Ad Servers vs DSPs, SSPs, DMPs & CDPs: Differences Clarified
It wouldn’t be adtech if there weren’t hundreds of terms and acronyms for us to decipher.
If you annunciated the abbreviations in this guide’s title aloud, anyone listening might think you’re a beatboxer.
If you’re lucky, maybe you can impress the panel on America’s Got Talent, get signed with an agency, and escape this whirlwind of confusing adtech jargon once and for all.
But, assuming you’re committed to walking the long and winding path of the “adtechian”…
Let’s sort out this POS (positively optimistic serving) terminology to avoid a FUBAR (failed understanding because adtech related) situation during your next digital advertising meeting.
- What is an ad server?
- What is a DSP?
- What is an SSP?
- What is a DMP?
- What is a CDP?
- The Ultimate Showdown: Ad Servers vs DSPs, SSPs, DMPs, CDPs & CRMs
Ad servers have been a core component of digital advertising since 1995 and have continued to play an important role in modern ad serving.
The fundamental purpose of an ad server is to store and manage ads from across different ad campaigns, to automatically decide which ads to serve to which visitors on certain webpages, and to gather data about how users interact with the ads displayed to them.
An ad server is a type of adtech (advertising technology) that’s used by:
- Publishers (referred to as a “first party” ad server when used by a publisher)
- Advertisers (referred to as a “third party” ad server when used by an advertiser)
- Ad Agencies (Teams that are hired to operate DSPs on behalf of advertisers)
- Ad Networks (Platforms which help publishers to promote and sell their ad space)
Features provided by ad servers include the ability to:
- Store media that comprise ad creatives (images, videos, audio, etc.)
- Automatically serve selected ads to visitors based using a set of defined rules
- Collect data such as impressions and clicks as visitors interact with ads
While an ad server possesses the same capabilities regardless of who uses it, publishers tend to focus on using ad servers to manage their ad inventory and reporting, while advertisers tend to focus on using them for managing the creatives of their ad campaigns.
In both cases, ad servers are used to conduct a type of advertising transaction referred to as a “direct deal”, meaning that a publisher sells their available ad inventory directly (as the name implies) to an advertiser, based on a set of manually negotiated terms.
When establishing a direct deal, the publisher’s (first-party) ad server communicates directly with the advertiser’s (third-party) ad server to deliver the advertiser’s ads to users who are visiting the publisher’s website.
In marketing and advertising, DSP stands for “Demand Side Platform”.
A demand side platform is a piece of adtech used by advertisers.
A DSP serves as an interface for advertisers to connect to the programmatic advertising ecosystem – and to manage how they purchase advertising media from that ecosystem.
Features provided by DSPs to advertisers include the ability to:
- Track ad performance
- Perform real-time bidding (RTB) on publishers’ available ad inventory
- Define target audiences
- Control frequency capping
- Establish ad campaign budgets
- Access and purchase exclusive ad inventory from partners of the DSP service provider
DSPs allow advertisers to connect to the programmatic advertising ecosystem in several different ways:
- Ad Exchanges: By connecting to an ad exchange via a DSP, advertisers are able to purchase impressions from ad networks and publishers that are connected to the exchange using an SSP. The ad exchange facilitates these programmatic transactions between publishers and advertisers as an “open market” without the involvement of intermediary parties.
- SSPs: DSPs can integrate directly with one or more SSPs, a connection which allows the advertiser to purchase a publisher’s ad inventory without the use of an ad exchange.
- DMPs: Data management platforms provide supplementary third-party audience data to DSPs. With this data, DSPs are able to establish more accurate audience criteria before placing bids on available publisher inventory within the programmatic ecosystem.
In modern adtech, the most confusing thing about ad servers and DSPs is the fact that many DSPs come bundled together with ad servers – making it seem like DSPs perform the same actions of both technologies.
However, this is not the case:
- An ad server does not allow advertisers to connect to the programmatic ecosystem.
- A DSP cannot store ad creative files and serve ads to a website without the functionalities offered by an ad server.
- One of the few capabilities shared by both ad servers and DSPs is their ability to report on ad performance – a feature which is handled based on the selected service provider(s).
