AMSTERDAM – Communication Service Providers (CSPs) need a self-imposed shift in corporate culture in order to capitalize on the massive amounts of potentially useful data they have at their fingertips.
That's one of the conclusions from the first morning's debate at the Big Data Analytics Summit here in the Dutch capital.
The TeleManagement Forum (TM Forum), the event's organizer, attracted a number of CSP executives from the likes of Belgacom SA, Cricket Communications Inc. and Orange Polska (with many more to come) to talk about why and how they are approaching the so-called Big Data challenge.
The overall theory is that by extracting useful information from the information that is currently hidden and untapped in many systems and networks, CSPs can provide a better customer experience, develop more targeted and intelligent marketing programs and develop new business models that will ultimately enable them to maximize their top and bottom lines.
Here are my takeaways from the conference sessions so far:
The ability to derive value from the information generated by networks, devices and subscribers is reliant on a dramatic, almost counter-cultural shift (for telcos) in corporate strategy and internal processes. Telcos "need to be able to prioritize and make decisions -- neither of these happen today as [telcos] don't have the right structure to do this," said Peter Crayfourd, a former senior customer experience executive at France Télécom – Orange, who is now an independent consultant.
"The human aspect is as important, if not more important, than the technology," stated Belgacom Business Intelligence Manager Wim Castuer at the end of his presentation. He outlined how the Belgian incumbent reorganized and brought together its IT and business/marketing teams in an effort to figure out how to get from "data spaghetti to [a] structured data model."
Telcos have been talking about these ideas for years (without the Big Data tag) but have done little about it: Now, with margins squeezed, profitable legacy services in decline and competitive pressures coming from Web services players, there is evidence of action. CSPs need to make sure they don't get left behind: It's time to jump on the bandwagon.
The CSPs need help from the industry -- there's only so much they can do themselves. "We need a lot of help from the business intelligence experts," said David Muragi, head of multiscreen operations at Portugal Telecom SGPS SA. But total reliance on Service Provider Information Technology (SPIT) systems and third party suppliers isn't recommended -- there needs to be some bespoke in-house action: "We need to identify the right people internally and then build on top [of the systems implemented]," he added.
Big Data strategies need buy-in from the very top of a CSP and then all the way down to the customer-facing staff if they are to have any impact.
New metrics are required, said a number of speakers. Traditional KPIs (key performance indicators) are practically meaningless, noted Crayfourd and other speakers. "I can't think of a worse measurement of a customer's value than ARPU [average revenue per user]," noted Jim Hayden, executive director of business intelligence at Teoco Corp. (See Teoco to Buy TTI Telecom.)
Big Data projects can't be forced into a fixed timescale: Once started, it will never stop. "You are never done," said Cricket's director of OSS, Stefan Spaar. "It's a journey, not a one-off investment," added Belgacom's Castuer.
Big Data projects are not big bang projects: They need to be implemented in an incremental fashion and CSPs should experiment as they figure out how the data from their multiple systems and silos can be organized in a more structured fashion and then mined for useful data. "Experiment by using [an] analytics sandbox," suggested Teoco's Hayden.
In terms of the technology underpinning CSP business intelligence/data warehousing/analytics strategies, some of the key IT suppliers are Teradata, The Apache Software Foundation (Hadoop), IBM Corp., Oracle Corp., EMC Corp. and SAS Institute Inc. That Teradata's business intelligence IT systems appear to be embedded in just about every CSP's back office looks like good news for emerging telco analytics specialist Guavus Inc., which is also here in Amsterdam and which has just announced a partnership with Teradata. (See Guavus Teams With Teradata.)
But CSPs should try not to be too hung up on the technology, noted Teoco's Hayden, and should not get obsessed either by the return on investment (ROI) on implementing a data analytics strategy (where the costs are a mix of operating and capital expenditures). "It's very hard to quantify how valuable [in financial terms] it is to know more about your customers. If you try to quantify it [then you'll] probably be wasting your time," said Belgacom's Castuer.
But there are ways to develop new revenue streams from Big Data operations: Cricket's Spaar explained that his company is currently studying how the business intelligence data it generates can be used for mobile marketing applications.
And to finish off, here's the quote of the day:
"Big Data sounds sexy but it's actually very dirty" -- Tapan Bhatt, senior director of industry solutions, Splunk Inc.
— Ray Le Maistre, International Managing Editor, Light Reading