About Us | Login | Follow CITO Research:

The Crude Hammer of Enterprise IT

Written by Stefan Dietrich | February 11, 2013 | 0 comments

On a recent flight from Germany back to the US, I sat next to a retired automobile industry consultant. He spent most of his life implementing the Japanese principles of Kaizen and lean production for German automakers to optimize production speed, minimize cost, and improve quality.


The Kaizen Way

With Kaizen, there are no lengthy analysis and decision processes; it’s a continuous process of small improvements. Changes—often proposed by workers themselves—are evaluated, and if accepted, implemented immediately. What needs to change is changed, right on the spot.

Experts are hired to go through the assembly lines to observe what is happening and compare it with what they have seen at other assembly lines and studied in best-practices research. They focus on specific bottlenecks at specific plants and how they can be improved. If a shelf is too high or parts are organized in the wrong way, they decide that the shelf must be lower and the parts reorganized. And no, this is not done on paper; it has to be implemented by tomorrow.

Too Many Hammers

The fellow I met on the plane told me that after the government bailout of the US automakers in 2009, he was called back from recent retirement to help with the restructuring of American automakers. To get an overview of what was wrong with the competitiveness of the production facilities, he started with a walk through the assembly lines and noticed an unusual number of hammers being used. He said it seemed like every other part being installed needed at least some gentle pressure to fall in place and he started to investigate some particularly worrisome examples of interior trim pieces.

It was soon determined that this was caused by slight irregularities during parts manufacturing. However, nobody really cared as everybody assumed that hammering the part into place was acceptable. In fact, the initial quality of cars was the highest ever, and tight fits meant no rattling or squeaking. It also meant that over the years, with summer and winter temperatures expanding and contracting them, the parts would eventually become loose, rattle, and although not dangerous, not compare to the precision other automakers put into their cars to keep them rattle free for 10 or more years.

Addressing Business Processes in Silos

Let’s look at a direct parallel in enterprise IT. The transition to mobile devices requires the implementation of incremental changes to existing legacy platforms to allow integration. At the same time, these tweaks mean that the overall production process itself has changed.

For example, the fact that support for mobile devices is mostly provided through the handset manufacturers means that enterprise help desks now have to address a different set of questions. First they need to understand which questions they should handle and which to refer to the handset manufacturer. For the questions they need to handle, they need to be able to support devices that are probably personally owned.

Like the automobile industry, enterprise IT departments have a tendency to address business processes in silos, optimizing them, but not addressing how they fit together in the bigger picture. A few hammers may be necessary in the interest of expediency, but how many should an enterprise IT department deploy?

The Hammers of Enterprise IT

Hammers address symptoms but ignore underlying problems. Here is a sampling of IT hammers I’ve thought of so far.

The nightly build. This so-called best practice used by many Java-based development shops is actually a remedy for the lack of a complete model for module dependencies. If you could tell which modules depended on which, you could recompile only what you needed to when you are checking in changes.

Buying more and more storage. The amount of storage at data centers is growing rapidly, sometimes for good reason. Often, though, it’s a mystery what data is being stored, how it is being used, and what archive processes make sense. It is easier to write a check for more storage than to classify data and implement appropriate data lifecycle policies to make better use of existing storage.

Adding network bandwidth. As network bandwidth becomes cheaper, bandwidth can be added year after year at no increase in cost. However, adding bandwidth without analysis eliminates potential cost savings from using less network capacity.

Bolting on more security. Adding firewalls to control traffic between machines (default settings intact) and requiring specific ports to be opened on a machine-to-machine basis are just two common practices that complicate the security architecture. The architecture becomes overly complex, making it difficult to resolve security dependencies between applications and network infrastructure.

Using VPNs to access an internal network. Sometimes VPNs are needed, but if access to a few specific services, like email, intranet, or file servers, is all that is needed, VPNs create more problems than they solve. VPN connectivity opens up complete connectivity between networks, and while providing centralized authentication, they introduce risk of unintentional access to data and then require elaborate mechanisms for access control and monitoring.

Blind loyalty to Microsoft. IT departments often commit to continue to buy Microsoft solutions even though non-Microsoft solutions have become feasible for a majority of users in a company and provide a better user experience.

Doing it all in-house. IT departments continue to create internal solutions for commoditized services, such as email or simple computing or storage tasks, that can often be handled better through an outsourcing agreement with integrators or using a cloud-based solution for much of the user population. By working on undifferentiating activities, IT can’t strategically help the business.

Keeping it local and more expensive. Some local versions of services are better than what is offered globally. Shift to global solutions that are perhaps slightly less flexible, but provide better overall savings for the company.

Experts with an angle. The practice of hiring a vendor’s consultants as experts to solve strategic questions means that the answers are necessarily colored by the vendor’s perspective. Strategic questions should be addressed by in-house staff.

Overpaying IT staff. Too often, existing IT staff are no longer up-to-date on technology and stay because they are not receiving better offers. In addition to the financial drain, they too often try to make themselves irreplaceable by implementing complexities that only they understand, which flies in the face of collaboration.

It takes a new breed of IT leadership to address these issues. To see how you can make a change, you’ll need external experts with significant experience. Though expensive, they will be able to quickly help you see what should be changed to improve your IT organization.

The implications of Automatically Created APIs.

CITO Research managing editor Deb Cameron pointed out a data governance problem described in this Forbes.com article, Box Partners With Roambi To Attack The BI Market.

Discover what mobile capabilities mean to a particular business.

Creation of VMs from a golden image can be automated, but the requirements are not easy to discover.