How to Reduce IT Demand by 50%
One of my favorite Druckerisms is, “I have no interest in someone who plays the minute waltz in 56 seconds.” Drucker goes on to explain that, “In terms of technology, we have people trying to play it in 56 seconds when it shouldn't be played at all. Very little of our computing capacity is well used.”
When it comes to IT, this quote hits home, big time. Do a random sample of the typical IT request queue and at least 50% of the requests there would bore Mr. Drucker senseless. In my experience, 30% of the IT requests aren’t worth the effort and 20%-30% can be accommodated by leveraging existing systems.
The problem with the bad, 56-second type, IT requests is that they slow down the good ones.
Imagine a road with a tollgate. The tollgate represents the IT investment governance process (including logging, defining, estimating, justifying, and evaluating). All requests must go through the tollgate before entering the road on their way to their final destination (or delivery, in our case). The bad ideas slow down the good ones by competing for governance airtime and creating congestion that slows down delivery overall. It doesn’t matter how well tuned the IT supply processes are; too much IT demand has the same impact on progress and innovation as quitting time does on my ability to get home when I’m driving in Los Angeles traffic.
Many a senior executive team has tried, and failed, to reduce IT demand. Problem is, there are too many requests requiring too much detailed knowledge to manage IT demand from above. IT congestion can be reduced -- but only if every driver does their part:
1) Evaluate the idea with the cold, cruel eye of a VC. Ask if the idea directly supports the business strategy and creates tangible business value (e.g., speed up orders, improve customer retention rate, etc.). If the answer to either of these questions is no, make the idea more interesting by tying it to the multiplier in your business.
2) Check to see if the capability already exists. Only about 20% of systems' functionality is frequently used (and an estimated 45% is never used), so it’s highly likely that the systems can do more than you think they can. It may be necessary to develop some new skills -- for example, using the data extract and reporting tools -- or modify your business process, but it’s much faster and cheaper to do it yourself than wait in the IT request queue.
3) Ensure that you are ready to devote the necessary resources. A project manager and analytical expertise will need to be assigned -- from your organization -- to define requirements, redesign processes, perform testing, and manage change.
4) Verify that the idea is as good as you think. Test or pilot the concept using the available tools (e.g., using a combination of current systems supplemented with Excel and some manual effort). This will not only validate the business value, but highlight any people and process issues that need to be addressed.
McKinsey was right to say, “The problem is that IT governance systems have become a substitute for real leadership." Leaders at all levels should perform the bulk of IT prioritization on their own -- before the formal governance process kicks in.
Sign up for the Harvard Business Publishing Weekly Hotlist, a new weekly email roundup featuring the top highlights from HarvardBusiness.org.
- Comments (4)
- Join the Discussion
- Email/Share
Comments
Here at the Town of Castle Rock, I've take the last paragraph to heart. IT Governance shouldn't be a substitute for a CIO. When I designed our governance process at the Town, I created it to help establish communication about the various tech projects across very diverse departments. The water department needs to be aware that the police want to install a massive new system and IT won't be able to just jump on other major projects they decide to initiate. With all the departments learning about the major initiatives for the following year, they start to collaborate on timing.
In addition, I use the process to gain better visibility for tech spending during the budget process. Before this, technology spending requests came in through many different channels to the budget process. Now, all major tech budget requests come in to the executive team as a clean, clear, prioritized list. The projects with the best alignment to the organizational goals and mission, with the lowest risk and with the best financial results are at the top.
So does the Town need me if the IT Governance process is working so well? Yes, because before the list even goes to the IT Governance team, I spend countless hours developing the initial list, researching the details of the project, talking to executive sponsors, and developing the initial prioritized ranking. I then educate the team on the projects and my recommendations based on my detailed prioritization process. Inevitably the team will make a few changes, but they always come away from the process with strong support.
Kevin Capp, CTO, Town of Castle Rock
- Posted by Kevin Capp
May 30, 2008 4:04 PM
Many companies get in ERP (Enterprise Resource Planning) solutions merely as a replacement for existing systems. i.e. the effort and expenditure to put in an ERP solution is to replicate processes that the existing system handles rather than re-engineer processes and get in change management to improve productivity.
