Networks Smarten Up
September 25, 2011 - Openet
It’s a graph we’ve all seen at conferences: a hockey stick-shaped line forecasting data use overlaid with a flattish line showing sluggish revenue growth. Everyone understands the implications of that graph. Operators need to rein in data use and run their networks more efficiently while finding new ways to make money.
The U.S. wireless industry has taken heed. AT&T’s decision to end its unlimited data plans in June 2010 set off an industry-wide shift toward tiered data, with Verizon Wireless and T-Mobile USA coming out with their own pay-to-play plans. Those tiered plans will help operators address the problem laid out on the too-familiar data chart by holding back the traffic tide and getting data hogs to pay more for their usage.
But some don’t think operators should stop there. After all, wireless networks are more than just conduits for voice and data – they’re veritable gold mines of information on traffic and subscribers. With the right probes and analytics in place, operators can glean information on demographics, location, mobile content, applications, traffic and devices.
“Today’s use of intelligence is a lot more about understanding opportunities to improve the performance of the network,” says Yankee Group analyst Susan McNeice. “I’d like to see more market trials of innovative applications, innovative offers... For too long, the only lever we’ve been pushing is on price.”
In theory, the data could be the solution to the dumb pipe problem, allowing operators to use the information latent in their networks to power out-of-the-box products like premium mobile video services, targeted mobile ads and app-specific service plans.
In reality, however, things are a little different. Most operators have so far shied away from the pie-in-the-sky implementations dreamed up by the more forward-looking members of the wireless industry. Instead, the information is finding a low-profile use in something far more fundamental: optimization and customer service.
Wireless operators are putting network intelligence to work, but they’re starting with the basics.
At Sprint, network analytics are being used to help call center workers handle subscribers’ calls about their wireless service. The information also helps Sprint stay ahead of massive increases in data traffic, allowing it to see where congestion is currently occurring and predict where capacity will need to be increased in the future.
“We use information about how our customers use our products and services in our care organization and as an efficient planning approach to our network – what we need to provide to our customers so they have the right experience,” says Lonnie Sauter, senior manager of network planning at Sprint. “We’ve shown we can increase the customer experience and also lower our costs at the same time using these tools.”
The operator has seen a year-over-year doubling in data traffic, and that growth has increased its need to accurately predict trends on its network.
Part of that is keeping track of applications. Sprint is able to see how applications are affecting its network and has used that information to help developers make their apps more efficient.
Developers don’t prioritize data efficiency when designing an app. For example, research conducted by Onavo found that Twitter’s own in-house app used twice as much data as Tweetdeck because of the way it cached data.
“It isn’t just about optimizing the network. It also allows you to find the root cause about high traffic volumes,” Sauter says. He declined to name names, but said there were a “few cases” where Sprint was able to identify and address apps that were especially egregious data hogs.
Sauter says Sprint isn’t interested in using its analytics to upsell its customers on premium services, which would go against the ethos of its all-inclusive unlimited plans. For now, primary use of the information is making customers happy and keeping Sprint ahead of usage trends that could impact its network.
Tiered data plans are an important step to getting customers to pay premiums for new services – consumers accustomed to flat-rate pricing aren’t in the mindset to fork over additional cash for an upgrade.
AT&T, Verizon Wireless and T-Mobile all have their own flavors of tiered data. From a marketing perspective, there’s nothing holding them back. But so far, most seem to be following Sprint’s lead and have steered clear of using information about their customers and network to offer premium services.
“There’s an infinite potential of service offerings that could be provided, but it’s completely unclear which will have resonance in the market and which could actually hurt the revenue curve,” says Openet marketing and product manager executive Chris Hoover, citing some of the more outré service offerings suggested by industry insiders as a way of squeezing more money out of wireless networks. “Operators who already feel their revenue is under threat are risk averse as a result.”
Operators are playing it safe. Products like Telefonica’s opt-in mobile marketing service aren’t likely to pop up in the United States any time soon.
COMPLEXITIES OF THE COTTAGE INDUSTRY
Customer care, optimization and traffic management are likely to be the core uses for network intelligence technologies for some time to come. However, that hasn’t dampened prospects for companies selling new products for measuring, analyzing and managing traffic.
Operators are on the hunt for technology that will let them better understand and control the increasingly complex surge of traffic on their networks, and vendors have stepped up to the plate with solutions.
