IBM's Software Czar On Smart Grids

[admin]

Focus shifts to quick payback

AustralianIT | Stuart Kennedy | May 19, 2009

THE M in IBM might stand for machines but these days the IT giant is more about software and services than boxes with blinking lights.

IBM's software czar Steve Mills believes there is a vast new opportunity opening up in building smart grids

Software makes up more than 20 per cent of IBM's revenue and more than 40 per cent of its profit and Steve Mills who runs IBM's software division presides over a behemoth that stands shoulder to shoulder with Oracle and SAP.

But with the global financial crisis puncturing business confidence, and with it many of the grandiose enterprise IT deals that form Mills' usual happy hunting ground, the software business has had to quickly and dramatically adjust the way it operates.

Rather than selling grand, long-term business transformation projects, Mills, known as "The General" inside his division, has his captains and lieutenants flogging short-term, quick-payback IT exercises.

However, as the world's governments pour billions into nation building fiscal stimulus schemes, Mills has set his sights on a whole new game for IBM's middleware software, analytics packages and telemetry expertise.

He believes there is a vast new opportunity opening up in building the smart grids and sensor arrays that will eventually encircle the planet and spit out endless waves of data for analysis and automated action by IBM's middleware and analytics software.

We are all in the grip of the global financial crisis right now and seeing bedrock US companies like Chrysler filing for bankruptcy. From your perspective what's happening to business IT plans in the wake of the GFC?

You see a couple of common behaviours. One is a shift to payback-based analysis. The extent to which companies had traditional techniques for doing return on investment and internal rate of return analysis, those methodologies, although they haven't been thrown away have become subservient to the payback analysis.

When does this investment give a full payback for the money we put in.

With budget money so scarce, as a CIO I am trying to operate in a mode where everything I invest in provides a payback in the short term.

Ideally the payback is in the current budget period.

If you are going in for multi-year payback timeframes you can't win those arguments in many companies today. They want to talk current year or single year types of investment.

Doesn't that make it hard for you selling SOA (service-oriented architecture) projects where it can take years to get a decent return on investment?

It makes the projects smaller.

You get two classes of projects. The IT industry has the ability to self fund investment through infrastructure elimination because of the tens of millions of servers deployed and all the different devices there are plenty of ways to leverage consolidation and therefore displacement of existing infrastructure.

Previously you had 1000 underloaded servers. Let us show you how to take 1000 underloaded servers down to 20 or 30 heavily loaded servers. That's a dramatic reduction in cost.

There are some very big projects under way focused on server consolidation, data compression, getting rid of storage devices and improving network efficiency.

These things tend to justify themselves in the short term so they are the hot projects today that are the easiest to get done and in many cases the CIO doesn't have to anything more than demonstrate the returns to get approval.

On the business transformation side the projects are somewhat more challenging but they are still being invested in. Some companies take a forward view and say we are going to invest today with the idea of making ourselves more competitive so that when things turn we are going to gain a lot of market share.

That philosophy is out there, not everywhere, but it is out there.

But less so than 18 months ago?

Clearly it is not a dominant philosophy. The dominant philosophy is save money now. The secondary ideas that are out there relate to strategic investments -- better analytics, better business understanding and business dashboarding and these smart planet projects.

Many of these have short-term payback, they can be surprisingly economical because you find that with a little bit more data you can save a lot of money on maintenance around physical assets.

Anything that provides relief on asset operations and maintenance often justifies itself.

You have said the business opportunity around wiring up the planet's infrastructure could be dramatically larger than any of the analysts' estimates so far. Could you elaborate on that?

It's a lot of large numbers. We have been accumulating those assets for decades and decades. Typically power companies are running everything from recent to fifty year old assets. The electrical wires strung across towns and cities have been there for decades. There's efficiency opportunities in maximising that investment where you get some proportional IT value tied to the asset value itself.

No one knows what these numbers are but you can implicitly begin to talk about numbers in the forty trillion, fifty trillion dollar range. Massive numbers as far as the world's infrastructure and its relative replacement value goes.

Knowing you can't get capital dollars to replace it all, what is the incremental investment that's justified to make it more efficient and in making it more efficient can I demonstrate a near term payback? Not in capex, but in opex. An example in the energy arena is the work we have done with Centrepoint Energy in the US where doing grid instrumentation reduced the number of trucks that they had on the road every doing repair work. Because they were getting more data they were analysing the state of their grid and were out doing preventative maintenance and they cut down the number of outages which had been driving up their overtime cost and they also cut down the number of trucks and repair crews they needed to have at the ready.

