This book has the honour of being the most expensive e-book I’ve ever purchased.

US$73.85.

Ouch.

I had just heard that Programmatic Advertising was about to explode in New Zealand and I wanted a definitive introduction, and this book promised to be it.

In the free preview I could see that every chapter had 2 or 3 different authors for a total of 45 authors.

That must be where the cost of the book adds up!

But actually, on the very last page it says that the authors waived their fees, and the proceeds are going to a charity called the Children’s Shelter Foundation.

Anyway, the book delivered what it promised.

Interestingly, this book doesn’t provide a concise definition, of what Programmatic Advertising actually is, so I’ve put my own together from a variety of sources.

What is “Programmatic Advertising”?

• Programmatic Advertising is a way to deliver digital advertising to the right person, at the right time in the right place
• It is the algorithmic purchase and sale of advertising space in real time, using a vast network of hi-speed machines and software to automate the buying, placement, and optimisation of media inventory via a bidding system
• Therefore, advertisers are able to tailor a specific message to a specific person to maximise the opportunity to convert exposure to an ad into an action that the advertiser wants the person to take

In a very short amount of time I have come to believe that this method of advertising will quickly dominate all forms of advertising online, and instead of being called “programmatic advertising” it will simply be called “advertising” because it will just be the standard way it’s done.

Without further ado, here are my notes on “Programmatic Advertising: The Successful Transformation to Automated, Data-Driven Marketing in Real-Time (Management for  Professionals)” edited and collated by Oliver Busch.

Part I: Concept

1.1 The Programmatic Advertising Principle

Major multinational advertisers in particular are facing immense pressure to optimize their processes due to globalization.

If no tangible performance improvements in brand advertising are achieved, prices are simply dropped even more resolutely.

In doing so, list prices remained stable over the years, but the gap between gross and net earnings widened as time went by.

Downscaling online editorial teams due to insufficient capitalization now appears to have reached its natural boundaries, which goes to show that further price reductions do not represent a solution to the problem of brand building performance.

Programmatic advertising actually emerged from a number of different developments initiated well over 5 years ago and only became a game-changer when these developments came together:

1. Immense computing power:
   • Complex data needs to be computed in milliseconds in order to allow for real-time marketing decisions.
2. Inexpensive data storage:
   • Millions of anonymized stored data points can potentially be used to forecast campaign-specific values for each individual ad space.
3. Science in marketing:
   • New technology allows marketers to persuade mathematicians and quantum physicists to join them in their work.
4. Stock exchange structures when trading ad space:
   • Pricing based on real-time supply and demand for each individual ad impression opportunity at the moment it is created helps to make markets more dynamic.
5. Globalization of advertising markets:
   • Advances in standardization, a global infrastructure and globally operating market players help to speed up the pace of development.
6. Fast data connections
   • Correspondence on how to get the most out of an individual ad impression can be sent around the world several times over within a matter of milliseconds.
7. Personalization
   • “Real value comes with real identity” is the basic principle now banishing the pure anonymity seen on the web during its infancy.
   • Log-ins and permissions enable dialogs to be conducted across multiple devices.

Below are the main characteristics of the programmatic advertising principle:

1. Granularity,
   • i.e. full consideration of individual ad impression opportunities together with their general parameters, specific recipients and specific advertising environment.
2. Real-time trading,
   • i.e. deciding on a specific advertiser or a specific ad impression opportunity at the time of its creation and based on the latest data.
3. Real-time information,
   • i.e. assessing the available opportunity based on its highly specific characteristics and relevant empirical data collected to date.
4. Real-time creation,
   • i.e. advertisers serve a (where possible: dynamic and data-driven) ad that is best suited to the opportunity immediately after winning the bid.
5. Automation,
   • i.e. an automated booking and posting process. In a nutshell: “Programmatic advertising describes the automated serving of digital ads in real time based on individual ad impression opportunities” (Bardowicks and Busch 2013).

It connects media with creation in order to serve the right people with the right messages at the right time, and all of that on a large scale.

1.2 Programmatic Disruption for Premium Publishers

The buzzword “native advertising” is used for this, and has become a nightmare for compliance officers at publishing houses. While Facebook uses the same text and same layout, just with a small note that says “sponsored” to let users know that a post represents paid advertising, publishers must ensure the strict, clear separation of content and advertising, to prevent their independent journalism from being corrupted by the advertising industry.

