Telcos will be a force in Cloud Computing

This is a cross post from

Ben Kepes asked me to write my thoughts on Jeff Kaplan’s post Can Telcos dominate cloud computing, in which Jeff argues, that based on their history, this is 'unlikely'

 While there is nothing wrong with delivering reliable services, in today’s rapidly evolving cloud computing environment reliability is quickly becoming table-stakes as businesses of all sizes seek cloud computing services which can give them greater agility, better economies and added functionality to reduce their operating costs and strengthen their competitive positions. These are not attributes which people associate with telcos…

…Telcos are still struggling to figure out managed services, which have been around for over a decade, as a new wave of Software-as-a-Service (SaaS) and cloud computing services become mainstream.

Jeff also throws in a comment about Telcos failing it their own markets:

 In fact, even responding to new ideas and business models in their core communications business continues to be a struggle for the telcos… Now, Skype is the largest international long-distance carrier.

This is a red herring and I will tell you why later.

First let me be very clear; in this post when I talk about cloud computing, I mean utility computing, IaaS or Cloud Computing services. I am excluding PaaS & SaaS from that definition. These are different propositions and to the best of my knowledge only BT, Telstra & Telecom NZ have commercial propositions launched.

Within that definition, I do believe telcos will have a large role (maybe not dominant), here’s why;

All cloud services require a network, who better to provide you that? SLAs that are currently conspicuously absent in cloud computing will become table stakes as businesses demand greater surety from their providers (just look at the buzz when gmail goes down). Unless you can deliver the end to end service, you aren’t easily able to do this

Telcos know about reliability. I just don’t get Jeff’s point above. If reliability is table stakes then can you seriously consider an internet delivered model as being better than an integrated private network and cloud model?

Cloud Computing is only just becoming mainstream. Really, just look at the numbers. The vast majority haven’t moved. Cisco, IBM, Microsoft and HP have only just entered the market. Customers look at these guys for legitimisation of any trend. No offence to any existing provider, but most companies are risk averse and naïve about today’s market leaders. By endorsing cloud computing, these vendors will make the cloud computing market much more real to end customers.

The other critical component of these vendors entering the market is their view on Telcos as a channel. That is the vendors that are going make cloud computing real for the majority of business users are all looking to work with Telcos for local delivery. Just look at the recent announcements (Cisco, IBM Juniper), they are all partners Telcos are familiar with.

It is also my belief that Governments are going to force regionalisation to occur within the cloud market. When Governments realise what this movement to the cloud will do for their tax intake (less IT spend, combined with less employment) they will figure out ways of protecting that. This again is advantageous for Telcos as they can enforce geo boundaries and they are generally large regional players.

Telcos have access to capital in the sort of multiples most start-ups would kill for. And make no mistake; it takes lots of cash (see this for speculation on Google’s spend) to build this sort of stuff.

Convergence of data and IT means this is a logical extension of Telco services. This is important because as Jeff accurately states, Telcos are under financial pressure. The difference between Cloud Computing and tolls is incumbency. One of the reasons Skype was successful is that they disrupted the incumbent Telco business model, a business model that Telcos are still very eager to protect. That is not the case in cloud computing, all that utility hardware is an upside for them. That’s why Jeff’s comments about Skype are a red herring, completely different challenges.

I’m not clear if Telcos will dominate this industry, compared to many other types of organisation they are slow to move and aren’t that innovative. That certainly isn’t universal and I would question whether it’s altogether relevant to mainstream customers. But I do believe that when the dust settles, some Telcos will be major providers in this space.


Disclosure – I work in the ICT industry – the views in this post are my own and may, or may not be, shared by my employer

PaaS – the game is on

Back in December I wrote a piece about how Cisco, through its partial acquisition of VMWare could make some very serious inroads into the PaaS market. Then in January the NYT leaked a story about how Cisco was getting into the server market, which in my view supported my hypothesis.

Today my RSS reader is loaded with bits about how IBM with its network partner (and arch Cisco rival), Juniper is extending its Blue Cloud construct.

Techchuck does a great job of describing it here, stating that the

“[technology will allow] businesses to install hybrid public-private cloud capabilities across IBM’s 13 "Cloud Labs" spread across the world. The companies have created technology that would allows enterprises to extend their private clouds to remote servers in a secure public cloud at the click of a button. Once the technology is installed in the Cloud Labs, businesses can easily switch clients workloads when resources become constrained...” My emphasis added.

 This is a great leap forward toward much more widespread adoption of Cloud computing as it addresses several issues for corporates (not SMEs) .

These being security and control. Juniper (and Cisco) are key players in Telco private network constructs, if you could extend your private network into the Virtual data Centre (VDC), then a major performance and security hurdle is overcome. After all you want the servers performance to feel like it would if it were in your LAN / WAN. It also addresses security because all that stuff is encrypted.

The second thing is control, read this from Mercury News

While many companies say they can deliver computing power on demand, most are using relatively old technology, analysts say. They aren't able to move software programs around the planet, dialing up a data center in Dublin, Ireland, when a server farm in San Jose reaches full capacity.

IBM demonstrated it can do exactly that at an event at its Silicon Valley Lab on Monday morning, as Rahul Jain, IBM's cloud architect, clicked on a box representing a computer in San Jose and dragged it over to a panel representing a data center in Dublin.

Notice how the CUSTOMER does this, not the service provider!!  This is gold because the customer feels like they are in control (which they are) and the service provider doesn’t have to have personnel doing it.

