Workday announced Tuesday that the Workday Big Data Analytics offering debuting this week is ready for all the data scale and variety customers can throw at it. But the company said it expects that most of its customers will want to start with more pressing, smaller-scale analyses that make use of internal data.
Workday also announced new endowment accounting functionality and a planned major new Workday Student application for the higher-education market, where the company competes head-on with PeopleSoft. But the big news of the day was Workday Big Data Analytics, a new component of Workday 20, the latest release of the company's cloud-based human capital management (HCM) and financials applications.
Workday announced plans for Big Data Analytics late last year, and, as promised, it's delivering a system capable of helping organizations "unify multiple sources, sizes and structures of data with Workday data to deliver insights business leaders need for critical workforce and financial decisions." But in the process of collaborating with approximately 20 customers to develop the system, Workday discovered that many known, internal analyses are more in demand than speculative big data opportunities.
[ Want more on cloud application progress? Read Salesforce.com, Workday Keep Cloud Momentum Rolling. ]
"We're meeting them where they are today, which is dealing with enterprise structured data primarily," said Dan Beck, Workday's VP of technology product management in a phone interview with InformationWeek. "They're interested to know that they can grow into analyses of [multistructured] data from outside the enterprise as they're ready."
Examples of data that might come from outside the enterprise might include compensation benchmarking data relevant in HR analyses or sentiments about suppliers from social networks relevant to financial analyses, Beck said.
That's not to say that Workday Big Data Analytics' ability to handle a variety of data will initially go to waste. The last mile in many desired internal analyses involves data from outside of Workday. That means it doesn't necessarily conform to Workday data models, yet business users won't have to worry about complex ETL or data transformation to make use of that information, Beck said.
Workday has developed a set of 10 prebuilt templates for the most popular types of analyses requested by customers. A prebuilt template for Retention Risk and Impact Analysis, for example, calculates employee retention-risk and impact-if-lost scores by combining compensation and performance data in Workday with external CRM data. In another example, a Customer Profitability 360 template crosses revenue and cost metrics from Workday Financials with data from third-party (but internal) systems tracking support cases filed and customer satisfaction levels.
Other templates do involve analyses of high-scale, semi-structured data sources more worthy of "big data" billing. Compensation and headcount analyses, for example, blend internal, Workday-managed payroll and HR data with external salary surveys, benchmarks, and job postings and stats from recruiting sites like Monster.com and social networks such as LinkedIn.
The infrastructure in use behind Workday Big Data Analytics says it all: The service is running on Amazon Web Services, specifically Simple Storage Service (S3) storage buckets and Elastic Compute Cloud (EC2) processing. For now, the analytic software component from partner Datameer runs on top, in what amounts to a single, simple S3/EC2 node for each customer.
Datameer normally runs on top of Hadoop, the industry-standard big data platform. Workday is ready to switch to the Hadoop-based Amazon Elastic MapReduce service as data scale and variety warrant, Beck said, and plans call for the eventual behind-the-scenes deployment of Hadoop by Workday. For now, that's not necessary, Beck said.