Sunday, March 15, 2015

Look Ahead With IBM To 2018

Look Ahead With IBM To 2018
Published: March 16, 2015
by Timothy Prickett Morgan
The world today is a much different place from the one that the AS/400 was launched into nearly three decades ago. Back then, it was exotic to be moving from basic accounting systems running on batch oriented machines to real-time transaction processing and database querying using relational database technology. It was also a big deal to be moving away from homegrown software to packaged applications tailored to specific industries. Software is far more complex and so are the systems it runs across, and IBM reflects the complexity of the market it serves.
Ever since it nearly went bankrupt in the early 1990s, IBM has been obsessed--many would say correctly--with continual self-transformation. The company did far too much navel-gazing back in the late 1980s and early 1990s when the AS/400 business was launched and growing and an important part of the revenue and profit picture for Big Blue, and IBM missed many opportunities.
If anything, IBM is perhaps a little too hair trigger to dump a business and chase the hottest new thing through acquisitions and a little bit of research and engineering internally, but IBM does not have anything even close to a monopoly in any part of the IT business these days (excepting mainframes and IBM i platforms, if you want to have a very tight market definition) and it has to chase growth where it can find it.
The question you have to ask is how IBM is aligning with your business and how you might be aligning with IBM's business. IBM has been giving Wall Street some insight into where it will be in 2018, thirty years after the AS/400 was launched and something that feels more like a millennium in IT time.




By now, you all know the mantra of the markets that IBM is pursuing with its "strategic initiatives."
Come on, chant it along with me as we could all have perhaps done at IBM's recent Investor Briefing event for Wall Street analysts. Let's say it together: Cloud, Data, Mobile & Social, and Security. (IBM actually pronounces it with an ampersand, which is a neat trick.)
These strategic businesses represented about $25 billion in revenues in 2014, comprising 27 percent of the company's revenues from the year, up from 13 percent in 2010.
"That $25 billion is a big business, even for IBM," Schroeter explained. "You have to have a B at the end to be relevant, but this $25 billion is absolutely relevant for us and for our clients."
Schroeter noted that the mix of that the mix of software in the revenues for these strategic imperative markets was around 50 percent, which is much greater than the percentage of software revenues for IBM at large. "So we have growth above the level of the market and the ability to scale even by our standards and the ability to craft a solution that is high value." IBM's growth rates in these combined social, mobile, security, big data, and cloud segments were roughly double the industry average from 2010 through 2014, too, so IBM is pretty sure it is running its business right.
Hardware, Schroeter noted, was about 10 percent of that strategic imperative pie in 2014, and services was around 40 percent.




To make his point, Schroeter said that eight years ago, IBM identified analytics as a key growth area, and now that business generates $17 billion in sales for the company. (He did not say what the baseline was eight years ago or how much money IBM had spent acquiring assets like Cognos, SPSS, Netezza and many others to buy a big chunk of that revenue growth.) IBM's hardware take for its analytics biz is about 10 percent, software is about 55 percent, and services is around 35 percent. This market is pegged at a $315 billion opportunity, and growing at around 7 percent per year between now and 7 percent. IBM's growth in the analytics business is keeping pace at 7 percent growth last year.
When it comes to cloud, there is no hardware sale at all, with the $7 billion cloud business last year being around 30 percent software and around 70 percent services. (There is plenty of hardware cost in this business, mind you.) IBM's software-as-a-service products, of which there are over 100 these days, have an annual run rate of about $3.5 billion, and are one reason why the cloud business at Big Blue is growing at 60 percent year-on-year. The cloud market opportunity that IBM is in love with has around a $400 billion total addressable market in 2018, and is growing at 23 percent compounded annually between now and then.
The engagement business--which includes mobile at $1 billion in sales, social at around $1 billion, and security at more than $1.5 billion--grew at more than 35 percent in 2014. At this pace, IBM will gain market share in this engagement business, which it pegs at $290 billion in 2018 and growing at 10 percent across the industry.
Looking out ahead, after making its forecasts, IBM is telling Wall Street that it can push $40 billion in revenues in these strategic initiatives by 2018, which will represent more than 40 percent of its total revenues. That is another way of saying that IBM will grow to be an approximately $100 billion company again by 2018, which is not huge revenue growth, but the strategic initiatives will fill in the gaps. Ginny Rometty, IBM's somewhat beleaguered chairman and CEO, has told Wall Street that scaling up the SoftLayer cloud and the applications that run on it will actually boost its earnings, but IBM has not said much about how profitable these other areas are within its strategic imperatives.
The interesting tidbit that Schroeter revealed in his presentation at the Investor Briefing was that 80 percent of IBM's customers buy all three of the elements of its product portfolio--hardware, software, and services.