To me this is PaaS at a level of maturity that will significantly improve adoption. This is true convergence (network, IT, and services) in action. Most importantly it targets a segment (corporates) with money (banking excluded), who experience genuine costs and are willing to pay… wow an initiative with a viable business model!

PaaS just got a new poster child, Cisco

What do Cisco, EMC and VMWare all have in common?  Virtualisation. Up until fairly recently they’ve all been playing in their own sandpits, but that’s is all changing. Cisco recently bought a chunk of VMWare from… EMC. So now you have  (arguably) the three leading vendors in virtualisation playing together. Not only in it, but tightly integrating.

Imagine one orchestration framework that could dynamically allocate you network, server and storage. That’s powerful… very very powerful. Think about why virtualisation hasn’t gotten 100% penetration? You’ve never been able to seamlessly link network bandwidth to server I/O or SAN space. Now you can, this is IaaS. 

Then extrapolate out what VMWare is doing with their View product. (Virtual Desktop Technology) and suddenly this triumvirate can deliver SaaS AS you WANT IT, and not just through a browser.  This is immensely powerful for an enterprise, it would be even more powerful for a MSP.

This isn’t just vapour ware either, reportedly Terremark are already doing this. Definitely watch this space .

Who will be the aggregators in a SaaS world

I promised myself I’d follow up on a previous post about who will disrupt old world ISV’s. Ben chipped in with some really salient comments which I’ve copied here.


Don’t forget to add to your list of companies that might disrupt as those who are in the value chain of the internet, and have every telco, isp and enterprise in the world as customers, and are all looking for more ways to gain revenues from end customers. ie the ones who actually make the internet work. Alcatel-Lucent, Ericson, Cisco,  Nokia. You might think of them as equipment vendors, but one thing is for sure, they know what is happening out there and have access to the network at a control layer that no software company will ever get”

I replied, essentially saying that there is massive potential for dis-intermediation here for Telco’s. Ben answered again with

“That’s right, and like every value chain, there are margins to be squeezed all the way. Who’s to say that the end game isn’t left with only Cisco and Google, and Alcatel and Microsoft, and Ericson and Apple as partners for example. ….. What I can’t see is actually the value that a Telco (or other aggregator for that matter) add to that equation as SaaS standards and protocols make integration of apps and networks seamless to the user” (my emphasis added)

Now I think I addressed the aggregation point with my previous post. In fairness to Ben I think he was talking about aggregation in 1.0 portal sense. Ie you just act as a gate, without reframing the content.

Ben and I dropped into and IM conversation in which we argued about this some. Ben took the position that unless Telco’s shored up their SaaS value proposition they’d get done like a dogs dinner because systems, technologies like widgets would effectively automate the value chain, thus neutralising the value that an aggregator would have. 

I argued against this. Stating things like automation cannot overcome the archaic and undocumented business rules. For example, if you look at different banks lending rules on a mortgage, you will find that 1) they are all different, 2) its not always formal, 3) the system can be circumvented if you know how. 

I also said that aggregators would be able to provide value in an increasingly complex and fragmented world because your everyday business would be overwhelmed by the choice available (if they aren’t already). And that there was real value in someone providing THE application or a shortlist which worked well for xyz vertical. 

So it was with great interest that I saw these two articles this week. The first by David Berlind is about BT’s new SaaS offerings. Clearly showing that an ISP has a play in SaaS. ISP’s are actually in a pretty good place to get into this type of play for a lot of reasons. Apart from the ability to bill, build platforms etc. One other reason has leapt to me over the last week. All ISP’s offer email. And given aggregation is about getting users back to a site again and again so you can offer them more, is there a better way do this than through an application you use 20 times a day? UPDATE; I just picked up this feed about Yahoo acquiring Zimbra. (Looks like the words out in the ISP community…. SaaS is in )

The second, (a bit more abstract) discusses how Google *MAY* buy some US mobile spectrum. This would be classic Telco dis-intermediation. They won’t have the content, they won’t have the identity or billing, and they won’t have the access. Google would become both the ISP and the SaaS provider. A really interesting play, and until I saw this from David Berlind it was out of context for me,

"Perhaps the other question raised by providing services at the onramp (as opposed to along the highway) is to what extent this move by BT and similar moves like it by other ISPs may force Google to engage more deeply in the onramp (ISP) business."

Other aggregation plays exist too. MS Live is one companies attempt to do this on their own. Not many companies have Microsoft’s resources at their disposal so I don’t see tis being that common unless you get an opensource equivalent movement going. 

Salesforce and perhaps Facebook with their Platform as a Service plays could do this if they could insinuate themselves into the value chain and appease the vendors / offer them enough value to do so.

Another aggregation play would be for someone to do the work for a specific vertical. Health, retail, legal etc. They all have some basic needs (horizontal) which are common like HR etc. But then you could as a smaller player do the aggregation for them. This lends itself heavily to the web 2.0 user generated content play because you could then get a community of recommenders and word of mouth working for you. 

Finally, and looping back to Ben's comments. The people that provide the fabric of our communications infrastructure, could partner with ISV's and content providers directly. These guys have potentially more chance to do this than you're average Telco / ISP. This is because of a couple of reasons. They ARE the onramp that David mentions above. Secondly, because just about every Telco / ISP on the globe has followed the global mantra to cut costs by outsourcing. All of the smarts that a Telco think's it owns, aren't really theirs at all….they belong to the likes of Cisco and Alcatel Lucent.  As Ben said, Telco's really need to shore up their value or they are goneburger…..