Last month, IBM rolled out a new service called Voice of the Customer Analytics (VOCA), a subscription-based service that combines analytics, data mining and text analytics to give organizations an integrated view of customer sentiment.
It was a natural step for IBM, which has built out its business intelligence and data management capabilities significantly in the last several years thanks to a number of acquisitions, including Ascential for data integration, Cognos for BI and SPSS for predictive analytics.
But while Big Blue would seem well qualified to do the heavy lifting for a company, does it make sense for companies to outsource their customer analytics?
"There's a place for it; I don't think as big a place as [IBM] may be hoping," said Jill Dyche, partner and co-founder of Baseline Consulting Inc. "What I fear is this is going to lull executives into a sense of complacency that they'll outsource what should be very strategic work."
IBM's VOCA offering promises to mine diverse customer interactions like audio recordings, call transcripts, email and survey results to provide a view of customer sentiment. IBM can combine the unstructured data with structured data from databases and CRM systems to deliver actionable insight, according to Kevin English, global offering lead for CRM analytics. The Voice of the Customer program is delivered as a managed services offering and can be purchased standalone or as part of IBM's business process outsourcing. It will focus primarily on the call center market, English said, with users looking to improve customer satisfaction, up-sell and cross-sell.
Outsourced analytics is certainly not a new development.
"Companies such as Axiom, Experian, Dun & Bradstreet, Epsilon, Info USA -- all of these service providers have teams of statisticians and consultants that do just that. They provide it on a flexible basis -- either project-based or as retained services," said Leslie Ament, managing partner of Hypatia Research and Consulting. "The difference is, it’s mostly structured analysis. They haven’t yet integrated online with offline or structured and unstructured."
Although IBM -- with its vast consulting resources and the technology of SPSS, DB2, Cognos and IBM Content Analyzer to support it -- is probably better equipped to handle a company's customer analytics than the organization itself, buyers should proceed carefully. A voice of the customer program, when done right, needs to encompass the entire organization and be viewed strategically, Dyche said.
"It's very tempting to use them as these little survey factories when that data can be so much more meaningful and strategic to a company," she said.
Companies which do that often end up marginalizing the voice of the customer.
When is it right to outsource customer analytics?
Yet there are instances where outsourcing customer analytics makes sense, Dyche added – notably, in call centers and business processes that themselves have already been outsourced.
"The threat of this is that it's not so much the analytics but the data that will stay siloed," she said. "Then it's not the tool that gets slapped on that's the hard part. It's the data coming from these new systems that's really the hard part."
Alternatively, businesses that don't require much in the way of creativity when conducting analysis of customer data may be well served by outsourcing. Repetitive queries that don't require ad hoc reporting can be good candidates for outsourcing -- for example, sales reporting off point of sales systems, Dyche said.
The sheer amount of data and complexity of systems is something that organizations should consider when debating whether or not to outsource their customer analytics, according to Ament.
"Some organizations have up to 10 transactional and/or database systems due to multiple business units or offerings," she said. "To try to create algorithms or do data mining and content mining is incredibly challenging and almost impossible within the confines and inconsistent database structures, along with the infrastructure that they have in house."
Another consideration is whether outsourcing means potentially losing out on competitive advantage.
"If I were IBM and I were sifting through all that information, I would certainly want to take my findings and best practices and bake them into algorithms that I could package and resell," Ament said.
It's not necessarily an all-or-nothing equation either. Organizations have been happily outsourcing analytics for years, accessing some slice of expertise from an outsider.
"Marketing services providers have been doing that for ages," Ament said. "Experian and Dun & Bradstreet have done this type of work for Staples, even though Staples has [its] own internal analytics team. It's not necessarily that expertise is lacking but the infrastructure is lacking, and the only way to manage and integrate and create consistency in how information is managed and utilized is to outsource it."
Beware the pitfalls of outsourcing customer analytics
Yet beware of outsourcing customer analytics as "analytics with training wheels" or as a way of developing skills within an organization, Dyche warned. Because once it starts getting done outside the organization, it's tougher to bring back in.
"You may have an exec who has a mini epiphany and says, 'We don't know what customers are saying; we need a project,'" Dyche said. "It may be a way to kick-start [a voice of the customer program], but it will take a really disciplined organization to bring those capabilities in-house -- not just gathering and measuring it, but bringing that data to broader places. Marketing needs that information; sales needs that information."
So what should companies consider if they do outsource their customer analytics?
Ament said organizations with significant regulatory and compliance requirements should be very careful about what they outsource. In addition, she said, consider short-term jobs, like monthly or quarterly projects when the price is right. Finally, she suggests that outsourcing can bring some order to what can be a chaotic process.
"Organizations are working harder than ever to capture and manage their information. They're doing reporting very dutifully, but they're not always using best practices in doing their reporting," Ament said. "So standardizing metrics is something most organization don't do well but a consultancy or an outsourced provider should."