Page 435 - ITGC_Audit Guides

As with other IT projects, this process must begin with gathering comprehensive requirements to
understand the questions the consumer is trying to answer or issues the consumer is trying to
predict. Involving internal consumers in the design and testing process helps them develop a sense
of ownership in the solution. Any post-implementation feedback should be addressed promptly
to sustain and increase adoption. Organizations should also plan marketing and training
campaigns to share success stories and educate internal consumers on the potential of big data
analytics. Stakeholder surveys are an effective tool for obtaining feedback and lessons learned to
improve development processes for subsequent implementations.

Analytics and Reporting

Reports should be designed with the appropriate flexibility of input parameters (e.g., start and end
dates, customer segments, and products) to allow consumers to narrow or broaden the focus of
their analysis. This flexibility enables consumers to ask questions that might not have been
anticipated during the initial development phase and supports adoption by empowering
consumers with self-service capabilities, which helps minimize the traditionally slow and costly
report development lifecycle. The available granularity of report data to support consumers’
standard or drill-down reporting should be balanced with consumer requirements, processing
capabilities, and data privacy concerns.

Self-service tools are important for activities that involve customers, vendors, or employees who
need to make fast decisions. For example, a customer service representative can use big data and
a self-service reporting application to view a customer’s product and service history across
multiple organizational lines on one screen. This would reduce the number of phone calls the
customer would need to make to answer product inquiries. Privacy and security concerns can be
addressed by restricting access to sensitive data fields to only those consumers who have a valid
business need to see those data fields.

Many people are familiar with the concept of predictive analytics, which attempts to explain what
will occur next based on historical data. For example, hospitals utilize predictive analytics to
determine which patients may be readmitted for additional treatment. Data scientists can apply
survivor analysis algorithms to help human resources departments predict employee
dissatisfaction, and that information can be used to support workforce management and planning
activities.

Analytic reports may also be alert-based, to help consumers identify which actions are needed to
address a particular situation. For example, sentiment analysis techniques can be applied to
determine a customer’s satisfaction with a product or service, based on information shared by the
customers via social media. High satisfaction levels can drive new distribution strategies, while
poor satisfaction levels may require immediate remediation actions to protect customer loyalty
and the organization’s reputation.

16 — theiia.org