The Programmatic Advertising Ecosystem Explained [The Complete Guide For 2022]
Programmatic advertising is one of the most confusing sections of the great AdTech jungle.
You’re exceptionally brave for exploring it!
Even top authorities in AdTech like Digiday can’t help but proclaim that programmatic ads are a tricky subject to navigate.
The good news?
You’re on the right path to understanding it all!
This roadmap will guide you through all there is to know about the industry as you map out your own understanding of the programmatic advertising ecosystem.
- Preface: Programmatic advertising glossary
- What is programmatic advertising?
- The programmatic advertising ecosystem explained
- What is RTB?
- The benefits and importance of programmatic advertising
- The future of programmatic advertising
- How can AdButler help with your programmatic advertising strategy?
Programmatic advertising is a complicated topic to say the least.
That means that it’s seriously packed with jargon and terminology that an aspiring “AdTechian” may (almost certainly) not be familiar with.
If you find yourself stumbling over unfamiliar words, be sure to refer back to this “field guide” of commonly used programmatic terms to re-establish your bearings before pressing on!
- Ad Campaign
- Ad Exchange
- Ad Inventory
- Ad Network
- Ad Ops
- Ad Server Priority
- Ad Tag
- Audience Segment
- Behavioral Targeting
- Bidding Strategy
- Contextual Targeting
- Demand Source
- Deterministic Matching
- Device Fingerprinting
- First Party
- Floor Price
- Frequency Capping
- Guaranteed Deal
- Header Bidding
- Media Buying
- Open Auction
- Open Marketplace / Open Exchange
- Preferred Deal
- Premium Inventory
- Probabilistic Matching
- Remnant Inventory
- Supply Source
- Third Party
- User Data
(It’s a long list – but these terms will surely help you to successfully navigate the jungle of programmatic advertising with confidence!)
Programmatic advertising refers broadly to the automated process of buying and selling digital advertising through the use of various programmatic software platforms.
In digital advertising, the term “programmatic” is frequently used interchangeably with the term “automated”, meaning that once the terms of an advertising deal are locked-in, a programmatic ad campaign is executed automatically according to those terms.
With minimal guidance from a campaign manager, artificial intelligence and machine learning are able to assist in the optimisation of a programmatic campaign’s performance over time.
Rather than having an ad ops team determine when and where to display ads, programmatic advertising makes these decisions based on the data it has access to about a visiting user.
Programmatic ads are the same as any other type of ads found online.
The only difference is that programmatic ads are served to a user’s web browser by using programmatic buying and selling processes.
If various programmatic platforms and software make decisions about when and where to display an ad, that ad is considered programmatic.
If an ad ops team is manually scheduling when and where to display an ad without the use of programmatic technologies, that ad is considered non-programmatic.
Programmatic buying, also referred to as programmatic media buying or programmatic marketing, refers to the act of an advertiser purchasing ad inventory from a publisher by using automated programmatic techniques.
In contrast, programmatic selling refers to the process of a publisher automatically selling their ad inventory by connecting it to one or more programmatic demand sources.
Some examples of programmatic buying techniques that are used to determine when ads should be purchased and where they should be displayed include:
These techniques are used alongside a number of targeting factors, allowing advertisers to fine-tune their media buys for maximum audience reach and performance.
Programmatic targeting refers to a number of different ways in which advertisers are able to specifically serve ads to audience segments that align with their ad campaign’s objectives.
Some examples of these targeting techniques include:
- Behavioral Targeting
- Contextual Targeting
- Deterministic Matching
- Device Fingerprinting
- Probabilistic Matching
- User Data
Advertisers are able to maximize their ROAS (return on ad spend) by maximizing the number of impressions, CTR (click-through rate) and CVR (conversion rate) of their displayed ads by making them more relevant to their targeted audiences.
There are many different types of programmatic buying deals which can be struck between an advertiser and a publisher – these deal types are covered in more detail later in the guide.
Some deals involve manual negotiation between both parties prior to initiating the programmatic ad campaign, while others involve no human interaction whatsoever.
In a situation involving manual negotiations, both parties establish the terms and conditions of the deal, often (but not always – it’s flexible, like any negotiation) including details such as:
- How long the ad campaign will run for
- How many impressions, clicks, and/or conversions will be received
- What model will be used to calculate price (CPM, CPC, CPA, CPI*, etc.)*
- The ad server priority level at which the ad campaign will operate
- When and where ads will be displayed
- Other details exclusive to the particular agreement being created
Once the deal has been created, the programmatic platforms used by each party will conduct the campaign according to the terms of the deal.
As visitors arrive at the publisher’s website, they’ll automatically be served ads from the advertiser’s ad campaign if they meet all of the programmatic criteria outlined in the deal.
