First off I want to say how much I enjoy EMC Global Marketing CTO Chuck Hollis’s blog. Crazy job title, really smart guy, its one of the few tech blogs I keep up with these days [kudos to EMC comms professional Matthew Buckley - his weekly email roundup of blogs always catches my attention].

Chuck says insightful things like:

“Internal cloud is just a user revolt saying to IT ‘can you just stop playing with knobs and give me what I need now”

Right On. But I felt I had to call him out on the notion that telcos and traditional service providers have “structural advantages” in cloud provisioning. In a post from EMCWorld this week entitled EMC Takes Atmos Storage OnLine he argued:

Any successful storage as a service offering has to meet two criteria; (1) it has to be cheaper or easier than doing it yourself, and (2) it has to offer certain well-defined and pre-agreed standards for performance, availability and security.

Unless you can do #1, you don’t have a market.

And unless you are pretty clear about #2, no one will trust you with their information.

The Service Provider Advantage?

Everyone points to Amazon’s S3 as the leading example of storage as a service. I’d agree that they do a decent job of #1 above, but as far as #2 — well, let’s just say that not everyone is convinced yet.

Now, consider for a moment the economic advantages that a service provider — such as AT&T — brings to the table. They have plenty of data center space in all sorts of useful locations. Those facilities are already built and already paid for. They have big pipes — wired and wireless — that can deliver information when and where it’s needed. Those pipes are largely already built and paid for.

They already have business relationships and trusted provider status with enterprises large and small. Those relationships are already built and paid for.

And they have sophisticated infrastructure and process in place to manage things like service delivery, security, provisioning, billing and so on. Most of that is already in place.

So, when people ask me about newer-style storage cloud providers (Amazon as an example) being viable in the marketplace long term, I just have to point to all the inherent structural advantages favoring the telcos and service providers in this market.

“Inherent structural advantages” has a really nice ring to it, but I think you could use the argument to drain the water off pasta. Lets consider some history.

We have been expecting telcos to make a substantive and successful play for the IT market since long before I came into the business 15 years ago. Remember “convergence”? The idea was that telecoms companies would use their inherent structural advantages to wipe out IT suppliers. Didn’t happen did it?

In fact if we look at the numbers- the latest telecoms company to make The Big Play has certainly suffered for it. Under Ben Verwaayen, ex-CEO, BT “went big” on Global Services. New CEO Ian Livingstone is now cleaning house. Costs were out of control, and the “synergies” and “structural advantages” weren’t realised. As Computer Weekly reported a couple of weeks ago:

BT is restructuring its struggling Global Services business to focus on three separate areas after its “unacceptable performance” led to the group losing more than £1.3bn.

My question for Chuck is basically- can you name a single example of a major telecoms company that is a successful IT service providers Now try and name some that have failed in that ambition…

Inherent structural advantages. Like the music business perhaps in the face of digital music? Or the mobile communications business perhaps?

What did Apple and the iPhone do to existing players with their “structural advantages”? Apple doesn’t rely on telco billing relationships – it uses them to subsidise the hardware for customers, then builds its own billing engine and relationships through the AppStore. Who would argue its wireless network partners got the better end of the deal?

My colleague Michael Coté talks of the “shackles of success” – and telcos have those in spades.

Lets consider AT&T, which Chuck says has “plenty of data center space in all sorts of useful locations”. Perhaps – but how modern are these resources? According to IBM

70% of data centers were built before 1985, and will need rebuilding (via @merv).

Now I don’t play tennis, but I would not be too quick to call “advantage AT&T”. Not when the competition will come from companies like Amazon, Microsoft and Google. Who builds the most efficient data centers- Google or AT&T? I’d be interested to know what you think Chuck?

Does EMC really want to bet against Amazon CTO Werner Vogels? Have you seen the progress they are making rolling out new functions? This week came Monitoring, Auto Scaling and Elastic Load Balancing for Amazon EC2. I am pretty sure IBM wouldn’t be encouraging customers to deploy production workloads to Amazon if the environment couldn’t be trusted.

Lets think about security for a minute – the commonplace anti web company canard.

How many serious breaches has salesforce.com suffered from since the company was formed 10 years ago? As far as I know the answer is… zero. How many banks, with vast security budgets, can claim that track record? How many customer identities have we seen lost by public sector and financial services incumbents in that time? Oops I lost my disk. Oops I left my laptop on the train. Oops who was looking after that tape again? Many of the organisations that have lost these customer details are major EMC customers. That didn’t protect them. Poor security is poor security whichever side of the firewall its on. There are no inherent security advantages in being an existing telco or service provider.

In the 1970s the FUD was that computing bureaus were insecure. Plus ca change.

Nick Carr has already said all this stuff far more eloquently than I ever could. One of my top Cloud thinkers is Simon Wardley from Canonical, the commercial arm of Ubuntu. I credit him with the photo I used to illustrate this post. Who is it? Joseph Schumpeter – the guy that coined the term “creative destruction”.

