The Foundation of ALM’s Big Data Initiative

ALM's Big Data initiative started more than four years ago -- before the term Big Data was even coined. It began with the editorial team setting up a content taxonomy. The goal at that time was to provide users better search capabilities and allow advertisers to target users broadly based on a content subject matter (i.e., display a particular ad for all content that is labeled e-discovery). This initial project has since become one of the primary foundations upon which ALM built out its Big Data Initiative. Given limited funding and resources, we took a pragmatic approach to Big Data building out our system with a modular investment approach rather than an all-in multi-million dollar upfront investment. Following are the 5 foundational elements of our build-out. (Read about ALM's use of behavioral data to boost marketing campaign effectiveness.) Developing a Robust Consistent Content Taxonomy: Leveraging the initial taxonomy work, we expanded ALM's taxonomy to include not only parent-child relationships but also grandparent-parent relationships. (Parent-child and grandparent-parent-child relationships are characteristics of a hierarchical model used to structure and organize data.) Since then we have been applying this taxonomy across the entire ALM organization -- both publications and non-publication business units. An important part of this process was to bring our sales and marketing team into the conversation, since the resulting taxonomy would serve all business units. In the near future ALM will complement editorial taxonomy efforts with automated solutions that will provide even more detail and consistency. Building Out and Leveraging Registration and Subscription Platforms: While not an absolute necessity in a behavioral data driven world, at ALM we've found that having a registration platform provides an additional level of accuracy and opportunity. More specifically, the registration and subscription data provides additional demographic information and provides a unique point of reference across platforms (i.e., mobile, tablet and desktop). Investing in a Data Warehouse: Next, we needed to invest in a data warehouse that could store not only our audience database, but all ALM data including transactional data, reading habits, and profile information. After significant deliberation, we chose MarkLogic -- a NoSQL database as the core to our MDM (master data management platform). The main benefits of MarkLogic are that it does not require a data schema (thereby reducing our time market), it can manage both structured and unstructured data, and it makes it easier for data analysts to work with complex data sets. Licensing Reporting tools: Having all of the information piping into a single repository, we next needed to be able to gain some insights into the information. To do this we licensed Tableau, a data reporting tool. With Tableau we are able to build out interactive data visualizations that make data interpretation and understanding much easier for ALM business units that may not be as data savvy. These insights can have direct impact on marketing campaigns, sales efforts and content strategy. Adding Tracking and Targeting capabilities: To begin to monetize the data, we determined that we needed additional tracking and targeting capabilities. Of importance, was an ability to do more than just track user behaviors, but to also: (a) create affinity groups based on the types of content users are viewing and how much they are consuming, (b) help confirm who users say they are by modeling what we know specific types of users typically read, and (c) help derive who we think a user may be if we don't know who they are -- especially helpful in adding value to the anonymous audiences (non-registered or subscribed). To do this we rolled out a tag management solution that in-turn allows ALM to leverage any web driven data element from a behavioral or content delivery standpoint in one platform. In the upcoming months we expect to make additional investments in predictive analytics tools and resources that will allow us to take targeting and lead generation to an even higher level of efficiency and accuracy. This should not only continue to drive sales and revenue but drive them with much higher margins. We see this as the next-generational opportunity for Big Data. Jeffrey Litvack is the President of ALM's Intelligence & Advisory division and Chief Digital Officer. Robert Schultz is Head of ALM's Business Intelligence department.

Emerging Technologies

The Foundation of ALM’s Big Data Initiative

Jeffrey Litvack

About

Subscriptions

Events

Publications

WORK SMARTER - LEARN, GROW AND BE INSPIRED.

Subscribe today!

Jeffrey Litvack