However, it’s also possible to purchase programmatic ad inventory while forgoing the negotiation process entirely – which is the most common use of modern programmatic technology, due to its efficient nature.
In a situation where no negotiations take place, each party simply configures their programmatic technology platforms to automatically buy and sell ad inventory when their specified criteria are met. For example:
- Publishers may set a floor price (a minimum monetary amount) at which to sell their inventory
- Publishers may indicate which of their ad tags (or “zones”) to sell programmatically
- Advertisers may adhere to a particular bidding strategy that adheres to their budget
- Advertisers may also specify their preferred targeting techniques and strategies
When each party has configured their platforms to their liking, the process of buying and selling ads becomes automated – creating a highly efficient method for transacting media.
Let’s look at an example of why and how a publisher and an advertiser in the automotive industry might conduct a programmatic media transaction.
Bob is a car enthusiast and owns a blog.
On his blog, he frequently writes content about car performance and repairs.
Bob wants to earn revenue through the content he creates – so he sets up a number of ad tags on his website to allow ads to be served to them.
Bob then connects his recently created ad inventory on his website to a number of demand sources, where thousands of eager advertisers are running campaigns.
Alice is one of those advertisers.
Alice is a marketing executive for an automotive dealership, and has been tasked with increasing digital traffic to the brand’s website.
As part of her marketing budget, Alice has allocated some funding towards programmatic advertising.
After defining her ideal target audience, bidding strategy, and a timeframe, Alice launches an ad campaign to attract new visitors to the dealership’s website.
The programmatic platform used by Alice scans thousands of publisher websites, and determines that Bob’s readership audience would be a good match for one of her ads.
When a user visits Bob’s automotive blog, an auction is conducted for the available ad inventory, and when Alice wins the auction against other advertisers, her ad is displayed.
Bob generates ad revenue from the impression he provided to Alice, while Alice receives an impression and possibly a clickthrough to the dealership’s website.
This is a basic example of how the programmatic process works, but the details of what exactly is taking place “under the hood” is covered a bit later in the guide.
There are countless examples of brands capturing major returns through programmatic ads.
There are a few ways to interpret this question.
The cost of programmatic ads themselves depends on the format of ads being served.
It also depends on factors such as the niche in which an advertiser plans to operate, as well as the bidding strategy they’ve chosen to use.
Typically, programmatic display ads cost an average of $0.50 – $2.00 CPM (cost per mile or “cost per thousand impressions”).
Programmatic video ads can cost around $12.64 CPM according to a report from the Association of Canadian Advertisers back in 2017.
Of course, these prices are subject to fluctuation as the market demand for different programmatic channels shifts with the evolving landscape of AdTech.
Another angle to consider is the actual effectiveness of programmatic ad spend.
An article by AdTech expert Dr. Augustine Fou draws attention to some prominent examples of major corporations, including Procter & Gamble, Chase Bank, and Uber, all of which saw no performance decreases after shutting off their programmatic ad spend.
The article proceeds to highlight some of the risks associated with the lack of transparency in programmatic advertising, as well as the risk of ad fraud.
Another cost to consider are the usage and transaction fees of the programmatic platforms themselves, which contribute to the overall cost of operation.
Yet another cost to consider is that of hiring programmatic ad ops professionals.
While programmatic ads are mostly automated, individuals who are trained in the operation and optimization of programmatic ad campaigns are often a necessary investment.
An ad ops team will also be responsible for monitoring campaigns for the effectiveness of ad spend, and any suspected fraudulent activity as mentioned previously.
Finally, the overhead cost of sourcing and configuring the various platforms and components of the programmatic ecosystem can also take its toll on business resources.
However, this cost can be alleviated in part by consulting AdTech vendors (such as AdButler) who can share years worth of programmatic experience with interested parties.
Ultimately, despite its many costs, the objective of programmatic advertising is to yield a net-positive return for both parties – with publishers covering their costs through ad revenue, and advertisers turning a profit by selling more of their goods and services to consumers.
All of this talk about programmatic advertising may have piqued your curiosity about how it all got started in the first place.
The truth is that, like all things related to ad tech, it’s complicated.
The concise version is that ad exchanges were developed in around 2005 – platforms which introduced programmatic advertising to the AdTech ecosystem.
If you’re keen to hear the full story, feel free to head over to our article on the history of programmatic advertising which covers its impressive evolution over the years.
The programmatic ecosystem is a subset of the AdTech ecosystem – a collection of technology platforms which power modern digital advertising.
The ecosystem also consists of a number of available advertising deal types, methods of purchasing programmatic media, and spans across a variety of different ad formats.
The programmatic ecosystem is made up of many different platforms which communicate with one another in order to facilitate automated media transactions.
This means that there’s no single platform that operates alone to conduct programmatic advertising (if only it were that simple).
