Friday, December 5, 2008

EA & SOA in Down Economy

Many articles, discussions, and presentations have been made on how the recent economic downturn affects the companies’ EA and SOA efforts. Some, like Joe McKendrick of ZDNet, believe that slow economy can spell boon for SOA but subprime SOA will suffer. Dave Linthicum published the results of his survey indicating the same. Beth Gold-Bernstein of ebizQ suggests that we bet the farm on Complex Event Processing. Ronald Schmelzer of ZapThink advocates that companies should continue to invest in SOA in a down economy. I would go a step further. EA and SOA investment is critical to the companies’ success – it will enable them to stay competitive, achieve significant efficiencies, and potentially even gain market share.

Pretty bold statement, some would say. I don’t think so. Let’s consider the facts.

When the times are tough, the first thing most companies do is slash budgets. IT budget gets reduced just like everyone else’s. The focus shifts from the strategic initiatives to simply keeping the lights on and completing projects as quickly as possible. Enterprise Architecture efforts are usually the first ones to be eliminated or significantly reduced. Point solutions become the norm resulting in duplication of software, hardware, and overall efforts. Smokestack applications rise up from the ashes of the Enterprise Architecture. Everyone becomes more concerned about keeping their jobs rather than doing the right thing for the company. IT managers shift to the aggressive empire building mode in order to protect their jobs and eliminate their own risks. (The old mentality of “I own more than you, therefore I am more important than you” is still alive and well, unfortunately. IT managers also think that if they can “own” and control every piece of their application, it will reduce their risk and allow them to deliver results faster.) Governance becomes unenforceable and largely forgotten.

Through this chaos, interesting trends emerge. While the initial IT budget is reduced through a series of staff reductions and some technology rationalization efforts, the costs begin to creep back up in subsequent years. When the economy finally turns around and the pressure to keep the budget low eases, the IT budget suddenly becomes larger than what it was prior to the cuts. Why? The explanation is simple. The empire building and unfettered decision making by IT management finally bears fruit. There are more software, licenses, hardware, and code in the data center, all of which requires more people to support. There is very little reuse and sharing because each group has built silo applications residing on their own unique platforms. Costs increase, efficiencies decrease, and it takes longer to deliver new capabilities especially if they require several applications to integrate with each other.

Enterprise Architecture and SOA can help reverse these trends and, in fact, keep the IT budgets low. Most companies have a number of redundant systems, applications, and capabilities that have grown through the type of uncontrolled behavior described above. EA, through an effective discovery and governance mechanisms, can eliminate these redundancies while maintaining the same capacity and level of operational responsiveness. Additionally, EA groups can influence or implement new architecture approaches to help consolidate resources and gain efficiencies. Examples of this could be virtualization, green technologies, cloud computing, etc. SOA, as a subset of EA, provides much the same benefits. Encapsulating key business functions as reusable services will help achieve more consistency, save money, and enable faster project delivery. An effective EA program can protect companies’ IT budgets from ballooning by establishing and enforcing standards, promoting reuse opportunities, and ensuring transparency across all IT systems.

The bottom line is that companies can not afford not to invest in EA and SOA. These programs will make organizations more efficient through the economic downturn and help achieve the necessary savings. On the long run, EA and SOA will keep the costs down while increasing business agility. Effective EA and SOA programs are a competitive advantage, not an overhead. They will easily pay for themselves and, what’s more important, enable organizations to avoid uncontrolled spending in the future. Enterprise Architecture and SOA is a must, not an option!


HRHead said...

It all really comes down to the business strategy for getting through the recession. If the business chooses to invest in reducing their cost of doing business so that when it comes out it can deliver more product at a lower cost, then SOA/EA is a logical place to put your money.

If the decision is simply to reduce over all run costs and wade through for a while, then the will won't be there.

This type of down turn is the exact right time to make a SOA move though. The typically powerful business voice is humbled during times of poor sales. When business is booming its much more difficult to implement SOA-like changes. Its an oxymoron that at the times when we are rife with cash, rather than invest it to reduce the cost of doing busines, which increases profit, the cash itself is used by the business as a shield to brush such efforts off. It becomes an "if it ain't broke don't fix it" scenario.

Leo Shuster said...

Excellent comments! I definitely agree with your statement that during the down economy, the business has less of a leverage. It adds further to my argument that EA and SOA investments should only increase during the bad economic cycles.

Mukund Narasimhan said...

I totally agree, in a down economy
the decesion for spending is being moved to the CFO's rather than CIO's. The CFO's want to look at plugging costs and new product development is where the axe falls. Pretty obvious considering that it could have an impact on you P&L statements.

I work for a S/W consulting firm and this is what our customers had to say.

The decesion clearly lies with how the CIO wants to control his costs. A good option would be to outsource the "Business as Usual functions" where you would find a a large cost benefit.
I would rather call it impact sourcing. If the cost to keep my systems running were X, I would get that done in .80X. I would use the residual money for new development initiatives. This could be a viable option considering exisiting systems maintenance eats more into your IT spending budget.


Leo Shuster said...


Outsourcing, as any other federated work, requires strong governance, which, in turn, needs to be supported by a strong EA organization. Thus, if CIO wants cut costs via "impact sourcing", he needs to ensure that the mechanisms to control this work exist. Otherwise, you will end up spending more money trying to fix the problems introduced by unmitigated outsourced activities.

Mukund Narasimhan said...

Yes!. A part of my job included building a framework to address these issues. It systematically irons out issues that you would face in the traditional outsourcing gambit. The journey to outsourcing should move context to core. I am not saying EA should be outsourced rather low risk IT functions like Manual Testing, Documentation should be outsourced.I can send you the white paper I wrote on this subject.