An entire cottage industry has sprung up around network intelligence. While this is helping push the technology forward, it also has created a good deal of complexity and fragmentation in the market.
With dozens of products to pick from, each offering different capabilities, operators looking for simple ways to understand and control what’s happening on their networks face a complex task. Operators are bombarded with new products claiming to be the must-have solution to the capacity crunch. To be sure, some of them may prove to be useful, but that doesn’t make sorting through them any easier.
Some products function as software that must be preloaded onto handsets. Others reside on the network level, but may not be open to all network elements. For instance, a probe might be able to inspect some layers of Internet traffic, but not all. A company that manufactures probes for deep packet inspection may bundle the hardware with some sort of analysis software, but might not provide the technology that then uses that passive monitoring for an automated network management system. That requires integration with an operator’s policy control systems – and a whole other sector of vendors.
Andrew Coward, who heads marketing strategy and product management at Carrier IQ, sums up the situation with an anecdote.
As Coward was explaining Carrier IQ’s product to a potential operator customer, the man expostulated: “We’ve already deployed 130 different management tools to try to get these different piece of information – why buy the 131rst?”
Coward’s response: “What took you over 20 in the first place?”
Carrier IQ makes device software that allows operators to see how customers are using their phones by gathering statistics on applications and other factors.
Operators looking for an end-to-end view of their network, from customers’ quality of experience to deep packet inspection, will have to patch together a hodge-podge of different technologies.
“In a lot of cases, this is all quite new and a lot of these solutions are still being offered by the smaller vendors,” says Informa analyst Mike Roberts, citing companies that lack the scale and reputation of established vendors like Alcatel-Lucent, Ericsson and Nokia Siemens Networks. “For larger operators’ bread-andbutter systems, they want to see the larger vendors offering these things as a stamp of maturity and robustness. A lot of operators are reluctant to try it until they’re convinced it’s bulletproof.”
Some companies have taken steps to address the fragmentation problem by joining the Network Intelligence Alliance (NIA), a group of 17 companies offering various products to help operators better understand and manage network traffic. The NIA strives to bring a measure of standardization to the technology by giving its members a place to collaborate and educate the market.
“This is really complicated stuff,” says Qosmos marketing executive Eric Larsen. Qosmos specializes in software engines for deep packet inspection probes and helped found the NIA. “There are a lot of different bits and pieces, and in order for them to work together they often require technical integration – there’s a whole chain of people that need to work together. What the NIA is doing is making them more efficient from a technical standpoint and educating the market on the solutions that are available, the ecosystem, what works well together and what’s important.”
Back at Sprint, Sauter says that he receives a fair share of pitches on new analysis technologies. Most of them, he says, aren’t up to par.
“We routinely evaluate a number of different companies that have something in this space, and my experience has been that very few of them do true analytics,” he says, defining “true analytics” as an insight a business can take action on to improve customer experience or reduce costs. “While there are a number of them, the number I would consider viable to adding to the customer experience – there’s just not very many.”
Monitoring and analysis tools have become a must-have in the age of soaring data use, and it’s a safe bet to assume they all have some sort of network intelligence technology in place. However, it’s not clear exactly what kinds of tools operators have in place to monitor and manage their networks. Sauter declined to specify which technologies Sprint uses to analyze network traffic, as did AT&T, Verizon Wireless and T-Mobile.
For all we know, operators could be holding off on value-added services because they lack the back-end systems to support the products, not just because they’re not sure the products are a good idea.
The number of new technologies on the market – each claiming they’re the must-have solution to analyze, manage and monetize traffic – suggests that operators are still looking for the right ways to gain insights into their networks and take action on that information.
The fragmentation in the marketplace doesn’t diminish network intelligence technology’s most basic function: to better understand and manage network traffic.
“The Holy Grail is to provide ways to turn on new sources of revenue, new services to the end user versus just trying to optimize based on raw traffic patterns,” says Stuart Benington, director of portfolio strategy at Tellabs. “If you can have the infrastructure in place that gives visibility and the ability to act on that visibility, that gives an operator the chance to create long-term profitable growth.”
Network intelligence will play a huge part in helping operators manage their networks and improve customer care. Whether it will eventually be used for value-added services is up to operators, and so far they have shown little interest in the upsell business.