A Glimpse of AdTech
This articles aims at providing a high level overview of the AdTech industry. The stakeholders, key players and potential opportunities coming up.
Advertising has evolved significantly over the years. From billboards on buildings to embedded ads on web pages, the industry has flourished. Reaching a global value of $760 billion in 2021 and continuing to grow at a steady rate.
Overview
AdTech can be defined as: All the tech tools to plan, execute, manage and analyse digital ads campaigns.
Which is not helpful to understand the domain in itself.
In my experience, taking an historical standpoint and going through the changes AdTech went through is the best way to grasp the ecosystem.
The Early Days (mid-1990s)
At the time, websites started to pop up all around the world. Personal blogs, news websites, brand websites, … They will be later known as publishers.
Publishers: Any digital platform that provides ad space.
On the other side, you have brands that want to promote their products/services. These brands (called advertisers) are already promoting their product through other means. But the growing Internet audience starts to be appealing to them.
Advertisers: Brands that want to promote their product/services.
At the time, the advertisers would simply contact the publishers directly to buy their ad space/ad inventory so they can showcase their product on the publisher’s website.
Ad space/Ad inventory: Space on the publisher’s website that is left empty for an ad to be displayed.
Technically speaking, a server was used to store and deliver the ads to the publisher’s website. A so-called ad server.
You could also see the beginning of functionalities like :
- Basic targeting based on the publisher’s audience: Since the publisher has information about the visitor, it can be forwarded to the ad server that determines which ad to deliver based on the advertiser’s targeting config
- Tracking the ad performance
- Frequency capping (limiting the max number of times a visitor would see the ad)
Here’s a simple schema below to sum up:
Birth of Ad Networks (late-1990s)
As you can imagine, this process was both manual and tedious. There was a clear need to streamline the buying and selling of ad inventory, making it more efficient for both publishers and advertisers.
Such was the beginning of ad networks.
Ad Network: Platform which aggregates ad space from publishers and sell it to advertisers.
In practice, ad networks use fixed pricing models where advertisers purchased a certain number of ad slots which would be distributed across the network’s ad space.
Say Hello to Cookies (~2000s or so)
Targeting is a very important concept in AdTech. It’s about showing the right ad to the right visitor.
And there’s a equation coming along with it:
Better Targeting = Better Ads = More money
This is where comes into a play something you may already have heard of, the cookies.
While the initial purpose of cookies was to store session information, they were quickly used to track user activities across websites.
At the time, the cookies were sent to the ad networks to build a profile used to deliver the most relevant ad to the visitor.
Note: Cookies also enabled retargeting strategies, where users who visited a website could be served ads for that site as they browsed other sites
Rise of Ad Exchanges (mid-2000s)
To replace the ad networks, ad exchanges were introduced.
Ad Exchanges: Platforms where publishers and advertisers can buy and sell inventory in realtime. This happens in milliseconds when the visitor lands on the page.
This process of selling the inventory in real-time is called Real-Time Bidding (RTB).
Ad exchanges offer 2 main advantages over ad networks:
- Automated Biding: Which reduces the need for manual negotiation
- Dynamic pricing: Price determined based on demand
The development of SSPs (late-2000s)
But as every tool, ad exchanges were doomed to be replaced by a better version of themselves.
This is where Supply-Side Platforms (SSPs) come into play.
Supply-Side Platforms: Like an ad exchange but under steroids.
SSPs are platforms that connect with multiple ad exchanges (or ad networks) to aggregate the bids of many advertisers.
SSPs offer several advantages over ad exchanges:
- Demand Aggregation: SSPs make the publisher’s ad space available to multiple ad exchanges. Which means greater audience, thus more competition, thus more demand, thus better yield for the publisher
- Private Platforms: SSPs can establish private platforms, where publishers offer their ad space directly to selected advertisers or agencies
- Programmatic direct deals: SSPs facilitate programmatic direct deals, where publishers and advertisers negotiate and agree on terms directly but execute the transactions programmatically (so skipping the RTB process)
- Inventory Management: SSPs provide tools for publishers to manage their ad space like setting floor prices, prioritising certain buyers, …
- Detailed Reporting: SSP report to the publisher who is buying the ad space, at what price, and how ads are performing
The creation of DSPs (late-2000s)
On the advertiser side, ad exchanges also had their setbacks.
This is where the advertiser twin of SSP comes into play: The Demand-Side Platform (DSP).
Demand-Side Platforms: Platforms that allows advertisers to buy ad space across multiple ad exchanges with advanced targeting options (behavioural, contextual, …).
To put it simply, SSPs and DSPs are just the evolution of ad exchanges tailored for the specific needs of publishers (SSPs) and advertisers (DSPs).
