x
Policy + charging

Bridging the Chasm[s]: Easier Said Than Done

It's possible to get a sense of weary déjà vu when analyzing the difficulties facing conventional telcos in the Web era. Year after year, in surveys and interviews, the same unsolved problems keep popping up.

Take Bridging the Chasm, for example. Some readers will recall that, back in 2011, Light Reading ran a series of articles with that title highlighting the gulf separating Networks and IT inside major telcos. (See Bridging the Chasm: A Manifesto and Why Bridging the Chasm Is Critical for Carriers.)

We'd like to able to say that the chasm is diminishing. In reality it is -- slowly.

But as we've found in several surveys during the past year or two, the gap between Networks and IT is only part of the picture. Just as challenging -- and in some situations, more problematic still -- is the gulf separating Networks and Product Marketing.

In a survey just completed on policy management, for instance, we found that this gap is, if anything, getting worse: The internal barrier dividing the Networks and Product Marketing teams is one of the biggest barriers to policy realizing its potential, survey respondents told us.

One reason for that is that the gap has simply become more visible, and that in itself tells its own story. When policy was used largely to manage traffic, there was little reason to consult the marketing folks. Now, with most operators saying that policy management is a strategically vital platform for differentiating services, the marketing people must be involved.

But poor channels of communications between Networks and Product Marketing at many telcos, surveys tell us, stall those efforts. The organization struggles to make the business case for further investment, and policy tools languish underutilized.

That's just in one area. As the Web and mainstream IT technologies continue to eat away at the conventional dividing lines between internal telco functions, those chasms will get a whole lot more visible, not least as telcos grapple with the opportunities and threats presented by network functions virtualization (NFV) and software-defined networking (SDN).

Indeed, in a separate survey we conducted earlier this year specifically on SDN and NFV, telcos themselves acknowledged that organizational resistance and inertia was one of the biggest barriers to deploying virtualization and SDN.

New technologies derived from the Web and IT present more and more opportunities to telcos, but they make life more than a little uncomfortable for those clinging on to the old ways of doing things. One frustrated telco strategist told us, half-jokingly, that SDN would be much easier to adopt once the current generation of 50-something technical executives is pensioned off.

For many, the threat is all too real: If analytics, policy control, and related tools enable telcos to automate the ways that service offers are bundled for customers, where does that leave marketing? If NFV means that technical staff specialized in managing network gear are no longer needed, where do they go?

None of this is to say no progress has been made. At the most senior levels, most boards understand what needs to be done, and are pushing through change. But making sure that this happens all the way through the organization is an enormous and sometimes under-appreciated task. Technology can only take you so far. To take things all the way, some difficult and wrenching organizational changes are required. Are big telcos with tens of thousands of employees really ready to make them?

— Graham Finnie, Chief Analyst, Heavy Reading

brookseven 9/18/2013 | 11:26:39 AM
I really think there is more to it than you think...  

So, in my mind if I am on the operational side of a telco I need to use a hypocratic-like oath around these things...First Do No Harm!  I think that is where changing network architecture and controls really struggles.

If I were putting a presentation together for a telco - on the operations side - I would start with a slide that started with...

"Nothing a customer can do can completely disrupt their service nor can they impact the service of any other customer using the network."

I think that is where closing any chasm's/gaps/holes needs to start and where folks are missing the boat. 

seven

 
gfinnie 9/18/2013 | 11:44:17 AM
Re: I really think there is more to it than you think... Absolutely, and I certainly wouldn't want to minimize the dangers here. In fact our survey earlier this year with operators on SDN/NFV found that the whole OSS area was seen as one of the biggest obstacles (for some, THE biggest obstacle) to deploying virtualized network functions. Several operators told us they would only deploy NFV if they could leave the existing OSS intact.

At the same time, some operators (sometimes the same operators) see it as a big opportunity to automate and reduce the amount of manual operational tasks.

There are plenty of reasons for telcos to be cautious on the operational side, given how much they have to lose when things go wrong.  But it's a double-edged sword—if that conservatism slows down or stalls service innovation, it leaves telcos vulnerable to a toxic mix of high opex that's hard to cut along with flat or declining service revenues—something that many European telcos, at least, are already struggling to deal with.  
Carol Wilson 9/18/2013 | 12:02:28 PM
Re: I really think there is more to it than you think... So Graham, if moving too fast is risky and moving too slowly is toxic - what's just right? 
gfinnie 9/18/2013 | 12:18:46 PM
Re: I really think there is more to it than you think... I don't think that's a question that can be answered in a message board post!

Looking on the brighter side, there's plenty of innovative work going on in the ETSI NFV group (NFV Management & Orchestration), and the NGMN Alliance and elsewhere to try to resolve the impasse. I highly recommend Caroline Chappell's new report on this, "Managing the Virtualized Network: How SDN & NFV Will Change OSS" which provides a huge amount of the detail here on some of the innovations going on -- but Caroline also notes that "Vendor "big picture" management architectures that aim to harmonize OSS, SDN and NFV are works in progress that may take years to complete."

Hopefully we can air some of these issues at the upcoming Ethernet & SDN Expo in NY.
Gabriel Brown 9/19/2013 | 5:46:12 AM
Re: I really think there is more to it than you think... brookseven, I think you're spot on with the operational issue. We get that feedback all the time. I wrote a bit about here: 5 Challenges for NFV in the 4G Core Network

But also, to some extent, operators are hamstrung by their own procedures and requirements.
brookseven 9/19/2013 | 11:03:45 AM
Re: I really think there is more to it than you think... All,

I think its all a trade off.

Graham...let me ask you a question.  How many operators are losing money?  I mean actually cash flow negative?  How many of those are the large Tier 1 operators?

I guess my view is that the operators are doing great....just not as great as Apple and their kvetching is basically jealousy.

My second question for all is this...Instead of adding all kinds of smarts to the network, imagine what the change would be if they instead made it dumb...dumb as a stump...and fast.  Could they lower their costs by eliminating lots of smart guys?  Would they be better off simple, fast and dumb or complex, slower, and really smart? (They all claim the latter...but I wonder.)

My personal view is fast and dumb is the way to go...but thats just me.

I think the 3rd question is more of a series of questions that you can use to get an idea of where folks heads are at:

- To get (whatever your favorite network change - NFV, SDN, whatever - from here on out called X) implemented, would you be willing to change out your entire Routing infrastructure?  By that mean EVERY router you have?

- Do you view X as a cost reduction mechanism or a revenue improvement mechanism?  How much will the cost or revenue will be changed?

- Will you use X to retire some other network or function?

- About how many $ will be used to deply X before you can see new revenue or cost reduction?

I think that is enough for now.

seven

 
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE