consequently market power. Premium publisher co-ops in smaller markets can achieve the reach and the inventory size to effectively challenge global advertising giants like Google and Facebook.
With a shared data management platform (DMP), publishers can offer advertisers high-quality third-party behavioral data with a significant reach. This is critical, as targeting data availability coupled with premium inventory enables local publishers to compete on par with data-rich global players.
demand sources – such as ad exchanges, networks and DSPs.
Publishers can decide not only who can buy what inventory through which channels, but also what is the minimum price the inventory can be purchased for. This
deal has its own “Deal ID”, a short unique alphanumeric string generated by SSP. The Deal ID is given to the advertiser/agency, and allows them to target a particular deal in their DSP.
Some of the well known supply side platforms are Rubicon Project, AppNexus, PubMatic, OpenX, AOL or Google’s AdX.
Data management platform (DMP) is a fairly complex piece of software used to collect, store, classify, analyze and manage large quantities of data from various sources, making it useful and actionable particularly in the marketing process. A DMP can typically take in data from a large number of disparate sources, including CRM systems, websites, apps, or external partners.
DMPs can be standalone (such as Adobe Audience Manager) or integrated (Google Doubleclick Audience Center). Standalone data management platforms can be plugged into any DSP, but data loss in synchronization poses a common problem. Integrated DMP can be easier to manage, but are typically tied to a particular advertising platform.
For publishers, a DMP serves to augment inventory with data to make it more valuable. A common use case is creating audience segments based on data collected on their visitors (such as “Travel intenders” or “Avid photographers”). Audience segments can then be used to target individual users with relevant online advertising in prospecting campaigns. This is achieved by plugging data from a DMP into a DSP (as a part of a private deal, or through direct integration) – the DSP will then bid only on the impressions aimed at the correct audience. Augmenting impressions with data makes them much more
...more
Some of the big DMP vendors include Adobe (Audience Manager), Oracle (BlueKai), Nielsen (eXelate), Krux, or Lotame.
It seems.like.it wpuld make sense.for.us to psrtndr with a dmp for easy segmentstion. Or to habe our audiemces easily sent to a dsp. In this case hoe eould they bediffsrer t than an onboareer?
Programmatic advertising currently encompasses four basic transaction types – open auction, invitation-only auction, unreserved fixed rate and automated guaranteed
case the inventory price is fixed and no auction takes place, we can distinguish either unreserved fixed rate, or automated guaranteed.
Is adsquare an auction?
OpenRTB API specification is a very detailed guide to the RTB interface. It describes basic interactions taking place between an exchange and bidders, including bid requests, bid responses, win notices and ad markups.
auction enables buying and selling of individual ad impressions in a fraction of a second, letting all participants trade each impression at its current market price. Every real-time bidding auction follows a process established by the OpenRTB protocol. An auction happens in real time (while the user is waiting for a page to load), and typically takes less than 100 milliseconds.
How does our pricing im adquare effectthe auction?
User visits a website
In a real-time bidding auction, the request to fill an available ad slot typically goes from a publisher page to a publisher ad server, which (under pre-defined conditions) redirects the request to a Supply-side platform.
2) Supply-side platform issues a bid request
Deal ID[xii] is a unique identifier designed to help buyers and sellers set up custom programmatic deals. When advertisers and publishers negotiate a deal (regardless of the actual transaction type), they can for instance agree a certain price floor (or fixed price), inventory, targeting, priority, or url transparency.
The Deal ID is then sent over (usually via e-mail) to an advertiser or their agency, and assigned to a line item in the DSP.
To kick off this chapter, we will look at basic data categorization based on source (first-, second- or third-party) and collection method (declared, inferred or modelled). These classifications are commonly used across the industry, and it’s very useful to have a firm grasp of their meaning.
These include contextual targeting, behavioral targeting and look-a-like modelling.
One common data categorization based on source is into the first, second, and third-party. First-party data is collected directly from your own users. This is the most valuable, exclusive and accurate data marketers have (and it’s free!).