Ad servers and DSPs are distinct pieces of technology with their own separate functions and roles to play in purchasing and serving ads, despite many service descriptions which like to group these two pieces of adtech together for the sake of marketing convenience.
In marketing and advertising, SSP stands for “Supply Side Platform” or “Sell Side Platform”.
A supply side platform is a piece of adtech used by publishers.
An SSP serves as an interface for publishers to connect to the programmatic advertising ecosystem – and to manage how they sell their ad inventory through that ecosystem.
Features provided by SSPs to publishers include the ability to:
- Establish rules surrounding which types of advertisers are allowed to bid on ad inventory
- Showcase their available ad inventory to thousands of relevant advertising parties
- Automatically sell their available ad space impressions to highest bidding advertisers
- Create guaranteed deals and other basic contract agreements with specific advertisers
- Configure options such as price floors (minimum buy prices) to optimize website revenue
- Manage a wide variety of other aspects of their ad inventory through a single interface
As an important and distinguishing note about SSPs – while they are highly effective at selling remnant ad inventory (inventory which hasn’t been sold by other means), using an SSP to sell ads is almost always less lucrative than an instance where a publisher establishes a direct deal with their advertising partners.
However, due to the element of convenience offered by automation, the tradeoff in ad inventory value is typically made up for by the amount of time saved by leveraging programmatic advertising to sell large quantities of ad inventory which would otherwise go unsold.
SSPs allow publishers to connect to the programmatic advertising ecosystem in several different ways (but differ slightly in functionality from DSPs):
- Ad Networks: By connecting to ad networks, SSPs are able showcase a publisher’s ad inventory across numerous listings that are browsed by a large number of advertisers. Ad networks don’t sell this inventory programmatically, but rather, increase exposure and the likelihood of publishers being able to manually negotiate deals with advertising partners.
- Ad Exchanges: When connected to an ad exchange, a publisher’s remnant inventory can be auctioned off programmatically, meaning that ad space that would otherwise go unsold is able to generate ad revenue for the publisher automatically, albeit often at a lower price.
- DSPs: SSPs can integrate directly with DSPs used by advertisers, allowing a publisher’s inventory to be sold using a variety of programmatic direct methods.
- DMPs: Similar to DSPs, SSPs can also connect to DMPs to associate third-party audience data with their inventory, improving the likelihood that a publisher’s ad inventory will meet the specifications of an advertiser.
You may experience a sense of déjà vu here if you didn’t skim past the DSP section.
Modern adtech providers often bundle ad servers and SSPs together when describing their services – which can make it seem like they’re the same piece of technology.
However, this is not the case:
- An ad server does not allow advertisers to connect to the programmatic ecosystem.
- An SSP cannot store ad creative files and serve ads to a website without the functionalities offered by an ad server.
- Reporting options are one of the only capabilities shared by both ad servers and SSPs.
Ad servers and SSPs are distinct pieces of technology with their own separate functions and roles to play in selling a publisher’s ad inventory.
While SSPs focus on allowing publishers to conveniently sell their remnant ad space, ad servers are what ultimately allow an advertiser’s ads to be served to a publisher’s website.
In marketing and advertising, DMP stands for “Data Management Platform”.
In the context of advertising, DMPs are used by both publishers and advertisers.
DMPs are repositories of first, second, and third party data – and used to inform and improve both human and automated decision making processes (typically by segmenting audiences).
Despite some DMPs managing all of these data types, most DMPs focus on third-party data.
In the context of digital advertising, DMPs have relied primarily on third-party data in the form of cookies to provide enhanced ad targeting capabilities.
However, with the end of third party cookies on the horizon, the role of DMPs within programmatic advertising will become less influential than it has been in previous years.
As part of the programmatic advertising ecosystem, the role of DMPs has been to improve the functionalities of DSPs and SSPs by providing additional data about web users.
Until third-party cookies are discontinued, DMPs will continue to improve ad relevance, allowing publishers to benefit from increased ad revenue, while allowing advertisers to secure more clicks from people that are interested in their products and services.
This publisher-oriented article highlights different ways in which a DMP can be used to serve more effective ads, while this advertiser-oriented article covers the benefits of a DMP’s enhanced audience targeting and audience reach from an advertiser’s perspective.