1.3 Perspectives of Programmatic Advertising

Emerging Programmatic Advertising Channels, Formats and Approaches

1. Facebook Advertising
   • The entry into new formats such as video advertising and new areas like location-based advertising will provide more sustained momentum, unless the usage of Facebook diverse social media offerings will be slowed down by emerging social networks.
   • The latter is the fundamental threat to Facebook’s social media advertising ecosystem and explains the aggressive acquisition strategy of Facebook in this area.
   • Snapchat and Instagram are now strong assets in the Facebook Family.
   • We expect more significant acquisitions by Facebook in the coming years.
   • With its Atlas infrastructure, Facebook will furthermore attack in Display, Mobile and Video-Advertising beyond its own platform.
2. Mobile Advertising: Out-of-Home-Use of Mobile Devices
   • Sometimes disrespectfully referred to as “Bored-in-line” use, the local context and the interaction with the direct, local environment are most important.
3. Mobile Advertising: At-Home-Use of Mobile Devices
   • Often as a second screen as one big portion of mobile device use is at home.
   • End users often use several screens parallel.
   • Already, communication on Facebook and WhatsApp today wins over communication
   • Apple, Amazon, Microsoft and Samsung have been experimenting with voice recognition at home for quite some time.
4. Online-Video and Digital Out of Home Advertising
5. Content Marketing and Native Advertising

New Data and Findings as Programmatic Advertising Drivers

• Big Data Analysis
  • The explosive increase in data through digital media, mobile devices and social networks is facing a huge increase in data processing capacities.
  • While most companies are still struggling with the sheer enormity of processing the data, other companies already use relevant findings drawn from the environment. Machine learning thereby will enable systems to derive better and better forecasts.
• Web of Things and Virtual Reality
  • With the proliferation of the Internet of Things, (“ Web of things,” the connection of an almost unlimited amount of networking devices on the Internet), this trend will accelerate again.

Risks in the Programmatic Advertising Development

• Restrictive Regulation of Data Protection
  • there is the risk of a too restrictive or even prohibitive regulation of data use.
  • More and more citizens perceive the use of their data as being close to the edge or even illegal.
• AdBlocker as Self-defense of the User
  • They install ad-blockers and regularly delete their cookies
  • Behind the scenes there is a technical arms race happening, examples are anti-Ad-blocker services or new identification technologies such as fingerprinting, the cookie-free detection of users according to specific characteristics such as their PC configuration.
• Limited Branding Capabilities
  • Despite many efforts Digital Marketing often fails to deliver results for branding.
  • On the other hand, this can be attributed to the way we use digital media.
  • Most of the time we use it as a service, searching things, skipping through news streams, buying stuff, doing communication – getting things done.
  • Not a good branding environment.

Conclusion and Outlook

• Who would have thought years ago that today a majority of smartphone users willingly and continuously allow themselves to be located, and that they would be okay with transmitting this highly sensitive information using apps?
• Video advertising and native advertising channels are strongly sought after formats, particularly because of banner blindness

Part II: Components

2.1 Consumer-Centric Programmatic Advertising

The Evolution of Programmatic Advertising from a Performance Channel Only to a Method to Buy a Broad Range of Digital Media

Re-targeting is a very powerful tool to re-engage with consumers that have shown an interest but have not converted.

Programmatic Advertising 2.0: Audience Buying

• While Retargeting focusses at the very end of the purchase funnel, Audience Buying is more applicable for the upper funnel to reach consumer profiles that have not engaged with the advertiser yet.
• In Programmatic Advertising 3.0, advertisers move to a Consumer Centric view. They combine as much data as possible into a holistic user profile, combining sophisticated segmentation of the website visitors with CRM and even sales data and enriching it with 2nd and 3rd party data.
• Traditionally, media planners allocate budgets across media channels (on- and offline) to selected placements (based on affinity, reach or historical performance) to buy impressions that will hopefully reach the right consumer segments

The New Consumer Centric Advertising Model Starts with the Consumer:

DSP (Demand Side Platforms) sees a consumer profile and determines the value of an impression right now:

• Right context?
• Right quality?
• Right time?
• Right location?
• Right price?

First Party Data: Maximising the Value and Control of Owned Data

• First party data is the data that advertiser generates through direct consumer interaction on his owned assets, CRM and customer databases.
• It is the most valuable data as those consumer profiles have already engaged with the advertiser so are therefore usually much more likely to convert into the desired action.

Third Party Data: Worth the Money?

• Third party data is data that can be rented from data vendors.

Access to Media Inventory: Open Auctions or Private Deals?

