Leadership

Is it time to reduce the complexity of our solutions?


There are many information technology trends to observe if you’re in the business long enough.  For instance, when I began we spent much of our time replacing thin-clients (aging dumb terminals) with full-blown Windows installations and bloated software clients.  The trend was to push more horsepower out to the users and distribute the processing to increasingly powerful PCs.  We then turned around a few years later and began emphasizing bringing applications back to the data center and using thin computing once again.  Now, I am the unfortunate witness to end-user mutinies which are forcing us to return fat-clients to the desktop.  It’s classic centralization versus decentralization, and is a topic worth discussing in its own right.  But I will save that for another day.

Simplifying the solution.  It takes little effort to glance back a few years ago and discover the many IT headlines which quoted CIOs calling for less complexity in their environments.  Fast-forward to 2007 and again we have executives clamoring to reduce the clutter in their complex systems.  Companies have enough issues getting quality products to customers, and don’t necessarily need overly integrated computer systems which could crash and put a screeching halt to their entire business. 

Competition, the push for collaboration and the lack of interoperability between disparate systems has gotten us to our current state of IT dependency.  Custom interfaces and a large dose of middleware abound to create a web of intricate information pathways.  One down system or interface leads to adverse effects felt throughout the entire corporation.  IT pros feel the stress of supporting such intricate systems too, and typically spend more than three quarters of their time performing maintenance and upgrades to existing systems.

Doing more with less.  There seems to be an initial surge in the adoption of new technologies which can possibly, just maybe, get companies ahead of their competition.  It’s a juggling act for CIOs.  On one hand, they must attempt to provide support for a rapidly increasing number of systems, and therefore shoulder more responsibility for the success of the company.  And on the other, they must do all of this efficiently with stagnant staff levels and overall declining IT budgets.  It’s no wonder they turn to collaborative software tools and service-oriented architecture to deliver innovative solutions. 

Unintended consequences.  The problem with tying everything together and jumping first to implement new solutions is that it makes system rollouts and post-implementation support a much more complex endeavor than it ever used to be.  Project scopes have increased dramatically and there is now a domino effect resulting from one failed system that can have negative consequences on otherwise separate business units. 

These larger projects, which require more cross-departmental collaboration and an increased staff skill level, are also much more susceptible to scope creep and going far beyond the initial project budget.  This can ultimately lead to untimely conclusions to job tenures and management changes.  Just ask Philadelphia CIO, Terry Phyllis, who has faced harsh criticism after a failed water bill project based on an Oracle provided solution went more than two times over the budgeted amount and must now be ripped out and replaced.

Managing the solution.  There are many IT staffs walking a thin line between complete system failure and a good night’s rest.  The more bound a business is to its IT solutions, the more critical it is to properly manage and maintain the system’s health.  This is exactly why we will see more businesses spending IT dollars this year on systems management, alert monitoring and problem resolution tools than previous years.  For instance, major vendors HP and IBM have developed solution offerings such as the Adaptive Infrastructure and Autonomic Computing to assist with a more adaptive, self-managing IT system.

There are only so many new systems and technologies a team of IT pros, however skilled, can implement before it becomes critical to take a step back from the frenzied pace and analyze the existing solutions.  Integrating everything may not always be the best course of action.  Sometimes simplifying can lead to more satisfied customers and staff.