Hence your statement that 20% – 30% of IT requests can be accommodated by leveraging existing systems becomes even more ironical in the light of the situation where the entire existing system is replaced with a similar set of processes in the new ERP system.
This is often because of the following reasons:
1. Many of the methodologies for implementation of ERP solutions focus on quick implementation rather than getting more out of the implementation. Hence, there is no time for re-engineering processes, change management
2. ERP solutions are perceived as cloaks of respectability for the organization’s IT framework as it implies better controls, tighter security. It is more about getting an ERP ‘tag’ such as an ISO tag rather than getting into business transformation.
3. Both, the ERP implementation partner and the CIO view an ERP solution from a narrow IT programming and hardware perspective rather than from a holistic business consulting perspective.
4. The activity of evaluating the task from a VC’s perspective, as you put it is rarely, if ever done. Not many implementations are preceded by evaluation of key business processes and metrics for measuring the success of the implementation. And even if they do, this is quickly forgotten when the implementation process reaches a feverish pitch where everyone wants to somehow make the system ‘live’ and call it a success and go home.
- Posted by Prashant Swadi
May 31, 2008 12:36 PM
I do agree with the fact that very little computing power is actually used to drive ones business needs. Adopting the four ideas presented makes sense only when an idea is tested for its feasibility given the variables in a firm. Alignment of IT inputs to complement a business strategy can be designed with extra care but what is often missed is the following,
1. Diligence required to maintain the systems (based on the
concepts)
Plans are scrutinized, evaluated, re-evaluated and then
implemented. IT resources are well used to drive ones
business, but after a technology shift (not a business model
transition) it is assumed that systems are legacy. Existing
systems are now ready for an overhaul.
Instead of making plans to change the existing setup, these
systems can be re-factored. System administrators can be
summoned to reload the data and optimize the implementation
(Change the middle ware and firmware not the actual systems).
Money is saved, knowledge base is available and user training
is not required
2. Longevity of the idea - Looking into the future
Ideas end up as educated decisions, but do they cater for the
future trends in technology, consumers, and strategies ?
I have seen companies using an ERP solution successfully for
years until a sales rep shows up at the door. If an existing
implementation is well-thought and mitigated for all the
future changes, there should be no need to make any further
changes.
IT demands also increase if we haven't looked enough into the future and make changes to the existing systems as it would affect other interfacing entities.
- Posted by Venkat Mahalingam
June 1, 2008 8:57 AM
Susan,
Your title intrigued me. Reducing IT demand by 50% is pretty compelling. I have always preached that big investment in IT should produce significant change, not incremental change. Big investment needs to enable business strategies and real value for buying customers.
I like your metaphor about the toll booth, but I am not sure how that reduces IT demand. Seems that it just isn’t letting some cars onto the road. Without getting needs recognized, demand isn’t reduced. Instead the solution is minimized and put into an EXCEL solution. Shadow systems in EXCEL are turning knowledge workers into mini-data centers. Productivity is lost trying to supply the pent up demand without IT. I shudder at the mission-critical ACCESS systems I have seen on desktops.
Rather than keeping cars off the road, we need to find a way to divert them to the exact change booth. The exact change booth let’s cars on the road, creates a gate for standards, quality and payment, but doesn’t required IT support each time. We only need the toll booth, some collection/banking technologies and a maintenance team.
Susan, we both know that incremental change is about business process. Every time an EXCEL spreadsheet is added to a process, the process gets more complex. If we had a toll gate for the times when a business process exceeds its acceptable resource (time, personnel) threshold, the payment is approval of a past-threshold business process. You can see where I am going. Map the critical business processes and set thresholds. IT supplies the mapping technologies; business supplies the maps and thresholds.
This is a win-win for IT and business units. The big investment is the mapping; control over processes can be strategic.
Ellen Bonnell
ellen.bonnell@trendsavants.com
www.trendsavants.com
- Posted by Ellen Bonnell
July 3, 2008 3:05 PM