Month: July 2013

The departure of Microsoft’s top Xbox guy, Don Mattrick, for Zynga has raised again the profile of Microsoft’s still-secret-in-detail restructuring.  Ballmer announced that the company would be working to be more focused on services and devices and less on software.  Clearly there are going to be a lot of changes, but rather than speculating on the changes or on whether Mattrick departed because he thought he might be asked to leave later, let’s look at the fundamentals.  Does Microsoft have a shot at their goal of becoming a device/services company, and why would they want to do that anyway?

On the surface, the last question is easily answered.  PC software is what Microsoft does most, and PC software is sinking as PCs sink.  Windows and Office sales are both tied strongly to new PC sales, which are being lambasted by tablet sales.  So it makes sense to jump out of software and into services and devices, right?  Not that simple.

Devices aren’t replacements for software, they’re carriers for software.  Apple’s iPhone and iPad are at least as much software as hardware, and Android’s success as a software platform is linked to Samsung’s success as a hardware provider.  The more something looks like a gadget, the less users are willing to think about hardware and software as separate elements.  Apple knew that and created the first platform ecosystem.  But it wasn’t a slam dunk.  Had Microsoft responded to Apple aggressively, before both Apple and Google had a chance to establish commanding leads, the Microsoft model of “Wintel” would likely have been just as effective in phones and tablets as it was for PCs, and the current restructuring buzz would have never happened.

Now, with Apple and Google established, and in particular with Google already owning the “software-partners-with-hardware” model of an ecosystem, Microsoft has little choice but to create their own total platform.  They can’t compete with the incumbents through partnerships because they’ll never get the big hardware players to bet convincingly on Microsoft at this point.  No choice, in short, more than good choice.

How about “services”?  It seems to me that every company that’s been slipping in its core business, whether hardware or software, discovers the “services opportunity”.  The problem is that if you’re creating a platform ecosystem of consumer devices, it’s hard to see how services are going to come out of the process.  Does Microsoft think they’re going to be selling professional services to Windows RT users?  If Microsoft wants to be a professional services player they need to be that in the cloud computing evolution in general, and that’s going to be a challenge given their insipid positioning of Azure and the cloud overall.  The cloud as it is today is a cost-based substitution, which means that it loses money for Microsoft by reducing sales of premises software licenses while increasing the sale of hosting.  So even if there’s professional services opportunity, Microsoft has to first pay for the cloud losses with the professional services gains, then get some net benefit.

The solution to all of this, if there is one, is to create a “device cloud”, the kind of symbiosis between devices and cloud-hosted services elements that I’ve been touting for operators to create.  Microsoft cannot win the game that it’s already lost, the head-to-head war with Apple or Google.  It could win the “supercloud-device” war because both Apple and Google would face the problem of kissing off their current incumbency to address future gains.  That’s the problem that got Microsoft behind in phones and tablets to begin with, so why not force the problem down the other guy’s throat?

I think Microsoft could create this device/supercloud model easily at a technical level, probably easier than it could create a completely new structure that’s aimed at becoming a device success and also a services success.  As somebody in finance said, culture will always trump strategy.  Microsoft’s culture is too far from what Ballmer wants, particularly when there are better choices that are closer to Microsoft’s traditional model.  Could it be harder for Microsoft to see something that’s at its own feet than something on the distant mountain-top?  Or could it be that Microsoft has become such a bureaucracy that it can’t lead anymore, so it has to follow its competitors and accept perpetual marginalization?

Huawei is also a part of the equation.  Because the US Congress doesn’t trust them, they’re barred from being a big player in the US carrier space and also from enterprise equipment sales to government agencies.  That means that they’ll have to play in the US through consumer devices—like phones and tablets.  They picked up their own new executive, from Nokia, in a possible move to buttress this play.  If Huawei is really going to get into the device space, then for Microsoft to make a move now would be potentially suicidal.  The cloud is Huawei’s only weak spot, mister Ballmer.  Go for that, or face what might be a fatal level of risk.

