lutya - Fotolia
As business intelligence and analytics continue to grow in importance to business operations, some multinational organizations are grappling with problems that threaten to put up data siloes and minimize the effectiveness of their data analysis strategies. For example, they face issues revolving around how to comply with different data privacy laws in various nations while still ensuring that BI and analytics insights can be shared between separate offices across geographical boundaries.
One of the main challenges is regulatory in nature -- and it's likely to become an even bigger hurdle as the European Union works to finalize sweeping regulations that set higher standards for personal data collection and analysis in EU countries. The new measures -- which were released in draft form in 2012 and are expected to be finalized later this year or in early 2016 -- will heighten already stringent EU privacy protections. By comparison, the U.S. has relatively few data regulations outside of the financial services and healthcare industries.
As a result, businesses that operate in both the U.S. and Europe have been left to grapple with the question of how to do meaningful analytics on their data while complying with current EU restrictions and planning for upcoming changes. One step organizations have taken is to revise their data processing and management policies. Google is a case in point. In 2013, the company updated the terms of service for its Google Analytics offering to clarify how it processes, stores and moves data. Google also introduced a browser add-on that enables European users to opt out of having their Web browsing data collected.
Separate regions, separate systems
Technology can play a role as well. For Groupon Inc., the answer was to partition its data centers. Speaking in a webinar sponsored by data warehouse and analytics software vendor Teradata, Kotesh Mukkamala, senior manager of engineering for Groupon's enterprise data warehouse architecture, said the Chicago-based company has a massive data infrastructure. But to comply with EU data regulations, it maintains separate sets of Teradata systems and Hadoop clusters in the U.S. and Europe.
"There are different restrictions, which makes partitioning the data important," Mukkamala said, adding that the need to keep the data sets separate is something he and his team have to think about regularly. Groupon, which offers discount purchase deals through its website, collects and stores data about every user contact; the information it tracks ranges from how many people buy specific deals to how users interact with the company's mobile app.
As part of its analytics strategy, Groupon does a lot of A/B testing on the data to analyze what's working and what isn't. Running those tests on data from all its users potentially could deepen the insights, Mukkamala said. But he noted that the company could find itself in hot water if it were to combine data sets from its European users with those of its American customers. For example, the draft version of the proposed EU General Data Protection Regulation would authorize penalties for noncompliance, including more frequent compliance audits and, in severe cases, fines of up to 1 million euros or 2% of a company's global revenue, whichever is greater.
Analytics issues come in all sizes
Challenges in developing effective data analysis strategies across multinational operations don't only affect large organizations. Redbubble, an online marketplace for artists and designers, has offices in Melbourne, Australia and San Francisco. Compliance with differing data collection and analysis laws wasn't as much of an issue for Redbubble -- its analytics uses are mainly limited to data from internal departments, like finance and marketing, and Australia doesn't have as strict privacy regulations as European countries do. Manoj Yadav, the company's director of business analytics, said his big concern was that he didn't want to leave important BI insights stranded in one of the two offices.
To avoid that, Yadav and his team implemented a cloud-based data discovery and governance platform from GoodData. The cloud analytics technology allows users in both locations to access the same data sets and run the same kinds of analyses.
Deploying on-premises BI tools instead of a software as a service platform wasn't a viable option for Redbubble. "For me, to implement a non-SaaS tool, that would have been completely counter to what we do," Yadav said.
Analytics strategies are key to big data success
Think about how to get front-line workers to adopt analytics tools
Data analytics strategies boost business efficiency