• As Programmatic buying moves beyond Performance marketing, advertisers put much more emphasis on the “quality of inventory”.
• They want access to premium inventory, high impact ad formats and to be rewarded with lower prices for spending higher budgets or even going into spend-commitments with publishers.
• And they want to know where their ads run and work with publishers they know.

Brand Safety: Preventing Fraud, Controlling Viewability and Audience Reach

• Shocking reports on fraud, low viewability and brand safety claim that 57.4 % of ads bought via networks and exchanges are not viewable, that 14.5 % of those impressions are fraudulent and 17.5 % have a brand safety risk.
• Applying blacklist of known “bad” sites is the base level that should be done but has limited long-term effectiveness as fraudsters launch new sites every day.
• The most shocking examples of brand-damaging ads happen on respected news sites, where airline ads appear next to plane-crash reports or confectionary ads next to articles about obesity.

Customising Messages: Dynamic Creative Optimisation

• Dynamic Creative Optimisation (DCO) allows the automated assembly of ads from multiple components (copy, image, call-to-action, offer etc.) to give each consumer the right message at the right time to drive them to the most effective reaction.
• Simply showing the same products that the user has looked at a high frequency may generate some clicks and sales but also alienate many consumers as they may feel stalked.

People: Evolving Skillsets for Granular Optimisation of the System

• Some still believe that they only need a login to a DSP, push some buttons and then the algorithms do the rest.
• While some reach-focussed tactics like retargeting and audience buying benefit from large data and machine-learning based automated optimisation, the need for smart people has not decreased but is in fact increasing.
• Advertisers and agencies have learned that the workload of complex programmatic campaigns is higher than traditional display buying as so much more data can and needs to be analysed, used and optimised.
• In traditional display buying, the team sees the number ad-impressions, clicks and conversion per media buy and can change the creative or cancel the buy.
• In Programmatic Advertising (within the service, hybrid or product models), the team can get full transparency about every site, every impression, every user segment, optimise in real-time based on those, time-of day, day of week, geography and more.
• Algorithms will only pick up after long learning periods and human optimization still beats the algorithm in many cases.
• While Search may be twice as labour intensive as traditional display buying, Programmatic Advertising is at least three times more complex.

So we see the evolution of digital media planners into programmatic planners and the addition of data consultants and data scientists into the teams.

• Programmatic consultants
  • The evolution of the digital media consultant helps to understand the landscape, opportunity and risks and define the Programmatic strategy
• Programmatic planner
  • Builds the plan, defines KPI and benchmarks, inventory strategy, audience segmentation, messaging plan and optimisation rules
• Programmatic optimiser
  • Is the hero in the team, working on the granular levers to drive Return on Investment by optimising based on deep analytics of all available data
• Inventory manager
  • Negotiates private marketplace deals, monitors brand-safety and inventory quality and works with publishers on innovative formats and data segments
• Data consultant
  • Works with the web-team, CRM, DMPs, software engineers and data providers on the “data plumbing” to ensure seamless data flow while protecting data ownership and privacy
• Data scientists
  • Analyses the data with statistical models and machine-learning algorithms to generate deeper insights and optimisation rules

Moving Beyond the Web: Programmatic TV, Outdoor, Radio

• Digital Outdoor is already starting to be ad-served, so buying it programmatically is the next logical step.
• However, the one-to-one relationship of consumer profile and ad-space as in Display advertising will not happen: Outdoor ads can be seen by many people at the same time, most of them will not (want to) be recognised, so using user profiles to drive Outdoor buys will not be the core use-case.
• However other data signals like weather can influence the value of an outdoor ad.
• Being able to launch campaigns in minutes and change ads in seconds may make programmatic outdoor interesting for both buyers and seller.
• The same is true for Radio, but as music listening moves into one-to-one digital channels with web radio and music streaming services, opportunities for programmatic buying are already becoming available.
• TV, however, is the largest advertising channel and the power of video to drive both brand values and actions is unparalleled.
• TV usage is still growing in most markets, but younger audiences move more and more into non-linear video-watching online, so advertisers want to do integrated screen planning, combining traditional TV with online video, to maximise reach and impact.
• In the immediate term, we will see much more innovation happening on the web and mobile rather than in the traditional TV world, which is also limited by regulation.

2.2 Understanding Demand-Side-Platforms

An IP address in the USA however is considered to be like a postcode that doesn’t change and therefore represents an important optimization variable.

The algorithm, for example, is decisively based on the IP address and this criterion has been dropped altogether in Europe. Correspondingly, it would be very unwise to adopt an identical American algorithm in Europe.