 

Well, we’ve seemed to finally get beyond the rumor phase with NSN.  Nokia has offered to buy Siemens out of the joint venture, one of several possible outcomes that I’d heard rumors on over the last several month.  The big question is whether this will matter to NSN or to Nokia, and that depends on just what Nokia decides to do with the asset.

NSN is one of three telecom equipment giants who have suffered significantly as network operators’ revenue per bit has fallen and differentiation of equipment based on features has become more difficult.  Huawei has acted as a price catalyst for the market, particularly in the lower network layers, and NSN’s reaction to the commoditization was to shed the lines of business that were most impacted.  NSN dropped WiMax, broadband access, BSS, and IPTV and significantly cut back on its partnerships to sell transport/switching gear, so what was left was pretty much LTE.  Given that wireless broadband in general and LTE in particular has been a capex bright spot, that may have seemed a sensible approach.

Two problems emerged, though.  First, NSN jumped with both feet into a market that was itself just beginning to fall into the commoditization death spiral.  Wireless ARPU is peaking for about half the world now, and will peak and decline for pretty much all of it by the end of 2014.  By jumping into a single boat (one sinking, though slower than some others) NSN gave itself less latitude to move on to another area to compensate for LTE’s inevitable margin decline.  Second, NSN disconnected itself from most of the SDN and NFV opportunity because they don’t provide the equipment any longer.

Wireless transport is still transport, even if one considers that the current operator focus is more on the wireless space.  You can’t really differentiate much; it’s nice to talk about flexibility in the RAN and self-defining network elements, but the problem is that wireless is really about software first and metro second, and if all you have on the table is RAN and IMS you’re not in a strong position to make a case for yourself.

IMS opens perhaps the key question for wireless and for NSN, which is how to accommodate revolution in building network capabilities that have already been designed for the old model.  The Evolved Packet Core manages tunnels to cell sites to allow users to roam between sites without losing their connections.  The same thing could be accomplished with the aid of SDN, and perhaps in a more effective and simpler way.  The new IMS Multimedia Resource Function should probably be viewed as a combination of SDN and a NFV implementation of content delivery networks, all integrated with NFV-compatible session management.  My point is that while we would likely agree that these changes are needed, who’s out there talking about making them?  Logically, NSN needed to lead this initiative because NSN was doubling down on wireless bets, and they’ve not done that.

But perhaps the bigger question here is how Nokia makes a whole that’s greater than the sum of its parts.  Is it possible to leverage symbiosis between a handset business and an IMS/RAN wireless broadband business?  Sure, but why couldn’t Nokia and NSN have leveraged that when NSN was a partnership?  I don’t think Siemens would have stubbornly refused to deal with a mobile ecosystem opportunity.  The challenge was, IMHO, that NSN like most Eurogiant telecom players, is just not particularly agile and the current network market is full of real and imagined revolutions that demand an agile response.

Nokia is almost a poster-child for inertia.  Like Ericsson, they’re incredibly conservative in positioning and strategizing, and all of the dizzying movement in NFV and SDN have probably frightened a lot of managers weaned on Class 5 switching or something.  It’s pushed them further into a consolidative model of their own evolution, and you can’t consolidate without commoditizing unless you can innovate like crazy in the space you’ve carved out for yourself.  Nokia has just not been that innovative before, and frankly I don’t think that it’s likely to be that now.

Can Nokia do something compelling?  Yes.  All they have to do is to look at metro, SDN, and NFV and think outside the box to devise a new metro/mobile architecture that will jump over all the old albatross concepts of the past and into the future.  Do we need mobility management, and thus EPC, in mobile networks?  We need the former, but EPC is an implementation and not a baseline requirement.  We could have “virtual EPC” where one big virtual PGW/SGW covered the whole metro.  We could have an MRF that was one big virtual add-on to a big virtual CDN.  We could have voice supported by a virtual SBC, and of course IMS itself can be virtualized (as Metaswitch has shown).  Nokia can jump into this, make it happen, and create a win.  But NSN could have done that already, and since they didn’t you have to assume that they don’t have the thought leadership internally to cut the mustard.  If that’s true, then were would Nokia get it?