Each programmatic platform is typically owned and operated by either a publisher, an advertiser, or an intermediary which contributes various advantages to the overall process.
In the programmatic ecosystem, publishers (referred to as the “first party” or a “supply-source” in digital advertising) use platforms called SSPs (supply-side platforms) to manage the programmatic selling of their ad inventory.
The SSP is a platform which is typically owned and operated by an SSP service provider, and publishers pay these services to access and use them.
The DSP is similarly owned and operated by a DSP service provider, with advertisers paying a fee to access and use them.
Some of the largest publishers and advertisers may choose to build their own SSPs or DSPs, though the cost of this undertaking inhibits the majority of these parties from doing so.
As mentioned earlier, there are also platforms which compliment the programmatic process by conferring additional data to the automation of ad serving decisions.
In modern digital advertising, both publishers and advertisers make use of data brokers and internal data silos to improve the targeting functionalities of their ad serving process.
DMPs are large repositories of third-party data which can be drawn from to understand more about a user to which an ad is being served.
However, with the end of third party cookies right around the corner, most DMPs are either rethinking their business model, or rebranding entirely to offer the role of CDPs.
CDPs are platforms which are used to store a business’s first-party customer and lead data.
Because first-party data is only ever shared with a business with a user’s consent, its use in programmatic advertising is becoming far more pronounced as companies move away from using third-party data to accomplish their ad targeting.
Methods of anonymously associating first-party data with other information known about a user have also been developed – a technique referred to as identity graphing, which is becoming increasingly popular in the cookieless ad serving world.
Ad networks provide their service in the form of amalgamating or “bundling together” publisher ad inventory categorically for convenient purchase by advertisers.
In modern digital advertising, ad networks often facilitate these deals through programmatic arrangements that further help both publishers and advertisers to secure maximum value.
Ad exchanges are a bit different in their functionality.
They allow ads to be automatically bought and sold on an impression-by-impression basis through a process known as RTB (real-time bidding).
Unlike ad networks, human interactions are never involved on an ad exchange.
Instead, publishers and advertisers specify their conditions for buying and selling independently from one another. When a match between pricing and targeting is created, the media transaction takes place automatically.
It’s also worth mentioning that nearly all modern ad exchanges are built into SSPs as one of their components (a configuration which hasn’t always been the case throughout the history of AdTech).
If ad networks and ad exchanges still seem too similar for comfort, our article on the difference between ad networks and ad exchanges might help to clarify the confusion.
Additionally, if you’d like to learn more about the individual roles of any of the platforms which were just covered, our guide on ad servers vs DSPs, SSPs, DMPs, and CDPs is likely to contain the detailed information you’re craving.
We’ve covered a lot of information about programmatic advertising so far, but how do all of these different platforms work together to form a complete process?
Let’s assume that both a publisher and an advertiser have pre-specified their conditions for selling and buying ad inventory.
The various programmatic platforms we just covered might work together in a process that looks something like this:
- A visitor arrives at the publisher’s website, which causes an ad tag to be read by the visiting user’s web browser.
- An ad call is activated and made from the publisher’s website to any of the SSPs they’re connected to.
- The SSPs then forward the ad call to various demand sources that they’re connected to, which may include multiple ad networks and DSPs.
- Prior to forwarding the ad call to the demand sources, the ad tag associated with the ad call may first be enriched with additional user data from a DMP and/or a CDP.
- Once the ad call arrives at a demand source, numerous advertisers will place bids for the available ad inventory, based on their bidding strategy and any available data.
- Advertisers may connect to their own DMPs and CDPs to further enrich the amount of known information about a particular user prior to submitting their finalized bid.
- Once all of the bids are finalized, the publisher’s SSP selects the highest bid, and initiates the final process of serving the winning ad creative to the publisher’s site.
Despite seeming like a complicated series of events, this entire process often takes place in less than a second – completing before a webpage has even loaded in its entirety.
If you’re interested in learning more about how exactly the final process of serving an ad to a website is managed, you may be interested in taking a look at our complete guide to ad servers.
Is the room still stationary?
Not feeling dizzy from an information overload?
We’ve covered the basics of how programmatic advertising works – but unfortunately, there’s an entirely separate set of topics to unpack when it comes to programmatic deals.
That’s the wild nature of the AdTech jungle!
Covering programmatic deal types is seriously beyond the scope of this particular guide.
That’s why we’ve written a separate guide exclusively on the topic.
Once you’ve had a chance to take a breather (and perhaps a coffee and/or a stiff drink to steady your nerves), you’re invited to continue your exploration by visiting a complete walkthrough: The Different Types of Programmatic Advertising: Deals & Formats Explained.