2.3 Granularity Creates Added Value for Every Objective

Full Funnel Marketing

• Prospecting or Branding
  • The industry standard of a viewed impression is fulfilled as soon as more than 50 % of a banner is viewed for at least 1
  • Due to their high degree of viewability and effectiveness, video advertising media, rich media display and the Facebook Newsfeed are especially conducive to achieving high-level advertising impact, while standard display advertising media offer a cost-effective increase of reach.
• Predictive Targeting
  • The probability of a user’s behaviour, such as the probability of him purchasing products on websites of the advertiser, which this user has not yet visited, can be determined from information about the user’s technical characteristics, the context of visited sites and his anonymous behavioural data derived from external sources.
• Retargeting
  • Which data is actually the most effective in Programmatic Advertising? It is the advertiser’s own – so-called 1st-party data.
• Marketing to Existing Customers
  • A successful customer acquisition marks the beginning of further cross- and up-selling measures using classical CRM measures such as e-mail marketing, as well as other online marketing channels through which existing customers can be retargeted.

Channels and Advertising Formats

• Display
• Mobile
  • The distinction between tablets and mobile phones is becoming increasingly blurred.
• Video
• Social
  • Facebook Exchange (FBX) loses relevance as Facebook develops its own capabilities further.
  • Networks like Instagram, Snapchat, Pinterest, Tumblr, WhatsApp or Vine are getting integrated in the standard marketing mix of each and every professional brand.

Devices

• desktop PCs and laptops,
• mobile devices like mobile phones and tablets.
• connected TVs.
• Digital out of home and radio

2.4 Enhanced Success with Programmatic Social Advertising

Social Media Today

• A key characteristic of social networks is the amount of information that the users are willing to share.
• A wealth of information about its users is available, often self-reported by the users themselves. This can be a great opportunity for marketers to take advantage of to target a particular demographic segment or an audience based on self-reported user data.
• Additionally, the value of an ad exposure can be amplified by the social context on social networks. It is in our human nature to pay closer attention to the actions taken by those close to us and in our social circles.
• Facebook currently offers several solutions: Website Custom Audiences (WCA), Facebook Exchange (FBX), and Dynamic Product Ads (DPA).
• WCA is a retargeting vehicle, available by placing a cookie on users visiting a particular website.
• They can then be recognised on Facebook and marketers can target them as a particular audience.
• Similarly, FBX is a real-time ad exchange for reaching users that visit a particular website using a cookie, who can then be targeted on Facebook.
• While Facebook DPA allows marketers to automatically promote relevant products from their entire product catalogue across any device.

Atlas data has revealed that cookie and browser based views of the purchase journey are on average leading to a 26 % overstatement of reach, a 41 % understatement of frequency and at least 21 % of conversions are not being captured Last click attribution models do not include the power of mobile advertising in driving interest for products that are bought on desktop devices where transacting is typically easier.

Therefore, existing attribution models are largely inaccurate as they give much of the value to desktop clicks, which are actually driven from other (mobile) devices.

Even multi-touch models cannot track mobile or Custom Audiences impressions, since view tags do not work in mobile apps, on custom audience targeting, or on many mobile browsers.

MediaMath’s Joe Zawadzki sees the future of Programmatic Advertising as “a single dashboard where brands will be able to track all of the ways they interact with those individual consumers they care about, and who care about them. With sophisticated attribution modelling, advertisers will be able to consolidate a full array of customer interactions across a multitude of channels and devices into a holistic picture of consumer engagement.”

2.5 Programmatic Brand Advertising

The volumes of inventory available on ad exchanges seem to be unlimited: in France, for example, 100 billion ad impressions are for sale programmatically on supply side platforms every month.

2.6 The Creative Challenge

How to Transform Programmatic Media to Dynamic Brand Messaging

• At the heart of the change lies the desire of any marketeer: I want to pay more to talk to people who might actually buy my product as opposed to people who will not, at a time and place that is conducive to both my consumer and my brand.

Unleashing the Power of Greater Creatives for Brands

• There are major massive offline budgets that have not yet even hit the digital space as there is a creative problem in Programmatic Advertising.
• The DSPs are increasingly adept at deploying high-impact campaigns via Programmatic Advertising. The brands don’t always know this, though. This leaves the DSPs in a position of educating their clients about their high-impact capabilities.

2.7 Cross-Channel Real-Time Response Analysis

• The Contribution of Measurement in a Cross-Device, Data-Driven, Real-Time Marketing World
• Programmatic has also been applied to TV advertising; Simulmedia is one example for this development.