While DMPs are forecasted to be less impactful in programmatic advertising, these platforms are actively evolving and adapting to privacy changes, and still offer a range of marketing intelligence functions, which can be explored with this overview on how DMPs work.
The difference between a DMP and its DSP/SSP counterparts is very distinct when compared to other pieces of adtech, which often have overlapping features and functionalities.
While DSPs and SSPs both allow advertisers and publishers (respectively) to connect to and interact within the programmatic advertising ecosystem, DMPs perform no such function.
Instead, DMPs are focused exclusively on storing audience data, and applying that data within programmatic advertising to increase the reach, accuracy, and overall value of advertising, both for publishers and for advertisers.
The easiest way to remember the difference is to understand that DMPs are “data providers”, while DSPs and SSPs are not.
In the context of programmatic advertising, DMPs live as an integrated part of the programmatic ecosystem itself, and serve to fuel the automatic ad transactions that take place with audience data, rather than facilitating the ad transactions themselves.
Before diving deeper, in case you want a more in-depth explanation specifically on the topic of CDPs – The Ultimate Guide to CDPs by HubSpot should be your go-to resource to gain a baseline understanding of CDPs, DMPs, and CRMs.
If you’d like a “second opinion” (and let’s be real, that’s never a bad idea for anything that comes in contact with adtech) Econsultancy has also put together a great piece that covers similar information for advertisers looking to learn more about CDPs vs DMPs vs CRMs.
In this guide, we’ve done our best to tie everything back to how these platforms operate from an advertising perspective.
To start off simple, in marketing and advertising, CDP stands for “Customer Data Platform”.
In the context of advertising, CDPs may be used by both publishers and advertisers.
CDPs are focused on collecting first-party data and PII (personally identifiable information).
Through this first-party information, CDPs are able to construct and manage detailed customer profiles that contain valuable details about a brand’s customer base.
The customer profiles of a CDP are used to track customer engagement metrics such as conversion rates, customer retention, and other customer-oriented marketing functions.
In some cases, CDPs also use third-party data to enhance the detail of customer profiles.
While some functions of CDPs can be used in advertising (for instance, tracking ad clickthrough rates, and creating “mirror” audiences based on existing customer data), CDPs have not traditionally been directly involved in the process of programmatic advertising.
In other words, the first-party data collected by CDPs has historically been applicable when making strategic advertising decisions, but has had a limited application within advertising operations (ad ops) themselves.
However, with the end of third party cookies, it’s not unlikely that companies will begin to rely more on integrating CDP data into their newly created first-party data oriented advertising strategies.
If CDPs and DMPs seem to have frustratingly similar definitions – you’re not alone in experiencing the feeling.
In an article by Digiday, author Yuyu Cheb writes “There’s no industry wide definition for CDPs” and “…the line is blurring as major DMPs are evolving”.
Since the article’s publication in late 2017, the service offerings of DMPs have been forced to overlap in many ways with those of CDPs in order to remain relevant with all of the changes taking place in the shifting global user-privacy landscape.
Despite the confusion, there’s a simple way to separate and define these terms at a broad level – and it boils down to the semantics between the terms “advertising” and “marketing”.
DMPs are focused on “advertising”, meaning that they’re focused on putting messages in front of “unknown” audiences that may have extremely limited data available to define them.
CDPs are focused on “marketing”, meaning that they’re focused on putting messages in front of “known” audiences that have customer profiles and a history of brand engagement.
They’re two different systems that compliment one another – DMPs focus primarily on gathering third-party data from an incredibly broad audience, while CDPs focus primarily on gathering first-party data from a targeted collective of confirmed customers.
If that definition still leaves room for confusion, or you find yourself questioning whether a certain platform is a DMP, a CDP, or both, that’s because there’s still a lot of confusion within the adtech industry itself.
For further reading on the difference between these two platforms, the article “What Is This Thing We Call a CDP?” written by Martin Kihn, Research Vice President at Gartner, provides more insight into what’s referred to by him as “The Great CDP vs DMP Debate”.
CRM stands for Customer Relationship Management (platform).
It’s not uncommon to have confusion arise when discussing CDPs and CRMs in the same conversation.
Fortunately, the difference between CDPs and CRMs is a bit more clearly defined than the blurry boundaries of CDPs and DMPs.
As the name implies, a CRM is built to focus on managing and facilitating the actionable steps within the process of building relationships with customers.
