The Candy man can
Back in 2006, in a high profile bid to differentiate itself from the pack and drive uptake on its WCDMA network, 3UK launched a flat-fee model for mobile internet access. It may not have been the first such approach the mobile world had seen, but it was certainly one of the most influential.
Six years later operators all over the world are trying to extricate themselves from the flat rate data tariffs to which much of the industry made a move. The uptake these tariffs encouraged, and the resulting network strains, have been the defining narratives of the industry’s most recent years.
The Hutchison-owned 3 group has a history of disruptive approaches to the provision of mobile data services, which owes much to its position as a challenger across all of its markets. In keeping with this, and unlike the great majority of mobile broadband operators, 3 is sticking to its guns on data pricing—a stance that reflects a fundamental belief about the role of the operator in 2012.
That role must be seen in the context of a digital world made up of a series of clouds, says Ed Candy, 3’s group technology strategist and former group technology director. Microsoft, Apple, Google and eBay—and a growing number of other players— each have their own cloud or economic environment, he explains, and the job of the mobile broadband operator is to provide users with unfettered access to those clouds.
At a time when many senior operator executives are calling for these kind of companies to subsidise the cost of providing the traffic that their services are generating, Candy insists that the truly innovative operators are the ones seeing the opportunity that the new world offers. “Operators getting upset about over the top services is a bit of a joke,” he says.
The marketing department may have cooked up the pricing strategy but it falls to Candy’s team of CTOs to deliver the service. Given Candy’s insistence that all legitimate usage is good usage, there is no let-up in the pressure on the teams he oversees to deliver a robust network performance. While Candy is a progressive thinker when it comes to OTT services, he’s something of a traditionalist in his view of the network. For all the enthusiasm around small cells, wifi offload and other capacity relief technologies, there is no substitute, he says, for a well configured, capacityrich macro network.
“If you put in small cells, they won’t track the movement of the traffic,” he says. “Whereas if you have a well-configured, overlapping macro network, it can cope when the user density—the traffic hotspot—moves from one end of the street to the other. The cells are built out for capacity so there’s a good overlap at the edge. The real disadvantage with some of the small cell applications is that there are limits to the cell overlap you can get.”
The absence of any kind of handover is a serious handicap for wifibased offload strategies, Candy adds. “Wifi is very much a local solution, but the real art is making it technology- invisible.”
Unless the movement between different networks can be executed out of sight of the consumer, he suggests, offload can be a challenging user experience. “We’ve seen quite a lot of resistance to offload once you move away from the expert user,” he says. “As soon as you lay down a reliable, high data-rate 3G network, the users prefer to stay on that, rather than switching back and forward between cellular and wifi—even in the home. Users at the top end of the skill profile might prefer to go over to wifi for specfic activities but that’s not normal consumer behaviour,” he says. Candy also expresses misgivings over the “boundary conditions” of wifi offload, when the device has the option of both cellular and wifi connectivity to choose between. “There are serious issues around how you handle the boundary conditions; The interaction between the two technologies always causes disruption,” he says.
It’s a problem familiar to operators who struggled with the interworking between second and third generation systems after the introduction of 3G, he adds. “The worst thing that can happen is when the customer finds themselves on the boundary of both services and neither is working well,” he says.
This issue has cycled round once again with the introduction of LTE. So far 3 has launched LTE services in Sweden and Austria and Candy says the top line learning is that the technology works very well. But, he says, the dramatic improvements that some operators have reported (and all vendors have promised) are less readily available to 3 because the firm’s existing kit is more advanced than the legacy infrastructure operated by its older competitors. “LTE gives you a really good improvement if you’ve got a crap, second generation Release 99 network and you’re building out a new core and a new RAN,” he says.
“The networks we built were very much an IP enterprise core with a 3G interface. So once we stabilised and got them working really well, they were already quite efficient. When you compare like-for-like, with similar amounts of spectrum, similar technologies and similar cores, the improvements in spectral efficiency and performance are not as good as you might expect.”
Candy says that operators and markets that might be deemed behind the curve with LTE as deployments gain momentum—and the UK, home to 3’s flagship operation, is chief among them—have little cause for concern, however. The LTE terminal sector has a long way to go before it achieves the kind of scale and diversity that will enable serious consumer uptake, he says, meaning there is time to spare. Day to day there are few opportunities for CTOs to relax, though.
Like other operators, 3 must run to keep pace with consumers’ appetites for mobile data. While the different netcos in the group share a number of common platforms and architectures, they each have their own CTO and a good deal of autonomy. Candy’s responsibilities cover strategic planning across the portfolio and, organisationally, the netcos are structured so that strategic and tactical responsibilities are divided between different teams. “In today’s world clearly the growth in data is the major challenge, Candy says. “And that’s about short term activity which is the responsibility of the guys who are working the current asset base and executing the agreed upgrade plans. You have to be very fleet of foot.”
The data growth problem is multi -faceted. Usage per subscriber is declining as the market widens but the variety of devices, and shifting user profiles as they explore a range of devices is a complicating factor, Candy says.
The type of data traffic is another. In the UK 3 has recently completed the implementation of video compression software which Candy says has “given us a significant capex saving, because it means we can do more with the current deployment.” Video represents 43 per cent of data traffic across 3’s UK network, Candy says.
