The Real Reason ‘Mobile DMP’s’ Make No Sense
There’s been a lot of talk recently about mobile data platforms. Coming out of Cannes, it felt like the two things pretty much everyone was focused on were mobile video and mobile data. Recently, there was a thought provoking article in AdExchanger about mobile data management platforms, or mDMPs. This was a follow up to Tom Chavez’s post from a year earlier about mobile DMPs being dead at design time. I whole-heartedly believe that DMPs provide marketers and publishers with critical functionality, and I couldn’t agree more with Tom Chavez that a “mobile DMP” isn’t a good idea. I disagree with why, however.
Omnichannel marketers & publishers need an omnichannel solution. To start, let’s review what a DMP should do. A DMP is a solution that can aggregate data across multiple consumer touch points to form a holistic view of the consumer. To Tom’s point, the very essence of a DMP is to break down the data silos, and only looking at mobile gives a limited view of the consumer which doesn’t make sense for omnichannel marketers or publishers.
But if a company is mobile only or predominantly mobile — such as Foursquare, Spotify, or Hotel Tonight — wouldn’t a mobile DMP make sense? For these types of mobile-led companies a mobile DMP isn’t a bad idea because they need an omnichannel solution, it’s a bad idea because they don’t need a DMP the way it was imagined for the web. These types of companies need something different, they need a solution that addresses the unique data needs of a mobile company.
An omnichannel DMP is essentially a hub and spoke solution. The central database is the hub, and the data collection done at each channel (web, search, email, social, mobile, etc.) are the spokes. However it’s important to separate the global challenges (platform level) and local challenges (each channel) to build a successful understanding. While I can’t speak to all of the challenges within each of the local channels, I can speak to mobile and specifically mobile apps.
In mobile, where most time is spent in native apps, there are a number of business and technical challenges that are unique from web. The biggest difference centers on the data and the way that data gets operationalized. First, apps can take advantage of device level functionality which generates data types that websites never deal with. As such, apps use tools that their web counterparts don’t require such as crash reporting and push notifications, just to name a couple. If a legacy web DMP isn’t designed to capture the unique data types that are core to app businesses then the solution is incomplete.
The second set of challenges center on operationalizing mobile data. Adding new vendors is incredibly complex and time consuming, and it’s done by embedding software developer kits (SDKs) since web technologies like pixels and cookies don’t work. Every vendor SDK that is added to capture an app’s first-party data not only bloats the app, but jeopardizes user experience and slows down the engineering team. Since apps are more like shipped software than they are websites, the release cycles and time to onboard new partnerships can be painstakingly slow which creates even more complexity.
To successfully solve data challenges at the app level, the mobile data solution must also reduce/remove friction from the partnership process. Adding an additional SDK for a web DMP, which wasn’t built to capture important mobile data types or integrated with the ecosystem of mobile app services, is only adding to the overhead, not solving it. Thus trying to implement a global solution while compounding local challenges won’t work either.
The last reason web DMPs miss the mark on app data challenges is because web DMPs have become very campaign centric tools, while app usage is more purpose driven than campaign driven. The core use case for web DMPs is to drive audience analytics across media to increase efficiency for marketers and yield for publishers. They do a great job at this, and again this is very important but these aren’t the challenges that drive app businesses. Apps provide an important canvas to make it incredibly easy for consumers to transact, and most solve for the quality of the user experience within the app. The on-demand economy is proof of this, as is the inability for any CPG brand to create a meaningful app.
So to sum it up, mobile DMPs are not a good idea. But neither are web DMPs that try to solve app challenges with an incomplete offering.