To provide a clear contextual difference between CDPs and CRMs – think of it this way:
CRMS are focused on “sales”, meaning that these platforms are focused on providing sales teams with the tools and information they need to build relationships with leads.
CDPs are focused on “marketing”, meaning that they collect highly detailed customer engagement data that helps marketing teams to make informed strategic decisions.
Similar to CDPs, CRMs also store data and build profiles on users as information is gathered about them. This includes first, second, and third party data from different sources.
For example, a CRM may track and store information about which page a user first discovered a website on, and the series of pages that user visited prior to signing up for a newsletter or online service, prior to initiating them into a sales team’s pipeline.
The CRM would also store the individual’s PII data, including their email address, their first and last name, and may even include some second and third-party data manually added to the individual’s profile by a salesperson during or after a sales call.
To provide a fictional contextual analogy – imagine the process of converting a lead into a customer as a relay race – taking place on the CRM “sales relay race track”.
The CRM “sales relay race track” is focused on the process of passing the “baton” between members of a sales team, ensuring that all of the information about a client is consistently known by each member of the sales team during each “baton pass” until the sale is closed.
As soon as a lead becomes a customer, the sales “relay race” ends.
Once the sales process is over – imagine the process of tracking customer actions as individuals entering a CDP “marketing gym facility”.
Here’s where it gets interesting with some overlap. Customers who came from the CRM sales process are wearing bright red jerseys that say “I came from the CRM race”, which means that the CRM may continue to track some or all of the actions taken by these individuals.
However, there are other anonymous visitors (users who didn’t go through the CRM process) entering the gym facility who are also performing actions and interacting with various gym equipment and things like vending machines – all referred to as “touchpoints”.
The CDP doesn’t care whether or not someone is wearing a bright red jersey. All the CDP cares about is tracking how often any visitor is coming in contact with certain touchpoints – and it tracks this information anonymously, regardless of whether or not a customer’s identity is known (unique identifiers can still sometimes be used to track which anonymous visitors are taking which actions, building an “anonymous” customer profile).
In other words, the CDP is only keeping track of how many times visitors are using each weight set, how many volleys have taken place in a game of volleyball, how many basketballs have been dunked, and how many sales are taking place at the vending machines.
If an anonymous visitor ever converts to a customer, the history of their interactions with the gym equipment (aka, the software touchpoints they interacted with) will be transferred to their first-party customer data profile, assuming that the CDP offers an anonymous data tracking feature.
One additional distinction that can be drawn is that CRMs do not collect “offline data”, while CDPs can integrate, connect with, and store information from offline sources, such as mobile apps or in-store purchases (as just a few examples).
CDPs and CRMs are systems that compliment one another, rather than compete over the same functionalities.
A more detailed breakdown about how CDPs and CRMs work, and the distinction between the two, is available here.
A lot has been covered in this guide.
There are enough three-letter acronyms flying around to cook up an overflowing pot of alphabet soup.
In case there’s still some confusion surrounding any of these terms, it can be helpful to think of them in broad categories.
Serves ads and stores/manages ad creative media elements:
Connects advertisers and publishers to programmatic advertising:
Data platforms that assist in ad strategy and ad targeting:
DMPs, CDPs, CRMs
As a bonus, here’s a quick “speed round” of each term to wrap things up:
- Ad Servers: Used to store and serve ads.
- DSPs: Used by advertisers to connect to the programmatic advertising ecosystem (to buy ad inventory automatically).
- SSPs: Used by publishers to connect to the programmatic advertising ecosystem (to sell ad inventory automatically).
- DMPs: Focused on storing and leveraging third-party data in programmatic advertising.
- CDPs: Focused on storing and leveraging first-party customer data, mostly for marketing.
- CRMs: Focused on storing and leveraging known lead data, mostly for sales pipelines.
And that’s a wrap.
Now you’re all ready to rattle off these hot adtech acronyms to your team – and maybe even get mistaken for an up and coming beatboxer while you’re at it (if you decide to take up beatboxing instead of being involved in adtech, be sure to invite us to your next performance!).
If you plan to stick with adtech for the long haul, there’s always more to learn about digital advertising.
We’d love to share a conversation with you. Ask us a question today!