2.8 How to Be a Successful Publisher in the Programmatic World

“what is the next big trend in consumer engagement”?

Some believe that wearable computers could be the next revolution.

• With wearables, health data in particular comes into play for advertisers.
• The device tracks how sedentary one has become in recent weeks and that could in theory be used to help encourage users to find the nearest running club or gym via advertising.
• Bio data being shared with advertisers is not something consumers have been widely exposed to so far, but the potential for this device to help live a fitter, healthier life is significant, and actionable bio data for the advertiser unlocks a tremendous opportunity.

The Next Step in Programmatic: Automated Guaranteed

The advertising industry has seen the rise of real-time bidding and private marketplaces over the past years. However, these two segments are potentially overshadowed by the market for guaranteed orders, where inventory is still largely bought and sold manually.

Part III: Transformation

3.1 The CMOs Challenge

Fraud

• Little incentive to stamp it out by definition (Peterson and Kantrowitz 2014).
• Publishers make money from it, buyers’ performance looks great, and technology companies get paid to help stop it.
• In most cases, brands end up holding the bag and pay the bill.

3.2 Integrated Campaign Planning in a Programmatic World

1. Choosing a Framework for Programmatic Campaign Planning
   • For Programmatic Advertising, Avinash Kauskik’s See-Think-Do-Care framework works very well:
     
2. Briefing
   • Clearly Defined Advertising Objectives
     • It is key to map campaign objectives against the user’s consideration stage and then deduce the relevant KPIs.
     • Finally, it is essential to define a singular core KPI for success.
   • Well Defined Strategic Target Audiences
   • Clear Picture of Existing Digital Platform Architecture
     • This is not limited to media platforms such as ad servers demand side or bid management platforms, but should also consider CRM databases.
   • Wider Channel Mix and Strategy
     • build a picture of the marketing channel portfolio, as well as the topline strategy in place.
3. Measurement
   • Measurement must be clearly defined and agreed by all stakeholders upfront
   • Measureable Actions
   • Centralized and Integrated Tracking
   • Attribution Models and Campaign Optimization
     • Once tracking has been centralized and all success indicators defined, a vital decision that can easily be forgotten about is how to accurately attribute credit back to the right media sources. The default position is ‘last click’ which carries well-known deficits and often may lead an advertiser to not fully appreciate all consideration stages of See-Think-Do correctly.
4. Audience
   • Existing Customer Data Beats Assumptions
     • Existing customer data is gold dust. Not only is it a totally unique data-set to an advertiser, but it is the most powerful tool to discover key audience segments.
   • Site Visitors Tell Many Stories
   • Audience Profiles and Underlying Behaviours
   • Low Hanging Fruit Within CRM Data
5. Campaign Strategy and Set Up
   • Inventory Sources and Buying Technologies
   • What are the Priorities: Performance or Premium
   • 1st Party Data is the 1st Choice for Targeting
     • To add scale to 1st party data look-a-like strategies can be developed to algorithmically identify similar users to the top performing audiences.
     • If a campaign is for new customer acquisition, it is recommended to block all of the existing customers to minimize wastage.
   • Addition of Targeting Features to Scale Prospecting
   • Campaign Structure Makes a Big Difference
   • Recency and Frequency
     • There are few things more annoying to consumers online than being stalked by retargeting campaigns.
   • Controls and Media Quality
6. Creative Set Up
   • Granular Campaign Structure and Matching Creative
   • Targeting Data Informs Personalized Creative
   • Sequencing and Storytelling
7. Landing Experience
8. Campaign Management and Optimisation
   • Sufficient Data is Important for Informed Campaign Optimization
   • As Campaigns Develop, Humans Need to Guide The Machines
     • Machine optimization engines are great at making rational and logical decisions based on all the data they have available to them – but they can’t make decisions they haven’t been trained for.
     • Machines can’t create new tests, won’t appreciate trends or events happening in real life or know that there is a key seasonal date approaching unless you tell them.
   • Continuous Optimisation Drives New Campaign Briefs
9. Conclusion
   • Potentially the biggest issue programmatic advertising is facing is actually one that the whole digital industry is facing together – data.
   • Consumers are becoming more aware of personal data and privacy regulations.
   • Lawmakers are sometimes working with rules and policies that are still not quite fit for purpose and strict local data protection philosophies can occasionally close doors, which are open to many others.