EMC is a really strong data management company. No doubt about it. It has an awesome arsenal of “enterprise-class” products. EMC also deserves great credit for realising the threat Amazon storage services face earlier than most – but I would advise the firm to be very cautious about an assumption that your channels to market have inherent structural advantages.

disclosure: IBM is a client. Not so Amazon or EMC.

Share This

IBM Impact, Big Blue’s annual service oriented architecture (SOA) shindig for customers and partners held in Las Vegas this week, was a model of Big Company corporate messaging. The broad story was pretty technical – but with IBM’s Smart Planet agenda at work – there were also plenty of broad business and societal perspectives to think about.

Gerry Mooney is the guy tasked with managing IBM efforts around the US economic stimulus plan. In his main tent keynote session on Tuesday he used a phrase that got under my skin.

“We need to create a legacy for the 21st century.”

Reclaiming “Legacy”

Think about it for a second. Are we finally set to reclaim the word “legacy” from connotations of failure? The question has a personal relevance. When I was a reporter in the 1990s my news editor used to jokingly refer to me as “legacy boy”. I had the mainframe beat, and so had to cover all the companies and technologies like Boole and Babbage, job scheduling, MPE, SSA, SNA, VAX, virtualisation, VMS and so on. Anything “legacy” was handed over to me to cover. No problem. Of course – now, 15 years after I came into the business most of the cool new companies my colleagues at the time covered have long receded into history, sucked into the long maw of tech industry mergers and acquisitions. Not so IBM. Not so IBM’s mainframe, z.

Over time I grew to realise that “legacy boy” was a badge of honour, not an insult. What is more, the mainframe has contributed a lot of dollars to my coffers over the years. Not bad for something “dead”. As the current recession has kicked in the mainframe continues to win new customers and new workloads in a way that not even IBM mainframe executives truly believed 10 years ago.

Legacy technology always seemed an overly negative term. After all, if I had a great aunt Maud that left me a legacy I would be darned happy about it. But legacy has been unambiguously used as a pejorative term in the tech industry. Feeble reframing efforts emerged such as “heritage” technology. For someone that loves language as much as I do however, the idea of legacy meaning “bad” always seemed dissonant.

The Legacy Turning Point: Stimulus

Well perhaps the the legacy turning point has been reached. A legacy is a good thing, one that should not be squandered. IBM’s Mooney, in his speech on Smart Planet and economic stimulus, squarely made the point that after the Great Depression huge capital expenditure projects left America with a positive legacy of new dams, parks, roads and essential public infrastructure that stood the country in good stead throughout the 20th century.

We mustn’t waste the money we’re (printing and) throwing around today. Whether or not you agree with the details of the current economic stimulus package, it’s clear that much of America’s infrastructure is creaking. Road bridges, for example, are in an absolutely parlous state, as are dykes, as Katrina showed us all too clearly. The legacy of the New Deal is evidently running out, so its surely time to build a new one.

Food on The Table In Tough Times

IBM is a company that is built to last. America can learn some valuable lessons from one of its greatest industrial champions. At this point I will note that IBM is RedMonk’s biggest client. The outstanding performance of the firm in this excruciatingly tough financial environment has been, I must admit, somewhat of a relief. The way I see it RedMonk is feeding 3 kids now, with the arrival of Tom Raftery at the firm last year, with his two lovely little boys, joining my own son as a RedMonk dependent, so we need to keep the revenues coming in. Our man in Austin, Coté, and his lovely wife Kim are planning to bring us another mouth to feed in the near future.

I take my responsibilities as a shareholder very seriously indeed – we all need to pay the bills, and put food on the table for our families. At RedMonk we don’t hire people for a year, we hire them for the long term – which is one reason we grow so darned slowly. The strategy has held us in good stead though, and I thank my business partner Stephen for keeping his feet on the brakes over the years. Some other analyst firms had been using rolling lines of credit to make payroll- bad idea. Since November 2002 when we founded RedMonk we have never borrowed a dollar. Of course being a people services business we don’t have significant capital outlay to worry about; we are certainly not capital intensive – but Stephen still makes an awesome risk averse chief operations officer.

Smart Planet, Stuff That Matters

IBM may be cautious, it may take its time to turn in a new direction, it may sometimes seem a little out of touch, but no bloody wonder- its building a legacy. If you were Sam Palmisano, the company’s current CEO, you wouldn’t take unnecessary risks either. Built to last – indeed.

With over a hundred years of history IBM has been here before, and it seems to be making the right moves in the face of some brutal challenges. Other firms have fallen apart under less stress.

Its time for all of us to work our collective asses off to dig ourselves out of a hole, and i make no apology if that sounds socially minded. The recession is just one manifestation of some of the short-sightedness wreaking havoc with so many environments around the world.  Its time to build some alternative approaches. Tim O’Reilly – one of the guys in this business I most admire, has been beating a new drum lately. Rather than talking about Web 2.0 his new mantra is Work On Stuff That Matters.

Tim and IBM’s core narratives are evolving in similar ways. Just as is IBM moving from a technical discussion, SOA, to a societal one, Smart Planet, Tim is also talking more about really big issues and less about bits and bytes.

