With retail operations in eight countries and a growing portfolio of telecommunications services, Carphone Warehouse is currently Europe’s largest independent mobile retail outfit. Paul Scullion, head of business intelligence at Carphone Warehouse explains how the company is using a combination of big data technologies to help improve retail customer service and eventually, help telcos improve their offerings.
“We have a vast amount of data but at this point, I wouldn’t necessarily call it ‘big data’,” Scullion says. “The majority of our information is transaction-based. But when I think of ‘big data’ it’s all about the [high] velocity of the data; it’s about dynamic data, and the kind of analysis you do with that data.”
But, as Scullion explains, the company has made significant progress towards using old data acquired through traditional BI and capturing new data to do the sorts of things most would normally associate with big data, as hackneyed as the term may be.
A few years ago the company moved away from its Oracle-based data warehouse and adopted Netezza, which is hosted in an IBM datacentre (Netezza was acquired by the company in 2010). He says the data warehouse improved performance significantly because of the way it cuts up the data being targeted processes small bits in parallel before sending the results back.
“The performance over Oracle was epic. We had queries that took an hour or Longer with Oracle that now take less than a minute,” he says.
The company, which partners with Accenture for many IT initiatives, uses Informatica’s ETL platform and has recently adopted MicroStrategy’s mobile analytics and dashbooarding tools, which Scullion helped roll out to all 6,500 UK sales representatives at Carphone Warehouse.
Each sales rep has an Android-based tablet available to them, and it is increasingly becoming the central platform through which a wide range of internal data is accessed. Scullion believes is has the potential to improve sales staff engagement, and makes it easier for the company to support them.
He says some dashboards enable the retail staff to see their sales performance, including the product mix they’re generating, how they’re faring relative to other colleagues. In the main dashboard there’s 7 KPIs per colleague, and each has a ranking and an overall ranking so colleagues can see how they are faring relative to others.
“We are trying to put other supportive applications on that same device, so that retail staff have just one place to go for their internal news and systems rather than needing to navigate a series of back-office terminals and applications, and we are looking at potentially moving our MicroStrategy mobile dashboard platform out into the cloud so that we can benefit from the vendor’s economy of scale.”
Where big data can generate big value
In the UK the tablets are also used to help customers navigate the wide range of device and tariffs available to them, which is where Scullion says the company is really starting to get into big data. Through a series of questions retail staff help customers drill down into how they intend to use their devices, and what it is they’re looking for from a network operator.
“Historically the mobile industry has been a bit bamboozling for customers. They feel like they’re being pressure-sold to by a used car salesman,” he quipped. “But the sales journey tool that we use makes it really transparent for the customer. And critically, we record every step in that journey.”
“What we’re able to do is look at that complete journey, record every step in the journey, every button pressed, every bit of data entered, for journeys that end in a sale but more importantly for ones that don’t end in a sale. And what we’re trying to do is look at where in the process certain pressure points put customers off,” he says, adding that the company is looking to do the same thing with its online sales channels.
Scullion says that many customers come into Carphone Warehouse stores with a preference for certain networks, but after being guided through the tariff options often change their minds, ending up with a different operator for a variety of reasons.
And that data, as one might suspect, is extremely valuable to mobile operators. With telcos looking to big data to boost long-term customer retention among other reasons, and mobile markets becoming more and more competitive, it’s clear there is rapidly increasing demand for this kind of information – particularly high-volume, multi-channel, multi-operator retail outfit like Carphone Warehouse.
“Certainly the insights we’re gathering on decision-making throughout that sales journey could be used to improve the options networks provide. They could help operators come up with more attractive tariffs; they could also be used to help explain why they may be losing so many customers to another operator,” he says.
Arguably the most disruptive telco in the UK market today, Three UK is looking to leverage insights generated from its vast troves of network data in a bid to improve customer satisfaction and how the business caters to segmented customer needs, as well as derive incremental value from existing customers.
Stefan Grew, data management & information service manager enterprise information CIO tells Business Cloud News that the company is at the outset of a challenging yet rewarding journey that will ultimately keep Three ahead of other large UK telcos.
At over 8 million subscribers Three isn’t quite the size of Vodaphone or O2 in the UK mobile market, but it’s a very respectable and sizeable chunk nevertheless – and one acquired in part as a result of its very favourable data pricing policies. It’s the only telco in the UK to offer uncapped data tariffs at competitive prices. And as a telco, regardless of its size relative to incumbents, Three is one of the few types of companies genuinely facing a big data problem.
Grew says that two years the company agreed that it wanted to move away from being a price-led company.
“We wanted to focus more on customer enjoyment, give the customer something they wanted. What was obvious was that customers wanted to use a lot of data on their devices. But moving from there it became about how we enrich the experience and give them new capabilities,” Grew says.
“The only problem with that is we didn’t really know enough about our customers, we didn’t have a consolidated view of that. The key question is, how we can deliver some value and pull together a 360-degree view of the customer?”
Grew, who is tasked with developing the data roadmap for Three and making sure processes and infrastructure is in place to support that, says that the company has already done a lot of work at the data warehouse level by integrating network data and customer / pricing data from its business support systems (BSS). That integration is essential for developing this comprehensive view of the customer.
It’s the first step in a long process that will initially see Three roll out business analytics to its various business units.
“I think part of the challenge comes back to good integration between operational support systems, business support systems and the data warehouse. I think we have latent problems there,” Grew says.
“As we mature and as we start to develop out new BSS capabilities, we need to have thought about that in advance rather than waiting for it to come and then dealing with the data side of it after, which we haven’t necessarily done so well in the past.”