10 comments
XEntity
XEntity

I have worked in a wide breadth of operations across many sectors in the economy. I have also been working with IT since 1982. The recurring theme / problem I have seen is multi-faced: 1. Leadership in the IT industry is polar. They think about cool technologies and dressing systems in pretty bells and whistles but fail to address the core of their own industry. The industry is not about technology. The industry is about information and information processes then applying technology to those processes. These processes often travel across a wide breadth of departments, organizations, and even industries. Information and knowledge 'ownership' is temporal for extremely short intervals of time before others gain access to it and begin to use it. I am not discussing personal information like SSN's but instead I am discussing practice. The actual owner of knowledge are the communities of practice. Managing and controlling information must be addressed at this level. 2. Common leadership in the various industries are inept at effectively utilizing, implementing, or managing information technology. This was something I thought would eventually fade as more competent leadership matured over the years but that has failed to fully materialize. At the center of the problem is that despite many IT professionals who are highly skilled, they often are hogg tied by CFO, controllers, and other leadership. I find that a large majority of leadership and management have little grasp at running an operation but have made their way into leadership roles. Decisions are made not made based on sound strategies that resolve the root problems and advance business performance. This is something I have seen repeatedly more than the cases where they have in fact made sage decisions. 3. Information techologies are strategic in nature. The effects are long term but leadership's attention span is about 90 days. This is evident in their calls to "Go Live" and to "Show Results". In talking these same people the attention span is about 90 secs. Most are incapable of seeing the abstract and almost always ethereal world of IT. What appeals to them is what "sounds good". Hence, the wild swings back and forth between strategies and platforms without realizing the long term cost and effects. In short, today I believe there is a clash between information architectures and organizational architectures. Organizations want to cling on to tradition establishing artifical boundaries that disrupt information processes. They employ expeditors and redundant efforts to resolve the conflicts between information and organizational architectures. The complexity enters at this point. Adaptive and autonomic systems are cute technologies fot IT professionals but what is really needed is adaptive systems for industry. The systems must be organized around information processes and be able to self organize. Staffing should be structured along these information processes. In the end, there needs to be a symbiotic union between staffing, business, and IT that is adaptable to emerging conditions.

Tony Hopkinson
Tony Hopkinson

It was a big pity we didn't come up with more memory and hard drive space before we ran out of space to store fragile opaque propriety binary constructs.

paul03pa
paul03pa

Is it time? Simplicity is, has been, and always will be ...the best approach. Whether GUIs, budgets, clueless management or fluff marketers are responsible is irrelevent. The best deliverable balances the reality of need-it-now demand with the basic requirements of good design. When substance-less Mr. Flash wants it to "fly around the room real time", just say no. If Mr. Flash is the CEO and it means you're out the door, well... that's the way it is.

jlawrence
jlawrence

Systems that are too tightly interdependant lead to inflexibility in the business model, whereas flexibility is a business attribute in demand today. If a business system increases profitability by sealing itself into a 'same-path' algorithm, it has decided to do business that same way, forever. It's a lot more difficult and expensive to redesign a system to respond to a new input when all your answers are highly complicated, tightly integrated components, then it is to tell a group of people, "Here's the problem, get on with it".

JamesRL
JamesRL

Companies that fail to do strategic planning for IT run into this all the time. What kills alot of comapnies is the number of IT silos that don't talk to each other. So we have java developers on this project, .nets on the other, difficulty integrating and the costs of supporting two platforms, and the additional costs of fighting it out as to which becomes dominant (Oracle versus DB2, Unix versus mainframe and so on). What needs to happen is someone needs to sort out the priorities of the organization in order to simplify things, you can't do it on a case by case basis, it has to be a strategic objective that everyone knows understands and plans for. Sometimes it doesn't matter which one they pick, as long as they pick one, as even the worst choice is better than living with 2 conflciting technologies. Blame the CTO/CIO -its their responsibility. James

Tony Hopkinson
Tony Hopkinson

to do the job properly. Application X carries out function Y, which we would NOW like in Z. Because application X was sold ten weeks before it was written from a specification written on a beer mat. The only way to get at function Y without rewriting X from the ground up is to add Z to XY. So do you want a less complex solution in six months time, or the usual crap in a week? New fully integrated XYZ application is of course a massive marketing bonus point. Not to mention two teams for two products, now only one product..... Come up with this, let the MD win at golf, even with the handicap of your tongue wedged firmly up his arse and you too could make CTO.

