Skip to Content
Author's profile photo Guido Wagner

SAP Wins UX Design Award 2017 with Data Encounter

The winners have been chosen and SAP is very pleased that the prototype “Data Encounter” has been honored with this year’s UX Design Award | Public Choice! The UX Design Awards are an annual competition hosted by the International Design Center Berlin. The Awards recognize excellent experience qualities in products, services, environments and future-oriented concepts. Particular focus is given to how human-centered design and new technologies can make a positive impact in peoples’ lives.

Data Encounter, which is currently a solution prototype and not available on the SAP pricelist or integrated into currently available products, supports users with the exploration of large data pools and the identification of formerly hidden structures. It was designed for a natural, intuitive interaction between users and the system.

One of the biggest challenges for business users is understanding what influences his or her business and analyzing how the situation is evolving. Not only data scientists, but many people throughout the organization like marketing experts or sales analysts need tools that support their intuition and offer a natural way to look at the available data from various points of view. Data Encounter allows users to explore business-relevant data, see the development of trends and visually discover similarities and anomalies, thus gaining new and often unexpected insights.


Data Encounter Video

Challenges of Big Data for marketing

In conversations with users working in marketing departments, the team found out that it is very difficult for them to understand customers’ buying behavior. Insight into the journeys of thousands of customers to see where the similarities are or what triggers them to purchase a product would be a huge help to sell more and waste less.

After numerous interviews with data scientists and other potential users and secondary research, including tangible embedded experiences, the team created several low-fidelity prototypes and validated them with people working in the targeted roles. In this way, they could understand what matters to the users and how they would use the tool.

The team’s research indicated that users want to explore large volumes of data relevant to their work in multiple dimensions. Data Encounter can display data in up to ten dimensions based on the business context. By visualizing the data in 3D, Data Encounter can show many more data analysis dimension than would be possible in a 2D space. The use case of the demo which the team submitted is based on marketing data from a sports retailer and supports the discovery of insights for a marketing campaign. Using anonymized SAP Hybris marketing data for the purpose of the demo, the team was able to provide thousands of marketing contacts (individual consumers) and their behavior (interactions on various channels). Product categories were represented in a three-dimensional space that the user can freely explore along multiple axes. These axes are customer interests, company products, marketing phases, probability that the contact will move to the next marketing phase, the age of the contacts, and the time since the contact last interacted on one of the channels with the company.

This marketing scenario is one area in which Data Encounter can support business users. Further business areas and use cases are for example:

  • human resources (matching employee or job applicant skills to business needs)
  • supply chain management (tracing reliability of product shipments)
  • finance (reviewing the amount and duration of sales transactions)
  • telecommunication (examining the profitability of subscriptions)

For the product definition and especially the product design, the team followed the Design Thinking methodology. Several potential users were interviewed to find out about their needs. Users indicated that they want to:

  • Recognize patterns or structures within the data, for example see data clusters that emerge from a certain point of view
  • Learn from historical changes in data and understand better the patterns evolving out of real-time data
  • Interactions that support people’s ability to identify patterns
  • Features, such as the time slider, that allow users to “travel” in time and identify trends
  • Choose the level of detail that is visualized, for example to learn if clusters appear on a product or product category level
  • Highlight areas in the data to easily focus on particular interests (additional data filters can be added)
  • Get all the details of a single data point if needed, like the contact address
  • Generate classical two-dimensional charts out of the three-dimensional representation to easily share in the classical form and how to their boss or others in the company
  • Use whichever navigation device best fits their personal needs, such as a 3D mouse
  • Get a tactile feeling when they use the tool, as if they can almost “touch” the data to better examine and understand it

The Data Encounter solution

To meet the needs that users expressed, Data Encounter represents large data pools in a three-dimensional space in which users can explore live and historical data (see Figure 1). Since humans naturally look for changes in patterns to identify anomalies or emerging trends, animating the data movement over time makes the interface particularly intuitive. Users can observe the movement of the data either in real time or by navigating back and forth between the past and present. The user can rotate and zoom his or her view on the data by using a standard mouse, a 3D mouse or with his or her fingertips on a touchscreen.

Figure 1: Initial Data Encounter screen with the marketing scenario. The points represent contacts’ interactions with the sports company.

Figure 2: Marketing data structured in the business context of interests, products and marketing phase

Data points are positioned in the data space along semantically related characteristics and dimensions. In this marketing scenario, these data points are the last channel interactions of each contact, such as a visit to the company web shop or to the physical store. The data space dimensions are shown in Figure 2.

The time dimension is very important to learn about trends and is used in two directions: Interactions of contacts who have not engaged with any channel and thus have no recent interactions are visualized with a movement upwards, away from the ground disk where the most recent interactions appear. Interactions of those customers who moved to another marketing phase are shown as a movement closest to the base level of the disk, along the marketing phases. This allows the user to directly distinguish “disengaged” contacts (who are literally drifting up and away) from “engaged” customers (who advanced to another marketing phase later). Figure 3 shows an example, including a timeline for the “disengaged” contacts, which indicates how long ago the last interaction happened.

Figure 3: Trends can be observed for „engaged” and “disengaged” contacts

The user can not only define any point of view and time, but can also highlight a certain data area in either of two ways: First, the user can click on segments of the disk to “spotlight” the data points above the segment. Second, the user can bracket a certain time frame on the time slider.

The system colors the data points within this time frame in blue. The user can highlight another time frame to support the exploration of the efficiency of a marketing campaign: When the user clicks on the campaign label in the time slider, the system colors the interactions (behavior) of the campaign target group in orange (see Figure 4).

As soon as the user has identified a data area of interest, one click allows the user to call up related statistics and further details, such as the distribution of gender groups, the most active age group, disengaged vs. engaged contacts and the most used channels. Figure 5 shows an example from which the user learns, for instance, that most people showed interest by using social media as a channel to ask for additional information about a product.

Figure 5 (left side): Statistics of a certain data area

Figure 6 (right side): Further statistics reveal additional details

More detailed statistics are available with an additional click. In the example in Figure 6 , the user learns that contacts who moved forward to the next marketing phase are generally under 35 years of age, whereas older contacts mainly showed interest but did not buy a product. Looking at the overall picture (see Figure 7), the user sees that these people are also interested in snowboarding. This reveals opportunities for further cross-marketing campaigns.

Customer and user feedback

Validation sessions with users and demonstrations at trade fairs clearly indicated that a tool and a user experience like Data Encounter would be enthusiastically received and needed by people who want to learn and gain insights from the wealth of data that their companies are now able to collect. The Data Encounter team is excited about integrating this tool into future SAP products so that customers can dig the “new gold” out of their Big Data.

Assigned Tags

      Be the first to leave a comment
      You must be Logged on to comment or reply to a post.