But he says Three, the youngest telco in the UK market (beyond EE, which was really a marriage of incumbents), has done a lot of work to adapt its approach as it endeavours to build out not just an analytics platform, but a sustainable infrastructure that will support its big data initiatives in the future – which includes ensuring strong integration between BSS and the data warehouse.
“We understand the changes we need to make. If we do that all in the right way, and capture the data at the start and think about how we want to use that data from the outset, we’ll be able to demonstrate our successes more effectively,” he says. “The challenge for us is to make sure we understand how the business wants to use that data.”
With business intelligence analytics being extended to new regions throughout organisations a very different approach to standard business analysis is needed, Grew says. Business leaders need information more quickly, and at the same time don’t always know what data they need in order to satisfy key business questions.
“Climate capture is still vital, so when you’re doing big structured deliveries you still need to have that very structured climate capture capability, but we need to complement that with something more agile. I think moving into an environment where from the data perspective we’re viewing it more transiently, working more iteratively, allowing users to prototype for themselves. The more we empower them to do that, the better.”
In June, Three launched the first internal pilot for its analytics and data visualisation platform. The company invited employees to bring one business question with them and, given certain data sets –service usage patterns, purchasing trends and network stats – the team was able to answer some of these questions fairly quickly.
Grew says the process reinforced business users’ ability to use the platform, which is often half the battle, and helped them think about how to connect the data the company has with potential business outcomes and actions. Self-service data analytics is where Three is heading, and the company hopes to be able to empower its users to direct these platforms towards other areas of the business in a bid to generate more incremental value from its existing customers.
“We already have the base data for usage, so bringing that up into that same mechanism so that we’ve got that additional flavour so customers start to generate incremental value to us. We think we understand what that really means in terms of how they’re using our services, and I think over the next 6 months or so we need to look at getting a good prototyping capability in place so that we don’t spend so much time building strategically only to find that we’ve not met what the business needs.”
“As we look at better ways of handling our vast amounts of network data, we’ll start to try and think about how we collect our customer sentiment information, how we interact with third parties around marketing and segmentation, use this information tonreduce churn and so forth. I’d really like to provide some of these things, but it takes time to get to step one”
“We’re really right at the beginning of our big data journey,” he says.
Infrastructure giant Ericsson forecasts total mobile data traffic growth will increase ten-fold between now and 2019, on the back of smartphone subscription growth that is expected to grow at four times the rate of total mobile subscriptions, and total global LTE subscriptions that are forecasted to grow 13-fold.
The company presented the findings from its latest Mobility Report at an event in central London Tuesday. Key themes of the report were the rapid growth in global smartphone penetration and LTE network coverage and the consequent jump in global mobile data traffic.
In 2013 global LTE subscriptions stood at around 200 million, according to Ericsson, and that number is forecasted to hit 2.6 billion. China is expected to account for an increasing proportion of global LTE subscriptions, with is total forecasted to hit 700 million by 2019, meaning it will account for over a quarter of the global total.
Speaking exclusively to Telecoms.com, Patrik Cerwall, the head of strategic marketing and intelligence at Ericsson, added further colour on China’s increasing dominance of the global mobile market. “The total amount of data generated by LTE subscriptions in China in 2019 will be greater than the global total is today,” he said. Cerwall also highlighted that, in common with many other technologies, the appetite for bandwidth is unlikely to ever be satisfied, with video streaming continuing to dominate mobile data use.
One of the major drivers behind China’s mobile growth has been the rapid uptake of smartphones by the country over the past couple of years. With much of the world’s technology manufacture happening on their doorstep, Chinese consumers have been quick to snap up cheap Android smartphones that cost little more, if at all, than the feature phone equivalent. Chinese smartphone penetration is already up at the same levels as many western markets so, given its massive population, it’s safe to assume China will continue to dominate the global mobile market for the foreseeable future.
US API developer Apigee this week secured a further $ 60m in funding, bringing its total raised to date to $ 171m for big data analytics development.
Pine River Capital Management and Wellington Management Company participated as new investors in this round, along with current Apigee investors Norwest Venture Partners, Bay Partners, Third Point, SAP Ventures, and funds managed by BlackRock, Focus Ventures, and Accenture.
Earlier this month the company upgraded its big data analytics platform, claiming to enable enterprises to increase customer satisfaction and revenue by predicting and immediately adapting to the needs of individual customers at scale.
Apigee Insights lets businesses apply sophisticated predictive analytics on fine-grained data and then proactively act on those insights in real time, delivering individualised services through predictive apps.
Analyst house Ovum recently flagged the company as one to watch for organisations embarking on digital transformation initiatives. “The combination of API/application infrastructure, analytics, and strategy services offered by Apigee is well suited to enterprise API initiatives focusing on driving business growth via digital supply chains,” said Saurabh Sharma, senior analyst at Ovum. “Enterprises should look to engage with Apigee as a strategic partner for digital transformation initiatives, and not as a typical API management solution/infrastructure provider, to realize more value from their investment.”
Two weeks ago Apigee was contracted by Telecom Digital Ventures New Zealand, a division of Telecom New Zealand, to power Qrious, a big data and analytics platform designed to deliver a better way for organisations to tap into data insights and combine them with other available data in the ecosystem.
Apigee has a strong presence in telco API initiatives. In 2012 the company was tapped by Telefónica to lay the technical foundations for an API platform that allows third-party applications to hook into the Telefónica network. The two companies have worked together in the past on Telefónica’s BlueVia application development platform.