Schroeter talked up how the mainframe continues to be strategic in core industries like banking, and he talked about how Citibank uses IBM System z and Power Systems iron to run virtually all of its applications, and that the bank is also a big user of its systems software and middleware. And to show the relationship between services and the Power Systems business, he explained that IBM Power Systems machines are behind about 550 million subscribers in the telecommunications industry in India, and amazingly, that all of this Power iron is delivered through a set of strategic outsourcing agreements with the big telcos in that country. (Why IBM has not mentioned this before is a bit of a mystery.) By the way, those Power machines are supporting about half of the telco subscribers in the entire country, which are mostly using mobile phones supplied by Bharti Airtel, Vodaphone India, and Idea Cellular.
Looking ahead, IBM believes that its "annuity like" software revenue stream--meaning subscriptions for software and support contracts--accounts for 70 percent of its revenues per year. IBM can count on it more or less like clockwork so long as customers don't shift their systems quickly. (Which very few enterprises can.) That annuity-like software business is growing at 3 percent a year, but the SaaS portion is growing at more like 70 percent. Larger customers are not doing software transactions the way they used to do in the past; they want subscriptions.
"For hardware, we have been saying for a while that our model is stable and profitable," explained Schroeter. "We have made tremendous, tremendous progress in turning what was a $1.7 billion decline in profit in Systems and Technology Group in 12 months, and we made a profit in the fourth quarter and we see pretty good growth coming out of that portfolio in 2015." IBM now has Power8 out for the first full quarter and the new System z13 mainframes ramping, too. As for storage, more of the value is shifting to software and away from hardware, both Rometty and Schroeter said in their presentations.
While this is all technically true, it is also a bit off the mark. Functions that are now sold as separate items in software are packaged as separate software features in storage arrays, and frankly, this is a change in packaging that is meant to aggrandize Software Group and hurts Systems and Technology Group. The same is true of the past two decades of Global Services taking a slice of what is really systems revenue.
If IBM wanted to show the real strength of the System z and Power Systems businesses, it would talk about them as a complete stack, with revenues for these platforms marked as such and software and services sales for other company's systems being kept in a separate set of buckets. I don't care about IBM's books, per se, but when IBM's executives and customers are making decisions based on the categories IBM chooses when it classifies its revenues and profits, it starts to matter. IBM is a systems company--80 percent of its customers buy all of its segments--and its bookkeeping and presentations should reflect that. Only then can the company, its customers, and its investors make sound decisions. My guess is that if IBM did that, it would show that System z and Power Systems sales have been subsidizing sales for software and services on other platforms, and that would raise all kinds of hell.




As far as IBM i customers are concerned, the message coming out of IBM is that the Power Systems business is being transformed such that it is all "high value" and has an intellectual property component, too, thanks to the OpenPower Foundation, which is opening up the Power systems stack to take on the hegemony of the X86 platform in the datacenter.
I will be attending the OpenPower Summit, the very first one, in San Jose this week, and I will be listening very carefully to how the Power ecosystem is expanding after being opened up. I will also be pressing IBM and its partners to open it up in ways that benefit the IBM i platform directly, rather than indirectly just by the strengthening and widening of the Power platform outside of Big Blue. OpenPower has been limited to Linux thus far, and I think IBM i and AIX should be brought into the fold. There is no technical reason why AIX and IBM i can't be ported to a bare metal OpenPower machine or a virtual one running OpenKVM.
Maybe what IBM needs to do is make a new AS/400. I will ponder that one my trip out to San Jose.