Improving data management and analytics has always been part of General Motors' three-year IT overhaul, but automaker's recall crisis makes it even more critical.
Can better data improve how General Motors operates?
When GM set about overhauling its IT operations and strategy shortly after emerging from bankruptcy protection, its plan to shift away from outsourcing and hire thousands of US tech pros grabbed most of the attention. But more than a year into that three-year overhaul, GM's strategy for improving its data collection and analysis is moving onto center stage.
GM's most pressing issue at the moment stems from the allegation that it was too slow to recall Chevy Cobalt and other vehicles with faulty ignition switches, failures linked to 12 deaths that have recently sparked a series of lawsuits. Creating a single enterprise data warehouse for quality analysis and other purposes was always part of the IT transformation plan drawn up by GM CIO Randy Mott, whom the automaker hired in early 2012. But the recall crisis makes improving the company's data analysis ever more critical.
Mott says GM now has moved 1.1 petabytes of product development, procurement, logistics, quality, manufacturing, customer care, sales, marketing, finance, and other kinds of data into its new enterprise data warehouse (EDW) and plans to move a lot more of it. Mott says the EDW's main purpose is to help the company look forward -- GM has started using that data to improve vehicle quality, Mott told InformationWeek as part of a broad update of the company's IT transformation effort.
Would the current data warehouse effort have helped GM spot a problem like the faulty ignition switches and deal with it sooner? we asked.
"Had we had an easy-to-use, broad way to analyze data, would we have come to different conclusions in the past? I'm sure we would, whether it's this one or any other," Mott says. Spotting vehicle quality problems early on has been complicated by the fact that GM didn't bring manufacturing data together in a single place for analysis, he says. "It's the real world -- things were divided by brands, thing were divided by different vehicle models inside those brands," Mott says. "And getting that information in a comprehensive way is not easy for anybody. Certainly it wasn't easy for us, and this [data warehouse project] certainly allows us to do that."
Mary Barra, GM's CEO since January, has apologized for how GM handled the recall, which now covers about 1.6 million vehicles built between 2003 and 2007. Barra has started an investigation into what went wrong. She has also made it clear that part of the job description for GM's first-ever global vehicle safety executive, VP Jeff Boyer, will be to use data to prevent safety problems. "He will look across the organization, identify and analyze the data that connect them, and give the company the most accurate, real-time safety-performance picture of any of our vehicles around the globe," Barra wrote in USA Today.
Mott has dedicated a team of 350 to 400 people to the EDW, which GM is now populating with both structured and unstructured data. It has a Hadoop layer, called the Landing Zone, at its core for ingesting data, and a Teradata layer, called the Enterprise Analytics Zone, for standard data analysis in areas such as production and finance. Alongside that analytics layer is what Mott calls a Discovery Zone, where GM staff can use tools from the likes of HP Vertica and Teradata Aster to query very large datasets as part of ad hoc research. Above those layers are software tools for consuming data, ranging from IBM Cognos reporting tools to HP Autonomy search tools to IBM SPSS advanced analytics tools to Tableau Software visualization tools. On top of those layers is a .Net-based information portal.
GM already has collapsed about 55 of 200 data marts scattered across its operations and geographies worldwide into this EDW architecture. In addition to existing sources, data from new applications GM's IT team is building, such as an app it's providing dealers to run their service operations, will feed into the new EDW. Mott produced a chart that shows, by subject area, how much of GM's data will flow into the EDW by the end of 2015, ranging from 100% of IT operations data and 90% of quality data down to 50% of logistics and procurement data.
Consolidating data into one place is meant to make it easier to correlate information and dig out trends and potential problems. GM is already collecting data from its 170 factories worldwide and plans to segment and analyze it down to the VIN number, to improve safety and quality and to assess profitability. "You're able to say very quickly, 'I've got these six incidents -- what's common about them?'" Mott says.
Insourcing: 12,000 IT staff ahead Beyond its enterprise data warehouse efforts, GM is moving forward with other elements of its three-year IT transformation.
As it moves from doing 10% of its IT work in-house as of January 2013 to doing 90% of it in-house by the end of 2015, GM has been on a hiring blitz. First, it brought on thousands of employees from one of its outsourcers, Hewlett-Packard, and it's also hiring experienced IT pros as well as recent college grads (600 by the end of this year) at four main development centers
Chris Murphy is editor of InformationWeek and co-chair of the InformationWeek Conference. He has been covering technology leadership and CIO strategy issues for InformationWeek since 1999. Before that, he was editor of the Budapest Business Journal, a business newspaper in ... View Full Bio
We welcome your comments on this topic on our social media channels, or [contact us directly] with questions about the site.
Digital Transformation Myths & TruthsTransformation is on every IT organization's to-do list, but effectively transforming IT means a major shift in technology as well as business models and culture. In this IT Trend Report, we examine some of the misconceptions of digital transformation and look at steps you can take to succeed technically and culturally.