Today’s brands aim to understand customers in every way, starting with collecting and analyzing purchase data like customer names and shipping addresses. Along with machine learning and data-mining techniques that help to understand intention, this data sets the foundation for these valuable customer profiles.
However, user profiles built this way have limits. If the key to understanding consumers is acknowledging the complexity and individuality of every customer journey, shouldn’t user profiles go beyond an aggregate, batch-and-blast approach?
In considering how to get the most of marketing data, let’s walk through Turn’s system of synchronizing information around the customer journey, culminating in our massive store of more than 3.8 billion anonymous user IDs, the Turn ID.
The ‘Big Event’ – and How Turn ID Can Help
Let’s say a soccer mom is researching hybrid cars. She peruses brands, leading her to a specific model with select features. Each step leads to the big event, the purchase, and knowing where she is in her journey can help predict her next move and reach her at the right moment.
That’s where Turn’s tech offering comes in. Built around a system called the Turn ID Store, the technology holds anonymized, event-level data about user interactions with advertisers’ online content, along with first-party or third-party data attributes. All user attributes can be accessed in real time. We collect data from sources without personally identifiable information (PII), and we don’t track data like email addresses or Social Security numbers. Instead, Turn ID is a proxy representing a device, and all online events are collected based off that ID.
Because data is tracked in real time, there’s no latency associated with targeting off that data. We target at a highly granular level, not simply in aggregate. We can track the soccer mom as she looks for hybrid cars with high safety ratings and good cargo space, and eventually, when she makes a purchase. Marketers’ options include building complex segments to specifically target users at different points on the conversion path or running sophisticated experiments to sequence creative effectively.
Fundamentally, we believe marketers will need to perform some sort of sequential analytics where the focus is on the customer journey rather than the touch-points themselves. Maintaining an event level data-warehouse is a prerequisite for enabling such sequential analytics. This is how we’re building a “what works machine” that helps marketers find out what works, what doesn’t and why.
Up to the Challenge
There are two ways a platform needs to access event-level user data. In the demand side platform (DSP), the event-level data needs to be accessed for making real-time decisions, requiring millisecond-level access to individual profiles. In the data management platform (DMP), a large number of user profiles have to be scanned to find meaningful patterns. However, there isn’t one system that can satisfy both distinctive access patterns with similar efficiency, so a platform needs to build two separate user profile systems specifically optimized for each use case.
At Turn, we built two separate data stores to best suit each use case. Our real-time system collects the user events with very low latency and makes them actionable almost instantaneously. The data gets replicated into our offline data warehouse for batch analytics and extensive data-mining. This is where Turn technology shines compared to its competitors who may have either a DSP or a DMP, but not both.
There are challenges, of course. We stay on our toes to scale for ever-increasing amounts of data, and we have to consider where next-generation hardware and systems are headed and how to manage costs for our clients.
Cross-device tracking, or reaching people wherever they are, is now a key tactic for marketers. Our hybrid car buyer might use her work laptop to research on her lunch break, continue to browse on her phone on her evening commute, and use a tablet to show her husband options before going to bed. The activities have to be tied together for accurate, effective targeting, and the device graph needs to have two separate storage systems, just like event-level data.
Processing event-level data with Turn ID helps our customers use data to create compelling stories. Over time, systems leveraging event-level data for sequence analysis will provide better ROI to marketers, and hopefully, serve a more relevant, effective ad experience to the soccer mom looking for the right hybrid.
Read more on our natively integrated demand side platform (DSP), data management platform (DMP) and analytics solution.
Read NextAll Blog Posts
Why You Need a Data Strategy
While brands have now heard loud and clear that they need to start using data in their marketing, many organizations often struggle to get off the ground because they don’t have a real data strategy.
November 9, 2016
Top Marketers’ Advice on Getting Started with Data
Where do you start as you’re trying to transform your marketing organization into one driven by data? Matt Westover, vp-product marketing, Turn, led a session with top brand and agency marketers on the topic at the Gartner Digital Marketing Conference in San Diego.
May 12, 2017
Why Programmatic Sticks with Marketers
Programmatic is moving from an emphasis on technology as a way to buy cheap, remnant media to a driver of invaluable customer insights that can impact everything from targeting to package design.
January 9, 2017