Much of this kind of data traffic is consumed indoors and here Candy reiterates the importance of “denser, stronger coverage and a higher signal strength” throughout the macro network. Engineers “sleep easier at night if they’ve got their feet on the pedals,” Candy says, and can be motivated by a desire to ensure the job is done right, or by a desire to ensure the job still exists. “When you talk to technical guys you have to work out whether they’re into job preservation or whether they’re into understanding the dynamics of the business.
“You don’t necessarily get the best results if the engineers have their feet on the pedals,” he adds: “Sometimes it’s better to let the computer land the jumbo.”
This tension really comes to the fore when operators look to network sharing or outsourcing strategies; both of which have been embraced by 3 in some markets. In the UK, 3’s JV with Everything Everywhere, MBNL, owns the network used by T-Mobile, Orange and 3, but outsources its management to Ericsson.
The most popular criticism of outsourcing deals is that they were driven by short term incentives at CEO level and that they put the operator in danger of losing control over its technology roadmap. Candy, who has decades of experience at telcos, says these moves are cyclical.
“Way back before cellular, operators would have the choice to install radio equipment themselves or pay someone else to do it. The accountant would breathe down the neck of the implementation department and say ‘you’re costing an awful lot to install this equipment, I’d like to give it to someone else’.
“So you’d dismantle the installation and commissioning department and send in a bunch of contractors. After two or three years the accountant would say: ‘Look at all the money we’re paying these people! We could benefit by doing that ourselves!’ And then you’d bring it all back in house.” He sounds a modern, pragmatic note when he says that, where financial management is the motive, it is easy to forever be arguing that the alternative is better than the reality. What is more important, he says, is that operators recognise the structural changes occurring within the industry.
“Once the networks become very similar in performance and coverage and start to link together, there isn’t really a lot of difference between your network and the other guy’s network,” he says. “So it is better to concentrate your management on the products and services, the routes to market, the sales and the high street business than to devote a large portion of your organisation to optimising the network.
“You have to put your effort where it has the greatest result and, today, putting the effort into the network doesn’t necessarily produce the greatest increase in customers.”
Outsourcing, of course, places an even greater importance on the relationship between the operator and its infrastructure and services suppliers. 3 is not yet ten years old as an organisation but, even in the short time it has been in operation, there have been dramatic changes in the mobile infrastructure vendor landscape.
The arrival of “competent Chinese vendors” has made a massive difference to the market, Candy says, and to 3 specifically. The operator now has equipment from Nokia Siemens, ZTE and Huawei, as well as its managed services deal in the UK with Ericsson. The Chinese vendors, he says, simply came to market with far denser systems in terms of capacity and efficiency per kilo of kit.
“One of the GGSN platforms which was standard in Europe offered 3,000 simultaneous transactions on a floor space of about two square metres. The equivalent product coming out of China offered 30,000 on a one square metre footprint at about two thirds of the price,” he says.
Such advantages are born out of the fact that the Chinese vendors cut their teeth serving the Chinese market, and so designed their products on a huge scale from the outset. “The processors were bigger, the BUS structures inside the kit were larger, the interfaces had more capability. And it probably didn’t cost a lot more in terms of the design process. They were just built to be bigger and that had an enormous impact on the cost of the product.”
And yet Candy argues that Ericsson still has the edge over other vendors in terms of “support infrastructure and the ability to carry out a turnkey or managed service.”
He continues: “Of the Chinese vendors, Huawei are good at a full service and support the product well. ZTE has some way to go yet. So ZTE requires a bit more hand-holding and support than Huawei and Huawei requires a bit more hand-holding and support than Ericsson. But in terms of technical equipment and performance spec there’s very little between them.”
Outsourcing partners have to be able to respond quickly to be successful, Candy says. The fundamental cause of network problems that so many operators have been confronted with is that the rate of growth in data traffic was faster than the rate at which new network equipment could be commissioned and installed. Service level agreements with outsourcing providers have to take into account the need to adjust the network at pace.
But outsourcing providers should not be given the responsibility for fundamentally important decisions about network technology, deployment and architecture, he says. They do not “wear the pain” when the customer experience starts to suffer. In 2012, mobile operator CTOs are more concerned with that customer experience than ever before, Candy says. “The most difficult question a CTO gets asked by a CEO is how the customer experience can be maintained and improved—and at what cost. Technically the CTO has to understand how to fulfil that customer experience. Which is new territory, because traditionally the CTO was the man who built the engine.” There are three core areas of focus for the CTO beneath this overarching requirement, Candy says. At a strategic level, he argues, the world is moving from a telco-dominated environment to an internetdominated environment. “The CTO has to understand how to manage that change, and the types of systems and architectures that are necessary to allow it to go ahead.”
The second requirement is for the CTO to decide where these services and systems will be deployed, he says. “Will they be under his direct control, with an outsourced agent, or delegated to a cloud? This is the second series of architectural discussions and structural decisions to be made in order to achieve the objectives of the business.”
The third is perhaps the most difficult. The CTO has to have his or her own understanding of shifts in consumer behaviour, of the changes in the cloud environments that he has identified as target destinations for those consumers. The CTO has to understand “The capital and financial issues which surround that, and try and be ahead of the requirements of the market, and the product guys.” That is arguably the tallest order in the industry. Candy, who introduces himself as someone who has retired but is still doing his job, could quite easily walk away from such a responsibility.
For some reason, though, he’s sticking at it.