3.3 Evolution of Digital Campaign Design and Management

• Big Data Becomes Relevant Data
  • Let’s be honest: Nobody cares about ad impressions or click rates. The advertiser either wants to generate sales, or at least lay the groundwork for a sale or improve his key brand KPIs such as brand awareness, preference/ top of mind or buying intention.

3.4 Realtime Data Accelerates Online Marketing

Differentiation Between 1st-, 2nd- and 3rd-Party Data

• 1st Party Data
  • “1st party data is information collected directly and stored by website publishers, retailers and other types of companies about their site visitors or customers.”
  • Because companies with this information have a prior relationship with their customers, they are able to use this first-party data – which may include names, addresses, phone numbers, site-interaction data and information about products purchased – to communicate directly with them.”
  • Search requests and search behavior of online users in Search engines, and especially the subsequent activities or transactions, deliver another very important data source.
  • All activities after the click on the sponsored link, means we are talking about 1st-party data, whereas the search information is 2nd-party data.
• 2nd Party Data
  • Data held by publishers collected from their sits or assets is highly valuable and relevant, but not very easily available directly to a marketer. Most publishers prefer to leverage data for their own benefits and do not see the value in selling it in an open market.
  • A far better proposition for them is to work with selected advertisers and match their data sets with the advertisers 1st party data.
  • In some markets, 2nd party data has improved effectiveness in the absence of a 3rd party data marketplace.
• 3rd-party data.
  • These include socio demographic data like age-, gender-, location-data, psychographic data like behavior-, opinion- or interest-data.

3.5 Redefining Retargeting

The Art of Finding the Right Attribution Model

• Lesson #1: Challenge Whatever Your Model Tells You
• Lesson #2: Stop Assuming, Start Testing
• Lesson #3: Focus on Touchpoints That Truly Have Influence on the Purchasing Ecision

Mobile Is the New Black

• Three recently-developed technological advances are helping them to monetize these valuable user bases:
  • Advance #1: In-App Advertising
  • Advance #2: Dynamic Ads
    • highly targeted dynamic ads customized to the users’ individual experience, provide a significant increase in value and boost the probability of a conversion.
  • Advance #3: Deep-Linking
    • Mobile deep linking creates a direct link to any location within an application.

Why Ad Blockers Are Not Useful for the User

• Since the dawn of advertising, there has been some resistance to advertising, particularly in regards to advertising on the ‘cost-free’ internet.
• The internet is not free.

3.6 Pricing for Publisher: Scaling Value, Not Volume

Pre-Negotiated Deals (Deal IDs)

• Unlike true auction-based realtime bidding, the Deal ID is used to determine together with the advertiser at what price he can purchase the ad impression rule-based via his DSP from the publisher.
• However – and it is very important to understand this – a predefined price with an advertiser does not necessarily exclude a bid on an impression. Why should it.

Pricing Strategies for Publishers

1. Fixed pricing (without auction)
2. First price auction (with RTB)
3. Second price auction (with RTB)
4. Dynamic floor pricing (with RTB)

Fixed Pricing

1. Rather, the SSP uses the predefined CPM prices of the marketing unit.
2. However, Programmatic Advertising only reaches its full potential in an auction-based process rather than in fixed pricing.

First Price Auction

• For higher priced inventory classes, it is recommended to use the first price aution.

Second Price Auction

• The bid of the buyer determines the price in a second price auction. In this type of auction, the highest bidder has to pay only the second highest bid plus 1 cent.
• Empirically, this auction model is more suitable for inventory classes with lower floor prices, which are intended for the largest possible number of bidders.
• The average CPM evens out very dynamically.
• The second price auction is the fairest auction model unless the publisher optimises with dynamic floor pricing at the same time.

Dynamic Floor Pricing

• Dynamic floor pricing is an optimisation opportunity where the SSP adapts the floor price based on the demand.

Client Trends

• Programmatic Advertising facilitates the direct access to high-quality inventory and interesting audiences. Many businesses and brands have therefore started to place their advertisements directly with the publishers, thus avoiding media agencies. That is why an increasing number of advertisers will have their own “seat” in the SSPs in future.

Common Arguments Against Programmatic Advertising for Premium Publishers

• Programmatic Advertising leads to a loss of control and cannibalises one’s own publishing business: The publisher can protect his existing campaigns and revenues through whitelisting and blacklisting.

Your Thoughts?

This concludes my notes on the book “Programmatic Advertising: The Successful Transformation to Automated, Data-Driven Marketing in Real-Time” by Oliver Busch.

What did you think? Was this summary helpful? Have you read the book yourself?

Have your say in the comment section below.