Policy Control & Policy Makers: New Settlement Needed?

Money is piling into the policy management sector, which is seeing a boom in business, vendors, and technology development. Policy engines are getting far more sophisticated and powerful, taking advantage of advances in computing, servers, and software to increase performance on metrics such as transactions per second by an order of magnitude or more. Still relatively small, it's a sector that could get a lot bigger very quickly.

Innovations in technology don't always determine market outcomes. Some technologies fail to catch fire. But with policy vendors reporting high double-digit or even triple-digit growth, no one can doubt the current enthusiasm among network operators. As one big telco put it to me: "Policy is going to be really big for us. We have to get beyond gigabyte pricing, and in the next 12 to 18 months, we will make much greater use of policy to differentiate ourselves."

And that means there's a good chance that more policies will be applied (or offered) by network operators to more end users than anything we've seen hitherto. In a recent survey, we found that the average telco with a policy architecture in place anticipates over the next three years adding a new policy every six to eight weeks. Policies on how services are delivered will be shaped by a growing list of triggers, including a subscriber's tier, location, device, network type, and stated preferences, real-time congestion information right down to cell level, time of day, type of application, URL, and so on (and on).

From the point of view of operators, the name of the game is competitive differentiation, not end-user control. And increasingly, the policy levers will be placed in the hands of end users themselves.

All of this raises a big policy issue for regulators: What should they allow network operators that enjoy significant market power to do with these powerful new platforms? If policy platforms enable new policies to be written and deployed within days, and dozens of policies are routinely deployed in the average network, how are regulators to maintain effective control?

This debate has to date been largely framed in terms of net neutrality, and the Federal Communications Commission (FCC) 's proposed rulemaking on the topic. Yet in some ways, net neutrality feels something of a creation myth that is getting in the way of a more nuanced debate. What seems to be missing from public discussion is any sense that the current and emerging generation of tools might actually benefit end users as much as network operators – especially if those tools allow users to control the way that policies are applied in order to save money or improve the performance of services they especially value.

As I've argued previously, the key requirements for regulators in this area are to ensure that there is effective competition, as well as transparency in the way that policies are applied, so that end users know what policies are being applied to their connection. Beyond that, the rule should be to let the market determine what policies work best – for both network operators and their customers.

— Graham Finnie, Chief Analyst, Heavy Reading

Justyna.Bak 12/5/2012 | 4:25:27 PM
re: Policy Control & Policy Makers: New Settlement Needed?

Great post, Graham. I fully agree that QoS is the new way for operators to differentiate and move beyond flat rate data tariffs.

I'd like to share a few words on the impact on end-user experience.

Network neutrality regulation will not change the fact that the radio interface of mobile networks is subject to physical constraints. If mobile data traffic grows 1800% per year [1], it is impossible for the operator to guarantee high throughput and low latency to all users and applications at all times and locations. Luckily, there is a solution to that challenge.

Introducing policy control to mobile networks is a win-win situation for operators and end-users. With policy control operators are able to prioritize delay sensitive traffic, such as video streaming, above traffic, where delay will have no impact on end-user experience, such as file downloading. Keeping subscribers happy is the most important driver for operators to deploy policy control. In addition, they can also improve profitability of their MBB business by introducing MBB packages where QoS level comes as the third parameter on top of speed and quota. Premium quality MBB package can e.g. cost twice the price of the standard one.

[1] http://www.telegraph.co.uk/technology/mobile-phones/6726623/Mobile-phone-networks-face-crisis-as-data-traffic-surges.html


cnwedit 12/5/2012 | 4:25:26 PM
re: Policy Control & Policy Makers: New Settlement Needed?

Well put, Graham. As I think you have noted in the past - and as I certainly have - there is a significant gap between how the telecom industry views policy and how consumer groups and others pushing the FCC to implement Net Neutrality rules view policy.

That gap is being perpetuated by the telecom industry's inability to adequately educate regulators as to how policy can be implemented in ways that will absolutely benefit consumers without violating their privacy rights, blocking competitors' traffic or stifling innovation on the Internet.

Too often, policy is seen simply as control and many people are loathe to cede control to large ISPs and wireless operators, even if it results in better services that make more efficient use of the infrastructure. Witness the reaction to the Google-Verizon proposal.

It would take a major initiative to both educate the regulators and consumer groups as to the very real benefits of policy and convince them such controls will be used for good, not evil. From what I've seen so far, ISPs would prefer to continue deploying this technology without discussing it publicly, much less mounting any real education effort.

PeterWThompson 12/5/2012 | 4:24:38 PM
re: Policy Control & Policy Makers: New Settlement Needed?

Operators have done themselves no favours by making their first 'policies' ones of denial rather than enablement. Seeing the issue as 'two much P2P traffic' rather than 'poor performance for other services' has set the tone for the subsequent debate. Provided (and this is a big 'if') network QoS mechanisms work effectively then there's no harm in allowing links to be filled by P2P traffic, as long as other services are appropriately prioritised to perform well.

More on this: http://bit.ly/aPQCdP

Sign In