Proud to Be Legacy Boy

I am prouder than ever to be “legacy boy”, and its nice to know that Mr Mooney is doing such a fine rebranding job for the term. We all have a role to play. Now, more than ever, we need to Work On Stuff That Matters. Ignore Everybody. Get Excited and Make Things.

For too long we have confused wealth creation with increases in debt. Our environmental credit card debt grows by the day. This tiny dot of blue and green is fragile. Legacy is not a bad thing, its the very best thing. It is something we leave to those that come after us.

related posts: here and here.

special thanks to Matt Jones, Dopplr’s design genius, for the Get Excited and Make Things image.

Share This

I recently posted about Tivoli Pulse, but it seems I also had some other notes in the hopper. For example – I found these facts cited by IBM pretty interesting. I think you may too.

• Over 78% of data centers were built before dotcom era
• Data centers use 10-30x more energy per square foot than office space
• For every 100 units of energy piped into a data center, only about 3 units are used for actual computing- more than half is cooling.
• in distributed computing environments up to 85% of computing capacity can sit idle

With those facts in mind, Tivoli has

a. integrated energy management into all of its console and service management tooling, enabling chargeback and accountability for energy throughout IT. This information can be used for carbon credits or utility rebates in certain geographies.

b. signed deals with building management firms like Johnson Controls and Siemens, to ensure information about energy consumption in facilities can be correlated with IT usage.

All of this stuff is driving IBM’s Dynamic Infrastructure story. Cote was at an event with that team this week, and cites some relevant numbers in his regular roundup, so I might as well include them too!

• 70% of data centers were built before 1985, and will need rebuilding (via @merv).
• Waste makes opportunity: “85% idle compute capacity. $40B of supply chain wasted. 70 cents per dollar just to run the systems. 1.5x data doubles every 18 months.” (Jim Stallings’ slide)
• Of IBM’s 194 datacenters worldwide (just under 6M sq ft), IBM’s own operations are run in just 7 (via @jeunice)

Automation and Virtualisation are the key technologies available today to improve data center efficiency. Sometime in the future Cloud Computing will be another potential avenue of improvement.

disclosure: IBM is a client. The numbers here were supplied by the firm.

Share This

When Dion Hinchliffe pointed to an article this morning – Killing Innovation with Corner Cases and Consensus – I assumed I knew what it was about, and wanted to read more.

However upon reading the post I realised Steve Blank was describing a different phenomenon than the one I had in mind.

As we were plotting marketing strategy, I mentioned that the phrase “Speaking to the Big Dogs” might end up as his corporate brand.  And that he might want to think about aligning all his video and Internet products under that name.

We were happily brainstorming when one of his managers spoke up and said, “Well, the phrase ‘Big Dogs’ might not work because it might not translate well in our Mexican and Spanish markets.”  Hmm, that’s a fair comment, I thought, surprised they even had international locations. “How big are your Mexican and Spanish markets,” I asked? “Well, we’re not in those markets today… but we might be some day.”  I took a deep breath and asked, “Ok, if you were, what percentage of your sales do you think these markets would be in 5 years?”   “I guess less than 5%,” was the answer.

You could say this tendency is the marketing equivalent of Donald Knuth’s “premature optimisation“.

No – what I had in mind, was the tendency of “innovation capture” by deep pocketed customers, where the needs of a tiny minority of customers dictate overall product generation, skewing the overall direction of innovation.  A good example of the tendency in action is The Java application server wars, where BEA WebLogic and IBM WebSphere fought a death match throughout the late 1990s and into this decade for the high endian needs of a handful of customers, notably Charles Schwab and eBay. The thing is though- the really big dogs have different needs from the rest of the market. They are edge cases. Enterprise Java app servers became something for the high end, for organisations with huge developer teams with samurai skills. Of course BEA and IBM’s focus on serving the needs of the Big Dogs created opportunities for other players- such as JBoss, and latterly Glassfish and dynamic scripting languages and associated servers such as Zend.

My point is that requirements in tech follow power laws – and a small number of customers are worth a lot. IBM is currently doing a great job of demonstrating just how much…

But as I have recently been arguing, Amazon is doing fantastically well for pretty much the opposite reason – it keeps things simple and delivers functionality for the many, rather than the few. Of course I am also the guy that said a while ago IBM should acquire Amazon (Had it done so Big Blue’s share price would likely have hit the magic $140 number that keeps its execs up at night, but that’s a different story).

Innovation happens at the “low end” not just the “high end”. If RedMonk had a dollar for every time we heard from a Big Vendor that a new technology “won’t scale”.

In summary, while the high end is obviously a great place to make money, it should not be the sole significant focus of engineering efforts and R&D dollars. In the long run volume wins. Sometimes its better to run with the pack. I also want to point out that innovation capture hurts both parties – it keeps costs high for customers. It is part of the pathological condition of enterprise sales, where vendors and customers are locked into a high margin dance marathon… The dog and pony show is exhausting for both parties.

Disclosure: IBM is a customer

picture courtesy of smoorenberg on flickr, under an attribution 2.0 license.

Share This