That pattern repeats itself over and over again because physical infrastructure is extremely maintenance intensive and the surrounding operating expenses are startlingly high for most businesses.

People are waking up to this phenomenon and now with the financial crisis it's just adding more impetus and IT plays a key part in it.

Maintenance is not typically an IT bucket and that's what provides the expansion for the IT opportunity -- it's freeing up other monies and translating them into IT.

How have you changed the software group's sales and marketing approach in view of the economic crisis?

Payback, payback, payback and more payback -- 30 days, 60 days, 90 days. You keep honing your analysis and equipping the field force with more references and examples and you try to help them point the customer towards the benefits of IT investment and that its not one of these things that has a long term benefit that no one will sign up for.

Don't you find that hard to sell sometimes. There's a lot of arcane terminology in IT that sounds very theoretical. SOA for a start sounds hard to sell to a board especially as a short-term payback?

Yeah. The chief executive doesn't really know what SOA means so you have to translate these things back to a business value view based on things the company can do better if it spends some money on IT.

Which aspects of cloud computing do you think are over hyped and which ones are real?

I'm not sure you could say there is anything about cloud that is not over hyped because you don't even know what you are talking about half the time.

What's real is this idea that its possible to do request driven provision and scheduling of IT systems.

If you think about any IT environment the amount of hardware there at any point in time is fixed. So is it utilised or is under utilised? Most companies when they analyse this would say they are under utilised or dramatically under utilised and enormously inefficient.

We need to pool resources and move the work around to the equipment as opposed to dedicating equipment for every piece of work.

There's tremendous financial motivation around this flexible model for application deployment. It's at the heart of every cloud environment -- being able to flexibly and seamlessly move things I and out.

Not everything lends itself to that. If I am a financial services company I need to make sure my trading systems are absolutely bullet proof and capable of handling peak load requirements so I need to be very careful how dynamic I make that environment because I might get caught in a short traffic burst when the markets go crazy. That would be a business disaster.

Many of your products are being enabled for cloud services like Amazon.com's online infrastructure offerings. How is that changing the way you charge for software and services?

We have offerings at IBM that are on-demand style. We have a very large monthly lease business at IBM. We are used to selling things to people on monthly charges.

Having had decades experience doing that we know how to implement those models.

If I sell you something upfront I get a much bigger chunk of change in the beginning if sell you things as you use them I'm not going to get as much money.

It has always had somewhat of a dampening effect on our growth rates versus many of our competitors but at the same time its certain, you get paid and it tends to be very sticky so we like it.

I'm comfortable with it and it's something we'd like to do more of.

Your group has been a frequent and forceful acquirer over the years with an acquisition tally so far of 75 companies. Is the global financial crisis throwing up once in a generation acquisition opportunities for the software division?

It's probably not once in a generation because the dot com bust drove valuations downward for tech companies. This time valuations have been driven down broadly in the market.

That makes it more interesting perhaps for acquiring certain business however in an environment where the market is not so robust the upside leveraging potential is nowhere near as strong as it is with a strong economy.

We are a GAAP accounting company and we buy things for synergy and for revenue lift not just to simply acquire for the sake of acquiring.

Prices are lower but the lift is not as big because the market is not as robust.

During the last downturn there was a nice inflexion point where the valuations where still at their relative low point but the market was starting to come back.

There will be the same inflexion point here but we are not there yet.

Is your acquisition targeting strategy still one of leaving the big customer relationship management and enterprise resource planning business application markets alone and stalking more middleware and business intelligence targets like Cognos?

It hasn't changed. There are things in our portfolio that have application like characteristics but we are very careful not to be competing with the core ecosystem providers from which we derive a huge amount of revenue. For every dollar customers spend with SAP they spend five dollars on surrounding technologies and services. Would you rather play for the one dollar or the five dollars?

From a software point of view were you disappointed to see the Sun Microsystems acquisition possibility pass by and Oracle pick up the company?

No. Sun is a hardware company when you look at that $US12 billion ($16 billion)-plus revenue its hardware based.

Sun has not been a very formidable software competitor, obviously Oracle is. Sun has been giving away software. It's easy to give away software its hard to put a price on it.

Sun has been giving software away in support of their hardware business and now Oracle is in the hardware business.

We know a lot about the hardware business. We have forgotten more about the hardware business than I believe Oracle will ever know.

We wish them the very best of luck.

Stuart Kennedy attended IBM Impact 2009 in Las Vegas as a guest of IBM