Tuesday, November 9, 2010

Gartner Symposium, Opening Keynote

After Andy Kyte's presentation on application overhaul it was time for the opening keynote, that was scheduled in the afternoon this time.

There are four big trends that change the IT industry as we speak. One is cloud, which is all about elastic capacity. Two is social computing, and the real impact will be more around ethics and culture than it is about technology. Three is context aware computing, which will be much more than location-based services on devices. Computers will also be able to detect language, emotions, desires and respond accordingly. Four is what Gartner has trademarked as Pattern-Based Strategy.

Pattern-based strategy is not based on traditional thinking in terms of 3-5 year strategic plans and large programmes, but one of continuous change. This type of thinking is not new. In an interview in McKinseyQuarterly  a few years ago, prof. Richard Rumelt already remarked that sustained competitive advantage is less likely to come from big dramatic shifts and more from continuously spotting tiny changes in the market and surfing that wave with a little bit more skill than anyone else.  In fact, I was part of a big research project in 2007 and 2008, in which we dubbed this "management excellence", as a next strategy after operational excellence. Gartner describes three steps to the strategy: sense, model, adapt. We talked about the three pillars of management excellence: being smart, agile and aligned. Sense and model would together be "smart", and adapting equals the agility we meant. However, alignment is equally important, because if your part of the organization (or your organization as part of the value chain) is the only one to adapt quickly, it leads to suboptimal results. It is necessary to be able to roll out changes to the complete value chain.

Gartner also predicted the end of the effectiveness of the acquisition strategy of the mega software vendors. It is indeed possible to acquire innovation, but it is not possible to sustain acquired innovation. In the words of Gartner: "No architectural mediocracy will be accepted". In other words, integration is not enough.

Then, three other concepts were introduced:
·      Smart control
·      Intelligent business
·      IT Dynamism

Smart control means managing IT by loosening grip. The old command and control style of managing IT cannot be successful anymore. It is not possible anymore to prescribe corporate IT solutions. Business users can get their own functionality in the cloud, or deploy their own departmental solutions. It is needed to engage users, and enable to do things themselves. For instance, allow them to define their own business rules and allow them to compose, decompose and recompose their own business rules.

Intelligent business is what comes after "business intelligence". BI has mostly been popular among managers and their staff. But in today's business, everyone needs insight. Business processes become increasingly knowledge-intensive. I think this is what drives pattern-based strategy. Not data warehouses, analytics and BI tools, but sensing change through business processes. After all, it is through your business processes that you interact with customers.

The IT dynamism trend entails not building systems that last a lifetime. From built to last to built to change. Interesting enough, this seemed to be the opposite of what Andy Kyte argued just a few hours before. We need to learn how to build systems without knowing all requirements. However, not all systems are equal (and this is what reconciles the keynote and Andy's presentation). There are three types of systems.
·      Systems of record, like a general ledger, do not undergo a lot of change. These can be standardized
·      Systems of differentiation need to be built for change. Likely they don't come from integrated software suites, but organizations should consider best-of-breed vendors here.
·      Systems of innovation, today typically consisting of internet mashups, and social media

Other topics were discussed as well, but at least this is what I took out of the presentation.

Monday, November 8, 2010

Gartner Symposium, Andy Kyte on Application Overhaul

Andy started by describing the current state of things. He pointed out that today's core applications do things they weren't designed to do. That despite all the application consolidation having been done, and budget cuts, there is still a bloated IT portfolio, fitting together in an "accidental architecture". As Andy said: "please don't tell me that how your IT landscape looks like, is by design".

And now comes the hard part. Most of the consolidation that took place was in infrastructure, where no one cares. But now you are touching people's application...

But something has to be done. The life cycle portfolio of applications in most organizations doesn't look like a neat bell curve. Stuff is still wheeled in all the time, in a disproportionate pace, while there is another large part of applications that are zombies, the living dead. They should've been eliminated a long time ago. 50% of all applications in existence today should be eliminated.

Organizations have an IT debt, Andy argues. IT debt is the deferred maintenance on their applications, because of the IT budget cuts of the last years. Gartner estimates the world-wide IT debt to be about $500 billion, and growing to $1 trillion.

In the meantime, IT people are in denial, and business people again start to avoid IT, and create their own solutions, running their own applications in the cloud.

How to take control? Money is not the constraint, the real constraint is an organization's ability to change... and if organization's don't improve their ability to change, cost and risk while go up, and the agility will go down even more. And the more combinatorial complexity (system interconnections) there is, the less agility you'll have.

When you replace a core system, it should be build for the next 30 years again, it is simply too invasive to the organization to switch them out every few years.  So design them for the 21st century. Don't automate the old business process with a more fashionable tool.

Andy also hit one of my hobby horses: option-based strategies. He argues that business cases are fairy tales, we don't really know what will happen. Build for unknown future change.

Lastly Andy describes the need for an Annual Report on Business Applications. He talke about it in terms of portfolio management, but as I have been arguing before in my own research, you can think of this report in a really financial way as well. IT truly is an asset. And the more fluid, like cash, the assets are (= agility), the more people are willing to pay for it. Also, the more fluid they are, the less you need to write off on them.

The most memorable quote (at least for me) in the presentation was: "It is NEVER  about application consolidation, it is ALWAYS about business process simplication".