stress junkie
stress junkie

This entire "article" is just a collection of business school catch phrases and unsubstantiated assertions. This is the kind of pablum that one expects in the first chapter of a text book used in a business school survey course. Where is the "clutter" in the computer environment. You have your office document tools and your database server. From the end users' point of view that's all of the "clutter" that most businesses face in their computing environment. How does that cause problems? Simplifying the solution. Dong more with less. Unintended consequences. Managing the solution. What a load of meaningless business school crap! I'm surprised that you didn't throw in a reference to Six Sigma and another to ISO 900x to round out the collection. All of these ideas are just used as a smokescreen among upper management types to pretend that they have some useful knowledge and some useful function. They do not posses any useful knowledge and they do not perform any useful function. The assertion that CIOs do anything more than collect exorbitant paychecks makes me ill. I've worked in plenty of businesses of all sizes. All of the upper management are either getting paid for doing exactly nothing or they are a liability to the business. Most CIOs, CEOs, VPs, and whatevers couldn't tell you the difference between a pixel and a key cap, not to mention the difference between Oracle and Microsoft SQL Server. They're a collection of stick-up-the-ass incompetents who managed to convince their frat buddies to hire them for a job which they do not understand. Except for those points I found the article to be very insightful. :D

michael.mulvihill
michael.mulvihill

System integration is a very valuable and important part of business process streamlining. Sure it's easier for the IT guys if the person who takes customer orders has to get into a different application to check customer accounts (simplified example), but it's not easy for the end-user and it's frustrating for the customer. Technology should not be invasive, it should supplement the result by reducing the effort and integrating the data at some point in the process. The only thing I would like to see changed in this movement is management's sense of urgency over everything they read in a magazine. Yes system X can share data and integrate with system Y and it could be done next week but that won't be the best possible solution. We need time to step back and really examine the process, find the points of failure and account for them. Integrating Peoplesoft HR with AD may be cumbersome in the backend but it is a godsend to the frontend. We have won over very important allies for our cause by simply giving them single sign-on access. As we move into a more integrated work environment I'm sure tools will become available which aid in the design, implementation and support of these new complex systems. Until then we will have to rely on documentation and forethought. But I would hate to have to go back to a time when break the links we have built and go back to dozens of systems unaware of each other. I do agree that none of our business problems will be solved by the latest catch phrase of the day. L. Ron Hubbard was correct that, in his time, there was no better way to make money than to start your own religion. I'm convinced that today there is no better way to make money than to create a buzzword heavy certification program and pass out annual awards for participants.

Bill Elmore
Bill Elmore

Well, I certainly appreciate your comments even if they are scathing. My apologies for the many catch phrases and corny headings. :D In response though, "office document tools" and a database server is not exactly the clutter I was alluding to. By clutter and complexity, I mean something more along the lines of a PeopleSoft HR system which integrates with an identity management system which is responsible for automatically creating AD user accounts and assigning group membership based on job titles. Or, as another example, companies that have systems which are responsible for customer/patient demographic data which have multiple outbound/inbound interfaces with other lines of business. These systems, a few of which may be Linux, some Windows, a few UNIX, have a mixture of Oracle and SQL (pick your flavor) back ends. On top of that, clients are connecting across the WAN via Citrix ICA to application servers hosted on virtual servers. And, oh yeah, printing solutions are being provided by print server appliances running on Linux. This is still a little ?dumbed? down, but throw in a few wireless devices and you get the picture. Enterprise computing environments have become incredibly complex. By the way, this blog post was targeted to tech managers as an opinion on anticipated IT market trends, not a detailed technical white paper. Thanks again for reading and providing comments though! :)

stress junkie
stress junkie

... for tolerating my tirade and responding so positively. I've taken my meds now. :D I suppose I understand what you mean by technical clutter. When I got into this business the idea of networking computers was reserved for large companies like Lockheed and IBM. I guess I just got used to the clutter since it accumulated slowly over time. I really do have a bug up my posterior regarding managers. I've worked with a lot of them (maybe 60 managers in 40 jobs with me working as a contract employee) and I get upset when I think back on those experiences. I think the worst was when I saw numerous higher level managers at Digital Equipment Corp take huge bonuses and then tell the wage slaves that worked for them that they couldn't have a deserved raise in pay. That didn't directly affect me but I felt a lot of sympathy for the direct employees.

Editor's Picks