Elsevier Optimizes Oracle B2C Service and Creates a Data Warehouse to Serve the Entire Organization

Ever wonder how a world-leading company keeps track of all customer interactions at each stage of the journey? How do multiple systems connect to provide a smooth, integrated experience? Here’s an exclusive, behind the scenes snapshot. Find out about the Elsevier Customer Service Solutions team’s data warehousing project from two of the software engineering professionals driving it. 

Marc Grant is a Software Engineering Manager at Elsevier whose team looks after Oracle B2C Service. Dom Smith is a Software Engineer at Elsevier.  Their team supports the Oracle B2C Service users (e.g. agents, managers), plus the data and analytics tools.

This is their team’s most complex data warehouse project to date, albeit not their first. Marc and Dom wanted to improve on past experiences, so read on and learn how they approached this ambitious project and what benefits they’re seeing. 

Marc Grant

Marc Grant, Software Engineering Manager, Elsevier

The mandate was to break down the silos!

Elsevier use multiple systems to gather customer data, including Oracle B2C Service (part of Oracle Cloud CX). Different stakeholders in the company needed different data sets and views. Before, when stakeholders needed data, they went to the different teams who owned the systems. Once they received the data, they had to sync it with data from other sources and work with the resulting spreadsheets. This required a lot of manual labor and effort from various teams and results were prone to human errors and delays.

Elsevier had years of valuable data in Oracle B2C Service, but they struggled to run historical reports due to the huge data volumes and the complexity of their implementation (e.g. many custom objects). Also, the underlying problem behind these performance issues was that Elsevier relied on Oracle B2C Service to be part of the data warehouse solution, not optimizing it as the transactional system, like it’s designed to be. To top it off, they needed to comply with policies like GDPR while still keeping historical data to support the massive keenness around the globe for people to do long-term trend reporting.

In summary, Elsevier’s technical systems were building up—not breaking down—those organizational silos and were undermining the different Elsevier teams’ ability to make clear data-driven decisions to reach business goals.

The process brought systems together to deliver actionable data 

Marc, Dom, and team embarked on a year-long project to fix and improve how Elsevier’s data from different systems could come together and support the various stakeholders of the organization. 

The project goals were to:

  • Establish a single source of trusted Customer Service data.
  • Improve Elsevier’s ability to access and report on data.
  • Contribute to Elsevier’s business goals through product improvements, reduced inquiries, and increase CSAT in a reliable, streamlined manner.
  • Enable their data analysts and scientists with complete data sets to support their distinct use cases. 
  • Gain more and better insights from Oracle B2C Service by optimizing it as a transactional system, not a data warehouse or reporting system.
  • Comply with GDPR while anonymizing and storing historical data.

Dom Smith, Elsevier

Dom Smith, Software Engineer, Elsevier

Dom said the overarching goal was to, “increase trust with senior leadership teams and internal customers. It was about providing them with the answers. We have the data and the solution, so you can make the decisions you need based on improved data.” 

The solution – design a system to deliver trusted data across the business

Before coming to a decision, Elsevier engaged its reporting teams from around the world and asked them what data they reported on, and to what end. For example, “How are we going to improve NPS for a product with this data?”

They gathered and analyzed the reporting requirements for each team over four cycles to produce the final version. This iterative process was crucial because every global team had different KPIs. One team counted incidents that came in for 24 hours while another looked at a 48-hour window. Elsevier realized it was treating customers differently.

This thorough process also informed compliance decisions. Just because someone had access to certain information didn’t mean they needed to or should have access to it. 

Their diligence paid off. They trimmed the Oracle B2C Service data set by 30% from the original requirements. This gave a big performance advantage and uses less storage.

The project also kicked off very important discussions. Looking back on their journey, everyone involved in the Elsevier data warehousing project committed to working outside their comfort zone. The Elsevier team moved into designing and delivering their solution with a focus on providing a holistic view of the customer, their organization, and their history with Elsevier.  

Elsevier use Oracle B2C Service and the API to run custom reports for extracting the support data, and then feed that through Talend to the different reporting audiences—data analyst and data scientists.

Talend is the software that goes to both Oracle B2C Service and other applications, extracts the data from those sources, and does all the processing before passing the data to these other pipelines. 

Previously, there were separate support groups for Oracle B2C Service, Talend, and others that would do separate bits of the work. Marc and Dom were keen to have the skill set within a single team deal with those requests. “We were very conscious of trying to avoid a siloed support structure for the data warehouse and widened our skills accordingly,” said Dom. “Now, they are responsible for all the technology until they hook that data up to the reporting server. From this point forward, reporting teams around the globe and analysts build the data sources in the way they wish and come up with dashboards and reports for product managers, marketing managers, and others.”

“At the end of the project, we were able to publish the Customer Service data sets to the Elsevier data lake, which other systems can publish into as well: marketing, sales, customer service, or product. These stakeholders can come with their data shopping list, put what they want into their shopping trolley, and report on it. We’ve seen things really move from having that siloed, team-by-team reporting approach to the joined-up trend reporting across the whole of Elsevier,” summarized Marc.

By having a single source of trusted data, Elsevier is starting to break down technical and business silos and give everyone the same view. 

Trusted data empowers proactive support

While this type of work is never quite complete and is always an ongoing evolution, Elsevier is reaping significant benefits with their new technical system and updated processes.

  • A new customer 360 dashboard: This provides Elsevier with a clearer view of how customers engage and signals if they should be proactively and intentionally changing how they interact and serve different customers. 
  • Identifying duplicate issues: Elsevier can now quickly identify people who experience the same problems and put a process around fixing those issues.
  • Look ahead to predictive analysis and machine learning: Marc explained, “At the moment, an agent takes a call, they assess what category it is, and they fill that in. We have our agents take that step for reporting, but if we can leverage predictive analytics based on our historical dataset to dynamically generate those categories and take that onus away from the agents, then that would be fantastic!”

Customer 360 dashboard. Blending data from multiple Customer Service, Sales and analytics systems.

Customer 360 dashboard. Blending data from multiple Customer Service, Sales and analytics systems.

Of course, there is still some fine-tuning to do and they have other datasets to bring into the fold, but they’re very conscious to avoid a siloed system or support structure for their data warehouse.

Advice for others starting a similar project

For other organizations launching into a similar initiative, their biggest advice is to start by working across other departments and teams to understand the data that people need.

Dom explained, “You really need to understand what data you’re bringing in and why you need it. Don’t just take a list of fields and put it in the database without understanding what they’re going to do with it on the other end. Expect your job, team and processes to evolve. View this as a growth opportunity. Get used to being outside of your comfort zone and set that same expectation with others on the project.”

If Elsevier’s story inspired you to make the most of your company’s data, find out how your organization can benefit from implementing state-of-the-art customer service with Oracle.


Leave a Reply

Your email address will not be published. Required fields are marked *