However, to give you the simplest and quickest overview possible, there are 4 main types of programmatic advertising deals. They are:
- Open Marketplaces – Environments where any advertising party may bid openly on available ad impressions.
- Categorized as RTB.
- Private Marketplaces (PMPs) – Environments where publishers and advertisers are invited to participate in exclusive, “invite only” bidding processes.
- Categorized as RTB and/or programmatic direct, depending on how the arrangement is facilitated.
- Preferred Deals – Arrangements in which an advertiser is granted the right to bid first on a publisher’s ad inventory, also known as a “first look”.
- Categorized as programmatic direct.
- Guaranteed Deals – Arrangements in which a publisher and an advertiser agree on a number of impressions to be delivered at a fixed price during an ad campaign.
- Categorized as programmatic direct.
The term RTB or “real-time bidding” was mentioned earlier in the guide.
You may have already glanced at the glossary to learn it’s basic definition.
As a reminder, RTB is a technology that operates on ad exchanges to facilitate ad buying and selling on an impression-by-impression basis.
However, in the context of programmatic ad serving, there’s a bit more to understand about RTB in order to avoid confusion.
RTB is also known as “open auction”.
Don’t let the name fool you – the term “open auction” is the official name for RTB, and isn’t quite the same as an “open exchange”, which is a type of programmatic deal (still with me?).
Many people think that RTB is the same as programmatic advertising – but this isn’t the case.
While RTB is a form of programmatic advertising, not all programmatic advertising uses RTB.
Only certain types of programmatic deals are categorized as RTB. They are:
- Open Marketplaces / Open Exchanges (same technology, different names)
- PMPs (Private Marketplaces) (which are also sometimes categorized as “programmatic direct” – confusing, to be sure)
The other two popular programmatic deal types, namely preferred deals and guaranteed deals, are non-RTB forms of programmatic advertising, because the bidding processes used for these types of deals don’t rely on auctions to determine when an ad should be served.
We’ve created yet another dedicated guide that exclusively covers the intricacies of RTB.
In the guide, you’ll find information about why it’s so commonly confused with programmatic (primarily due to its massive popularity when it was first introduced in 2009), as well as links to even more advanced RTB techniques, such as waterfalling and header bidding.
The benefits and importance of programmatic advertising can’t be understated.
Since its inception in 2009, programmatic advertising has redefined the way that advertisers and publishers interact with their audiences.
For publishers, some of the greatest benefits of programmatic include:
- A simple and automated way to sell remnant and premium ad inventory alike
- Pricing calculations that optimize the value of available ad inventory
- An ability to improve ad relevance and enhance overall user experience
For advertisers, a number of programmatic benefits exist as well:
- Programmatic ad campaigns are capable of attaining high audience reach
- Ad campaigns can be adjusted in real-time to fine-tune performance
- Advertisers are able to target very specific audiences to improve their ROAS
Despite the uncertainty introduced to the AdTech industry since the beginning of 2020, programmatic advertising has a bright future.
Many of the industry’s most popular channels continue to support programmatic techniques, and new methods of delivering privacy-friendly programmatic ads are being developed.
As technology advances as a whole, new opportunities are emerging for programmatic advertising:
- AI (artificial intelligence) and machine learning are becoming more sophisticated
- Identity graphing and device fingerprinting are allowing advertisers to reach desirable audiences without compromising personal privacy
- Dynamic ad creatives allow ads to be more personalized based on user interests
Another major opportunity for publishers and advertisers lies in contextual programmatic advertising.
When third-party cookies cease to be used in online advertising, Google will be rolling out its new FLOC (Federated Learning of Cohorts) program to anonymously categorize users into contextual-interest groups known as “cohorts”.
In order to make use of this new technology, publishers are advised to begin paying attention to how the content is contextually tagged to ensure their properties are prepared for the new update.
If you’re a publisher who’s looking to get ahead of the curve on this trend, we invite you to contact the AdButler team to learn more about our contextual tagging solution.
We’re excited for you to tell us!
With over 20 years of experience in developing custom AdTech solutions, if you’ve got an idea in mind, AdButler has what it takes to help you pull it off.
Out of the box, AdButler supports any and all connections to the programmatic ecosystem.
We also provide native support for direct S2S ad serving and header bidding.
AdButler is a demand-neutral ad serving platform, meaning that we’ll never snoop through, sell, or use your data in any way – something that many ad platforms will be hesitant to confirm!
If you’re building a programmatic ad serving tech stack, we’d be thrilled for you to consider AdButler for the role of your go-to ad server – with 100% uptime since 2017.
And, speaking of programmatic advertising, did you know?
AdButler also offers publishers a self-serve portal feature that allows advertisers to purchase ad inventory directly through a branded advertising order form.
These orders are configured to actively compete with your programmatic ad inventory – providing you with the best opportunity to maximize your ad revenue.