DMPs vs. DSPs – why you need the former
Published: November 22, 2016
Author: Feliks Malts
Recently, we ran a post on common misconceptions regarding DMPs (data management platforms), which have become more and more important in digital marketing given the proliferation and growing importance of data.
Two questions we often come across are, “What is the difference between a DMP and a DSP?” and, “Can an advanced DSP do everything a DMP can do?” To address these questions, let’s start with what a DMP is, what a DSP (Demand-Side Platform) is, and what an advanced DSP can do.
What’s a DMP?
Traditionally, a DMP has allowed you to centralize first-party data that you are collecting on users who are visiting your site and getting exposure to your brand in media. More recently, DMPs have been incorporating data made available through third-party providers, allowing you to get a lot more insight into your users without having to ask for more information.
The cycle starts with defining requirements, architecting the system, and managing data collection, and moves through implementation from a technical perspective and leveraging the platform to deliver insights. DMPs allow marketers to identify our customers’ prospects, users, abandoners, and most importantly converters. The main players include Lotame, Adobe Audience Manager, BlueKai (which is now Oracle DMP), and some smaller companies like [x+1] and xaxis. They give us incredible insight into data sources and allow marketers to react to signals in real time.
In essence, a DMP will allow you to bring all your first-party data together and layer some really, really powerful data on top of it to so you can get a more complete understanding of who your users are – beyond the information they are giving you.
What’s a DSP?
A DSP is a platform that helps automate media buying, allowing you to purchase in real-time. A DSP automates the programmatic buying process, reducing the back-and-forth and helping the buyer decide which options are most worth bidding on. An advanced DSP will have partnerships with some third-party data providers, so from the perspective of making data available to you for retargeting, a solid DSP will do the trick. However, compared to a DMP, a DSP won’t offer any insight that goes into that data and who you should actually be targeting.
The piece that’s really missing in an advanced DSP is the analytics component. With a DMP, you can load your behavioral and user data into a platform in a really granular way with granular signals. This allows you to get much deeper insight into who your customers are and what they look like before you begin targeting. If we had to think about being successful with a DMP, we’d think about it in two phases. Where targeting is phase two, DSPs are missing phase one, which in terms of strategy is even more instrumental because of how much it is needed to inform targeting. So while advanced DSPs are great tools, they are not currently able or meant to fulfill the same needs a DMP can.
Why do companies need a DMP?
There are a couple of reasons. The way we like to think about DMPs at 3Q is two-fold:
- A DMP can help you gain insight into who your users and ultimately your customers are. If you’re really in the dark and you want to replicate the traditional commerce experience where someone walks into your retail store and you learn about them through conversation, a DMP is a way to do it. As users visit your site you can start to understand who they are from a demographic, psychographic, financial, professional, and in-market perspective. Leveraging that data, you can almost replicate what you’d able to learn from a conversation with that user.
- The second part is taking that information and intelligently, strategically, and efficiently thinking about how you invest your budgets in media, using everything you learn to go talk to the people who are buying from you or look like the people who are buying from you so more of them to convert, buy from your brand or become part of your brand family, buy your products, and so on. Another really innovative way of thinking about DMPs that has yet to become that common is leveraging the platform and its data to not only understand who’s actually buying from you, but also who’s not buying from you, so you can alter your strategy or exclude them from your media buys to be even more efficient.
Obviously, we’re bullish on DMPs at 3Q. If you haven’t read the misconceptions post referenced above, I’d recommend it, since there’s a lot of bad or outdated information floating around. And as always, if you have any questions, leave a comment!