Ben Gaines is a Senior Product Manager for Adobe Analytics. In this role, he works closely with Adobe customers to understand their needs, provides input on product strategy and roadmap, and manages the planning and design of new features.
Ben is also a frequent contributor to the Adobe blog, with great articles on tips, tricks and updates on Adobe Analytics.
While at the Adobe Summit 2016 in London, I had the opportunity to interview him on the latest features of Adobe Analytics. I also interviewed him in 2015.
Nicolas Malo: In the Adobe Analytics Spring Release 2016, a new metric called "People" has been launched. Could you please explain how this metric is calculated and what is required to activate it?
Ben Gaines: The people metric is based on the Adobe Marketing Cloud Cross-device Co-op. The idea is to enable Adobe customers to share cross-device information based on username identification on a voluntary basis. If you share your device graphes, you will be able to get device graphes from other members of the Co-op.
Co-op members will give Adobe access to cryptographically hashed login IDs and HTTP header data, which fully hides a consumer’s identity. Adobe processes this data to create groups of devices (“device clusters”) used by an unknown person or household. Adobe will then surface these groups of devices through its digital marketing solutions, so Co-op members can measure, segment, target and advertise directly to individuals across all of their devices.
As we are able to idenfity that the same users are logging with different devices, then this metric is more reliable than the unique visitor one. It enables calculation that are people-based, for instance revenue per customer, customer conversion rate, etc... In the end, you will be able to market to people instead of cookies and devices.
No Personally Identifiable Information (PII) will be collected and shared. Adobe is working diligently to launch the device co-op and the people metric in North America first. Other parts of the world will be coming next.
Nicolas Malo: Segment IQ was also announced in the Adobe Analytics Spring Release 2016, but I haven't seen a lot written on this topic. Could you please explain how this new feature work?
Ben Gaines: One of the problems that marketers have is that they need know how they want to define their segments. Once they've built their segment and apply them to the data, there is not a good way for them to figure out what's missing from their data. In this initial phase, Segment IQ is a perfect way to solve "the don't know what you don't know" issue.
Segment IQ enables the comparison between segments. Adobe Analytics will look at the audiences of these segments and determine how they are similar and different. This will not be based at the segment rule level but at the visitor level in those segments. It will tell you what these visitors have in common and what sets are different. Segment IQ will be available beginning of this summer for all owners of Adobe Analytics.
Nicolas Malo: During Brad Rencher's keynote yesterday, the Virtual Analyst was announced as a new feature of Adobe Analytics. When will this be available and how will this work?
Ben Gaines: Our engineering team is actively working on it. Virtual analyst takes the power of anomaly detection and puts that power into your inbox or mobile. It's all about making sure that analysts and their colleagues have the most update view of their business.
Nicolas Malo: Regarding Predictive Analytics, what is already available today for Adobe Analytics users?
Ben Gaines: The first thing we did in this area was Anomaly Detection which predicts numbers for your metrics. Contribution Analysis tells you why you were above or below the expected range, so that you take some action. Segment IQ is predictive in the sense it predicts by looking at the differences and similarities between different segments. It predicts changes that you can make in your segmentation strategy. In Adobe Analytics Premium, we have a number of additional predictive workflows, such as Propensity Scoring which allows you to group your customers and prospects based on how likely they are to perform future actions.
Nicolas Malo: In my list for the Adobe Santa Claus, I still have cross-device and attribution modeling. Do you have any update on this?
Ben Gaines: Yes, I remember from last year's interview that these two ones are hot topics for you! Cross-device and attribution modeling are complex topics if you want to make them relevant on a business standpoint. We will not push these features to our consumers until we are confident with the quality and reliability of what we will deliver. Having said that, I can guarantee you that our engineering team is working actively on it.
Regarding cross-device measurement, the Co-op is a huge step towards visitor stitching. If the Co-op identifies visitors across devices, we will allow marketers to visualize customer journey across devices. Our goal is to provide the best and most complete visitor stitching solution in the world. Unfortunately, I do not have a specific ETA to share at this time.
We are also actively developing the attribution modeling feature, which will be based on machine learning. It will measure the incremental impact of each marketing channel towards success. We are hoping to have a first phase available in the next couple of major releases. For more information on attribution modeling, I encourage to watch John Bates' video from the last Adobe Summit in Las Vegas: https://www.youtube.