Print this article

Comment

01 June 2003

In the new world, influence comes from the bottom

Enterprises will have to find ways of producing exactly what the customer wants, and then delivering it overnight, an extreme version of the demand or pull model.

Imagine if a football game played out just like a business does. Rather than allowing the players on the field to react in real time as the game and plays develop around them, they would be required to execute a recipe-like, step-by-step sequence of plays sent down to them from the head office, developed months before by the owners and managers.

The thought of either side following a process from those not involved in real-time decision making is patently absurd. And yet it is precisely this model that describes the business operations employed for the last century or so—from the top down. In an era of mass production coupled with almost continuous economic expansion, a business paradigm that depends upon forecasting market demand and then building large batches of products to that forecast works fine. The catch is ever expanding market demand covers up the waste in this model. In a seller's market, buyers have low expectations, and will put up with high prices, wrong products, low quality, and missed delivery commitments just to get something.

And so it is with enterprise software. It, too, follows the same scenario from a functionality standpoint and from a structural standpoint.

IN THE BEGINNING

From the day the first general ledger program ran a half century or so ago, the function of enterprise software has predominantly been to facilitate the higher-order business processes of an organization, such as finance, forecasting, and planning. Calls to action for manufacturing and the supply chain were manual. In the 1970s, when materials requirements planning software appeared, the software developed top-down influence. It took a forecast, threw in bills of material and inventories, and then "exploded" out a list of materials requirements, a set of work orders for production, and purchase orders for suppliers. Even as a wider variety of applications began to appear, including those focused on bottom-level processes such as manufacturing execution systems, their output still came from input that filtered down from the top.

In addition, as enterprise software became a multibillion-dollar industry, software packages themselves became true top-down entities from a structural standpoint. A team of software engineers repeatedly tried to forecast what functionality "Fortune 100" manufacturers in the U.S. or Japan would need, and then wrote millions of lines of code to embed this view into an application. Much of the embedded functionality may have been of little use in any specific installation—sort of an "overstock inventory" of code—but economies of scale, the mantra of the twentieth century, dictated that it be there. And when it came time to actually install one of these monoliths of code, teams of "systems integrators" came in to harmonize a very abstract view of the enterprise with the day-to-day, working model that actually existed. It's no wonder that enterprise software has gotten the reputation of being fat, expensive, unwieldy, immutable, and producing return on investments far lower than expected. There is a new way to look at things these days, and it is from the bottom, not the top.

MEETING NEEDS

The global economy is transforming itself, a change precipitated by the instantaneous communications infrastructure provided by the Internet, coupled with levels of productivity in all of the manufacturing economies. Manufacturers are facing dramatically increased customer expectations in terms of availability, timeliness, quality, and specific functionality, as well as increased governmental requirements. You can call this new economy the "execution economy."

In this economy, producers operate their businesses at new levels of excellence in execution, using best practices. They must exorcise waste in time, labor, and material under the growing religion of lean supply. Defect rates worse than one per 10 million will be unacceptable under the Six Sigma mantra. Finally, enterprises will have to find ways of producing exactly what the customer wants, and then delivering it overnight, a version of the demand or pull model.

What is common to all of these strategic initiatives? They all focus on the bottom of the enterprise. And to actually be implemented, they demand a global, real-time software infrastructure embedded right among the people, processes, machines, and material that make it up, something called the "bottom-out" enterprise software paradigm.

In this paradigm, any business or manufacturing process can be described in software by an enterprise's own personnel, integrated with any relevant software control application, and then, through the Internet, influence or be influenced by any other process, anywhere in the world, for any reason. Thus, the influence of a bottom-level process radiates outward, horizontally, to other atomic-level manufacturing or supply-level processes, and vertically, through more abstract views of supply and manufacturing, until ultimately reaching the top levels.

Under the bottom-out model, enterprises can selectively implement bottom-out applications where the need is greatest, and then grow that infrastructure outward (in three dimensions) in a step-by-step fashion, as need and experience dictate. Bottom-out enterprise software applications will therefore go in quickly, in a matter of weeks, and begin paying for themselves almost immediately, in complete contrast to legacy enterprise software. It will allow business to read and react just like football players on the gridiron. IT

Adam Bartkowski, CPIM, is president and chief executive officer of Long Beach, Calif.–based Apriso, an enterprise software provider. Bartkowski holds a bachelor's degree from Amherst College and an MBA from Northwestern University.