The creation of DMPs (late-2000s)
Remember the equation:
Better Targeting = Better Ads = More money
So there was (and still is) a momentum to improve our targeting logic. This is where the Data Management Platform (DMP) shines.
Data Management Platforms: The DMP is a centralised platform to aggregate customer/visitor data from multiple sources (CRMs, cookies, mobile apps, third-party databases, …).
The DMP segments the users into categories (tech buyers, sport enthusiasts, …) and then integrates this data in DSPs and SSPs for better targeting.
In practice:
- Advertisers use DMPs to collect and manage data from their own websites, CRM systems, and other sources (like third-party data). With this data, they segment their audience and make sure the visitor the bid is for corresponds to one of their targeted segment (by integrating with DSPs)
- Publishers gather data (time spent on the pages, clicks, …) and segment their customers as well. This information is sent to the SSP and used to sell their ad space at a better price
Note: You may have heard the first party alternative to DMP, CDP (Customer Data Platform). CDP is basically DMP but specifically designed for first party data.
Alternatives to cookies (late-2010s)
Cookies were used everywhere: ad networks, ad exchange, SSPs, … They are a great tool to improve the ads’ targeting.
But concerns about privacy grew. And regulatory changes (GDPR, CCPA, …) caused the downfall of cookies.
These regulatory changes enforced strict rules on how personal data is collected, stored, and used. This made the major browsers phasing out support for browser third-party cookies.
But targeting is not a nice to have in AdTech. So many alternatives came out the past years like:
- Contextual Targeting: Contextual targeting means that the ads displayed are based on the content that the user views rather than the browsing history
- First Party cookies
- …
Step by step
Summing up what we’ve seen so far:
- Publishers list their available ad inventory on SSPs
- When a customer arrives on the publisher’s website, a bid request is sent from the SSP to the ad exchange.
Including information about the visitor, the context of the content and any other relevant targeting data (cookies for instance). This data can come from the publisher’s DMP. - DSPs receive the bid request from the ad exchange and determines whether to bid on the inventory and how much to bid. That decision is based on the advertisers targeting criteria (also created with a DMP) and its budget.
- The ad exchange now conducts a real-time auction, considering bids from multiple DSPs. The highest bid wins the auction and the corresponding ad is selected for delivery.
- The ad server delivers the selected ad to the publisher’s ad space and tracks various metrics such as impressions, clicks, …
- The performance of the ad is tracked and reported to both the advertiser and the publisher via their DSPs/SSPs.
Now we have a good understanding of the key players work together. We can move forward to the other mediums where digital content can be used as a platform for advertising.
A multi-platform ecosystem
Ad space can be sold in different platforms:
- Web
- Mobile
- Audio
- …
No matter what platform you use, you always have the concepts of DSPs, SSPs, and so on. The main difference (but not the only one) is the nature of the ad space proposed by the publishers.
With all the different platforms we use day-to-day, a new challenge came out : Cross-device targeting.
Cross device targeting
Being able to track a user across devices to ensure a consistent experience is key in AdTech. That is to say, if I switch from my phone to my laptop, I want the targeting logic to stay the same.
Which raises the question : How do you match the same user on different devices ? There are also several options:
- Deterministic matching: For instance, when users login to the same account across different devices, it provides a reliable way to identify and target the same user across all devices.
- Probabilistic matching: Uses algorithms to infer the likelihood that different devices belong to the same user. The algorithms can use the IP addresses, the location, browsing behaviour, …
- …
But of course you need to test and optimise your ads for them to reach their full potential, this is where a new area came into play: Creative Optimisation.
Creative Optimisation
Here, we’re talking about the tools to test and optimise ads so we can improve its performance and engagement.
Here are some key functionalities of those tools:
- A/B testing: Tools to test different versions of an ad to see which one performs better
- Dynamic Creative Optimisation: Tools to automatically adjust the content of an ad in real-time based on user characteristics, behaviours, or context. For example, a travel ad might show different destinations based on the user’s location or browsing history
- Creative Management Platforms: Platforms that provide a central hub for managing, organising, and distributing ads. They streamline the process of updating and deploying ads across multiple channels and formats.
- Performance tracking: Tools to provide detailed analytics on how different creatives are performing. Metrics might include click-through rates, conversion rates, …
Note that a lot of DSPs are already implementing those features.
Future Opportunities
I hope you have a better understanding of how AdTech works.
If you’re looking for an AdTech tool, remember that the line is a bit blurry sometimes. For instance, Google Ad Manager is both an ad server and a DSP, most of DSPs include creative optimisation tools, …
That being said, I hope it was somehow useful to you, let me know if you’d like me to dive deeper into a specific topic!
