To Build or Buy? That is No Longer The Question - Part I

By Daniel Gingras

(Back to article)

The past 15 years have seen revolutionary changes in the software market, driven by significant increases in both hardware processing power, and an improvement in the way software is designed and developed.

In addition, a move to blend development in lower-cost development centers across the world, coupled with standardized design methodologies, we are now seeing a world where there is an almost limitless supply of unique and specialized software available at “relatively” reasonable prices.

It may be that your particular industry now has a plethora of choices available for even the most specialized applications which previously required custom software development.

More Research on CIO Update

The ERP Renaissance

The Death of the Internet?

It’s Time for New IT Governance Models

Surviving as the First CIO

FREE IT Management Newsletters

Having taught system analysis and design, and technology strategy at Boston University’s graduate school for almost 15 years, I’ve seen the slow but exorable move from a world where CIO’s made complex and difficult “build or buy” decisions for specific systems, to a world where almost all software is now “commercial off the shelf” (COTS). This means there is almost no “build or buy” decision left to make.

Most informed management would look pretty closely at an IT organization which proposes writing an accounting or manufacturing system, and rightfully so, because there are a multitude of choices available to any size company from pure startup to Fortune 500. The questions we now focus on instead of “build or buy” is how to efficiently select COTS packages in the “post-build” era, and when should you think of retiring formerly home grown software in favor of something more mainstream.

The shift from development to acquisition of software in the commercial environment also has profound implications for IT as we focus less on software design skills to business skills, project management skills and communications skills. Increasingly, we see organizations that have difficulty understanding that both the skill level and mix must change as corporations move from writing their own software to acquiring and integrating standardized packages.


There’s a natural tendency on the part of many organizations to think their businesses are absolutely unique, and “no one else does what we do.” Anyone who’s implemented enterprise resource planning (ERP) systems, has heard that canard at almost every turn, but in truth, unless you have a proprietary, patented business, and you are the sole global supplier of your goods or services, you are like many other businesses.

You may manufacture a very unique product, but the essence of manufacturing is common across almost every manufacturer. If you’re hearing the “we’re really unique” mantra, you need to step back and re-evaluate whether you really are.

In truth, there are major commonalities in most businesses, particularly manufacturing businesses. They generally fall into a small number of common types, and there is COTS for all of them. In general, once you find the type of package—accounting, ERP, business intelligence, et.al.—you will find there are dozens of companies furnishing the types of systems you need.

These packages are generally used by hundreds or even thousands of companies very similar to yours, and the key thing to remember here is the “institutional knowledge of all of these companies is embodied in the processes defined by the software itself.”

This a very important concept to grasp, because knowing there have been thousands of software improvements over the years, representing the collective wisdom of hundreds of companies, is the key to understanding why COTS software now dominates the marketplace.

The concern most organizations have with the embodiment of institutional knowledge in the software best described by the question: “Well, if everyone in my business uses this how am I going to create any differentiation in my processes, am I going to be forced to do everything the same way my competitors do it?”

The short answer is no. Just as we’ve moved from writing our own software, we’ve also recognized modifications to standard packages are also very expensive. It is well known that 80% of the total lifetime cost of software is in the maintenance, so changes to COTS are as bad as writing it yourself.

We’ve moved from the era of customization, where we create expensive, hard-to-maintain changes to software, to the era of configuration, where the software vendors have figured out it’s much easier to create software with extreme flexibility and allow the users of the software to incorporate the uniqueness of their processes by changing the configuration of the system as they implement it.

In most major software packages you find lots of “switches” which allow subtle changes in the way the package works. This is not to say that configuring software is simple. Sometimes it can be quite complex, but the beauty is these configuration changes do not have to be re-done every time the vendor comes out with a new change to the package.

Owning an off-the-shelf package has a huge amount of benefits for most companies. The key is understanding how to select the package.

Daniel Gingras has been CIO of five major companies and is a partner at Tatum Partners, a nationwide professional services organization of senior-level technology and financial executives who take on leadership roles for client companies. He has more than 30 years of IT experience and teaches computer science at Boston University. He can be reached at dan.gingras@tatumpartners.com. In Part II of this article, Gingras will explore the COTS selection process.