Businesses provide many different types of fulfillment. Some fulfill by keeping shelves properly stocked, and then delivering from that stock. Some businesses go a step further and manufacture or assemble those products. Others provide professional services such as engineering, case management, analysis or legal services.

The fundamental problem of fulfillment can vary in complexity, but the premise is always the same: The business has to organize and deliver resources of some kind to their customers. It's the age-old adage of providing supply to meet demand.

Believe it or not, automating the process often is more alike than different. If we're offering professional services, then the "products" we are selling are simply self-renewing; every new hour means an employee has another hour of product (their time) to put up on the shelf. If we are producing those products ourselves, we have some sort of assembly process to organize. Finally, whenever we are selling physical products, we have to have a purchasing function to replenish those products as they are sold.

Therefore, let's assume that for most businesses we can generalize these operations at a high level as delivery (or scheduling), assembly and replenishment. In turn, each of these three processes involves some sort of logistical challenge, and herein we find the opportunity for automation, and the return on investment of that automation.

In this phase of the business, the benefits of automation can be more quantifiable. For example:
• What is the value of a 10 percent increase in manufacturing workers' productivity?
• What is the value of a 10 percent increase use of service workers (i.e. increased billable time)?

• What is the value of a 10 percent increase in use of a piece of equipment?
• What is the value of a shift reduction or reduced overtime?
• What is the value of needing to maintain 10 percent less stock?

Unfortunately, most people consider automation of this phase of the business to be the most challenging. The nuances of scheduling human beings with variable schedules in projects with equally variable schedules can be complex. Similarly, the notion of routing work in process through even a simple manufacturing environment becomes very complicated when a particular part is out of stock unexpectedly, or a certain customer's order is suddenly given "rush" status.

Because this appears on the surface to be complex, few organizations put any significant effort into automating it. They simply push all of the order information, or all of the scheduling information, to a human. That human usually is a person with decades of experience who knows all the nuances of the various "gotchas" that any given process can throw out. This person then jumbles the information around and makes the best decisions they can. Because there is no way to know whether something better could have been achieved, management usually accepts this as being "the best we can do."

The thing to remember about automating these processes is that the computer doesn't have to direct every single detail of the process. I highly recommend people to consider the value of simply providing better and more timely information to the human who makes the decisions. Rather than making decisions based on last night's data, why not update those decisions based on real-time data, or by looking at exceptions that are delivered in an automated fashion? That alone will usually yield a percentage increase in key metrics like utilization and productivity. Additionally, the value of self-directing purchasing or giving management insight into service delivery can be invaluable, especially during an economic downturn. This was originally posted on the Central Penn Business Journal Gadget Cube.

Next week I'll look at the billing process.