In When the time comes to upgrade IT equipment and systems, is it better to repair what you have-patching and tweaking systems and software to keep things running-or replace the old technology entirely and start with a clean slate? That was the dilemma facing Accenture CIO Frank Modruson.
I recently faced a decision familiar to any homeowner. My furnace was nearing the end of its useful life, and my choices were clear: replace a few critical parts at a modest cost in the hope of extending its service for a few more years, or replace the entire system at a considerably greater cost.
As I examined these options, a key consideration was the fact that an entirely new heating system would allow me to take advantage of the latest and greatest advances in home heating technology and efficiency. That would be impossible to do if I were just going to refurbish the current furnace.
As mundane as this decision was, my furnace choice exemplifies a dilemma familiar to many CIOs. When the time comes to upgrade IT equipment and systems, is it better to repair what you have-patching and tweaking systems and software to keep things running-or replace the old technology entirely and start with a clean technology slate?
Of course, the IT "tweak or trash" debate is far more complicated than replacing a furnace. IT's operational imperative to ensure uptime and avoid downtime at all costs creates a natural bias toward risk avoidance. A philosophy of "it works, so don't touch it" nurtures the growth of hybrid IT environments in which multiple systems coexist, documentation and qualified programmers are hard to find, and it appears safer and cheaper to make tiny patches rather than risk transformational change.
Over the last decade, we in Accenture faced a series of decisions more momentous than, but not unlike, my furnace call. Looking back to the year 2000, when we were essentially a new company (having recently gone public), we had embarked on an accidental experiment-even though we didn't realize it then.
We inherited legacy systems from our former parent and needed to build separate technology capabilities. We had no idea that Accenture would today have more than 244,000 employees or revenues of $25.5 billion. But we had ambitious growth strategies and knew we needed the IT infrastructure to support that growth. While we did not plan it in advance, our efforts over the last decade to create a world-class IT organization effectively became an accidental test of the tweak-versus-trash debate.
Set a Coherent Strategy
Our first task was to set a coherent IT strategy and build an independent capability. Then we concentrated on running IT just as we would any other business, with a focus on efficiency. We centralized, rationalized and standardized, cutting our global applications from 600 down to 267 and local apps from 1,500 to 255.
We migrated from multiple country-specific platforms to a single-instance global ERP. When we came up against the tweak-versus-trash decision, we opted for trashing what we had, mainly because innovations in IT kept making the newest available solutions much more powerful than any legacy systems.
By the middle of the decade, we had our IT shop in order and were in a position to tackle even bigger changes. These included a complete network transformation, which gave us the bandwidth to build out one of the world's largest high-definition video conferencing networks, as well as enabling us to introduce an entire suite of powerful collaboration tools. These tools have already had-and will continue to have-an enormous effect on our evolution as a virtual enterprise.
Take video conferencing as a case in point. I recently thought about how my business travel had changed over time since we started using video conferencing to replace travel. My sense was that I was traveling less, so I took a look at my mileage records for the past five years.
Although my air miles for client meetings had remained roughly constant, miles traveled for Accenture management meetings had been reduced 35 percent because we now do so many of our internal meetings via video conference. I benefit, my family benefits and Accenture benefits.
As our universe of 45,000 Webcam-enabled Accenture professionals expands and we continue to bridge our video network with client systems, we expect these benefits to multiply.
No Free Lunch
Just as my brand-new furnace took a bite out of my wallet, none of the changes we made in Accenture's IT came free. We invested heavily, with some investments aimed at reduc-ing costs and others at adding new capabilities.
When we started, we did not consciously set a policy of replacing everything, but we did not want to be encumbered by old technology because we knew that conflicts between systems impose massive burdens on performance. Looking back over 10 years of work, we discovered that we consistently chose to jettison the old in favor of the new, and when we were done, we had changed absolutely everything.
Each of the changes we made took time and entailed more than a little disruption. For example, we moved to our global SAP platform in a single big-bang implementation. The downside risks were obvious, but so was the upside potential: We got to a single version of the "truth" across our entire global enterprise much faster with this approach.
Our internal customers are pleased: The percentage of "satisfied sponsors"-the senior executives with whom we work very closely on each of our IT initiatives-rose from 67 to 92 percent between 2001 and 2010.
Clear lessons emerged from our accidental experiment. Roughly one billion dollars in cumulative investments resulted in three billion dollars in savings. Thus, for every dollar Accenture made in IT investments over these years, we realized three dollars in savings-a very impressive return.
Our IT operation became dramatically more efficient over the period: IT spend in total dropped 22 percent, and the IT expense as a percentage of Accenture's net revenue fell by 59 percent. IT expense per employee dropped by 70 percent, even though we made massive investments over this period to equip our professionals with the most advanced technology tools.
Defining High Performers
Our experience is consistent with the latest findings from Accenture's High Performance IT research, which looked at 226 of the world's largest private- and public-sector organizations in North America, Latin America, Europe and the Asia-Pacific region. We define high performers in IT as those that achieve excellence in three distinct areas-IT execution, agility and innovation-balancing the constant and sometimes opposing demands placed on today's IT function.
High performers in IT manage information technology like a business, and run the technology operation for and with the business being served. CIOs in these organizations are deeply engaged in mapping their company's business strategies and then envisioning how IT can best support those strategies.
When we conducted detailed assessments of these IT organizations, we discovered that they spend 29 percent more annually on developing and implementing new applications, rather than spending more on older applications that were delivering less. They are 44 percent more likely to recognize the strategic role IT plays in increasing customer satisfaction, and they were eight times more likely than peer IT functions to measure the benefits realized from IT initiatives.
So when it comes to tweak versus trash, the results from our 10-year billion-dollar experiment are conclusive. The faster you leverage the latest technology, the sooner you benefit from it. The longer you retain a legacy technology, the more it drags down performance, and the higher the price you pay in the innovations that are sacrificed.
Looking back over all the decisions made over the last decade, we would not have changed anything, except this: Everything we did, we would have done even faster.
Frank B. Modruson has been the CIO of Accenture since 2000. He oversees all business applications and technology infrastructure, enabling employees to serve clients in 120 countries and work anytime, anywhere.
Add new comment