Technology and Apparel-Will We Ever Catch Up? OR The Failure of Supply Chain Technology

Are We At the Beginning or the End?

We are at a critical juncture in the apparel industry: Prices have come under extreme pressure due to the Walmartization of our retail segment, and of the narrowing of choices of sourcing locations due to political risk, SARs, etc. Due to these risks being figured in everyone's cost, manufacturing costs have not declined commensurate with prices. And, while certain industries have enjoyed materially reduced costs due to improvement in information technology, apparel has not been one of them. In fact, the entire soft goods industry is facing a near-crisis with regard to process, technology and timing. This industry is facing the crisis that the automotive, aerospace and consumer product goods faced decades ago. They came through it not so unscathed, but smarter about how to work- so why are we having so much trouble? How can we insure the survival of the industry as a whole (except, of course, for Wal-Mart)?

The issues fall into two distinct categories: the business processes and the technology that support them. These two need to come together in order for any progress to be made.

In the last 5 years, technology has made a huge difference in our lives- primarily technology that makes it easier for all of us to communicate (c2c) with each other and communicate with companies (c2b, b2c):

1. Email was available prior to 5 years ago, but it was used only by companies and governments; now even your grandma uses email
2. Web tools such as IM allow us to speak to each other over the Web in real time
3. The emergence of Web-enabled cell phones and PDA's has extended communication to wherever we are at any moment
4. More than 25% of Christmas gifts are now bought online
5. Every company worth anything has an online presence
6. Kids can find information on any subject in seconds on the Web
7. Companies can communicate with consumers, sometimes unsolicited, and without even knowing your email address (hey, every good thing has its ugly side)

But what really has changed on the vaunted b2b side? Owing to the intellectual groundwork paved by people like Mohanbir Sawney of the University of Chicago, the terms "supply chain" and "vertical" came into use in the business lexicon. During and around the year 2000, there was a resultant "dot.com boom" where everyone and his sister claimed they could "automate" the supply chain-create communication around the development and execution of a product that would make paper, spreadsheets, and cumbersome systems like Excel obsolete.

Well, here we are in 2003 and nobody in the apparel trade has done it yet. There are web-based applications that claim to automate this part and that part of the process; mostly parts of the process where information goes in a stable flow (like logistics). But none that knits together all the different pieces from product development and sampling through the purchasing of materials, and the actual production planning and production of the item. Why? Here's an example of why a vertical like apparel may have failed to create a TRUE supply chain system:

1. There is no such thing as the supply "chain." Information flows in the development and execution of a product do NOT follow a straight line. They go back and forth in a somewhat predictable pattern, more like a sweater than a chain. For example, when you write a purchase order for a product that must be approved, you cannot assign an accurate delivery date to the production unless you take into account the approval status of the product-fabric, fit, testing, etc. So you must have a way to relate sampling and testing processes to production.
2. The information from each segment of the process impacts the other, so the entire process must be taken into consideration from the outset and included in the technology for it to work. The bits and pieces approach is costly, time-consuming and has never worked.
3. There is no stability in most of the process flow. For one manufacturer, each customer may have different requirements. And each vendor of materials may have different lead times. Any system we use must be able to program these variations into each individual order.
4. The actual durations of each development or production event are not perfectly predictable. Well, that's OK, we can report that something happened a week after it was supposed to on a spreadsheet-but what is the impact on all that comes afterward? Whenever something changes, or many things change, we must know the impact on what comes after and the final delivery date.
5. The changes happen often and quickly. Change management is intensely hated because the lifecycle of apparel goods is so short; unlike other products whose lifecycle allows for a "supply chain". The shorter lifecycle puts many more constraints on the scheduling of raw materials and goods, which leads to trouble in flat-file systems.
6. Apparel is just different. There is very little BTS (Build to Stock); most items today are BTO (Build to Order). Styles and colors are myriad and change often. And here is the key issue, so do the sourcing locations due to the intense pricing and quota pressure on the industry. Due to all the variables involved, the process seems so dynamic that nobody has figured out how to bottle it. Is it really so dynamic when reduced to common denominators?
7. Nobody knows how to put all the above into a schedule. In most cases we can create a schedule with expected finish dates for each activity in Excel or a database system or whatever. But it is not dynamic or predictive, which is what we need. And it needs to know how to respond when something changes. Because it will and it does.

So what is missing? Surely we need technology expertise to program all these events and activities into an application. That we have aplenty. However, history has shown us that technology expertise alone is NOT enough. These guys know how to program anything, but they don't know what anything is. What happens when and if is the province of domain expertise (back to process). That, my friends, is and has been the missing ingredient in the failures above: Technologists don't know how to program something they don't know; and domain experts generally don't know how to program. The development of these projects has been dominated by technologists without the partnership of domain experts. Result: failure. A legendary example of this is Nike commissioning I2 to automate their supply chain. Does a reputed $400 million later point to success?

How did we go so wrong? How did we forget something so simple and logical? At least part of the problem is another effect of the explosion of technology in the last ten years. The role of technologists in corporations went from guys walking around with a screwdriver and pocket protector to Senior Executive status. Once corporations separated technology on an organizational level and gave it the status of "Chief" (you know, CIO, CTO) that was the same as or higher than any operating position, we created a fiefdom. A department or position that was created to serve the users took on a life of its own: technology can because technology is. Don't tell me I have to do what "they" want? What do "they" know about technology? Well, they (the users) probably don't, but they know what they know, they should know what they DO, and they definitely know what they need to succeed.

Next: Users and Techies-Can't We All Just Get Along?

Mike Serwetz- President, Ecompartners, Founder, Technology-Fusion, LLC.
Nancy Johnson- President, Optimyze, LLC, Vice-President, Technology-Fusion, LLC.