## Step 2: Model Activity Value

The second step in Web Business Engineering is to build a value model based on the activity map created in Step 1. The value model is a computer simulation of the mapped process's value.

In a road map, you have extra information to help you determine the best route
to your destination. For example, roads between major points of interest are
usually annotated with the miles between those points. And if the map is drawn
to scale, you know that every inch counts for so many miles. In Web Business
Engineering, the *value model* is the "annotation" to your activity
map that helps you determine the best use of the Web.

One kind of value model is a cost model. Each arrow in the map shown in Figure
1 has a cost associated with it. If you sum up these individual costs, you
get the total cost of the process. Now because the organization pays those costs,
it *at least* values the process by that amount. Thus, a cost model is
a kind of value model. Next we look at how to build such a model in a spreadsheet
program such as Microsoft Excel™.

### Building a Cost Model, Part 1: The Interaction Matrix

Each arrow in the activity map represents an information transaction. The arrows
can be depicted in an NxN *interaction matrix *(see Figure
4), where the rows in the matrix denote the source of the information (the
agent at the base of the arrow) and the columns denote the recipient of the
information (the agent at the tip of an arrow). You don't include all the agents
in your interaction matrix—just the ones whose costs your clients have control
over. We'll assume that the client doesn't have control over the Mailing Room/Student
interactions because these occur outside the client's organization. Thus, the
interaction matrix for our activity map includes only the Instructor (I), the
Assistant (A), the Mail Express (M), and the Mailing Room (R).

*An interaction matrix (Reprinted with permission from Flor, N. [2000], Web
Business Engineering, Addison-Wesley Longman)*

Each cell in the matrix depicts the cost of an arrow—or, more precisely, the cost of an information transaction. Next, we look at how to calculate these costs.

### Building a Cost Model, Part 2: Determining Cost Equations and Programming Them into the Interaction Matrix

Each arrow represents an information transaction, and each information transaction has potential storage, production, material, and distribution subcosts. For example, the cost of the Assistant (A) packaging an Instructor's (I) assignment and mailing it to several remote student locations (R) via Mail Express (M) has the following subcosts.

The production cost is the cost of copying and packaging the assignment, which is a function of A's hourly wage rate (AWAGE), multiplied by the time that it takes A to do the copying and packaging (CTIME):

The materials cost includes the cost of the paper (CPAGE) for the copied assignments. All pages of the assignment (NPAGES) must be copied for each student taking the course (NSTUD):

Finally, the distribution costs include the fee for mailing an assignment package (MAILXFEE) to all the remote sites (NSITES), as well as the cost of A's time delivering the packages to a drop-off box (MTIME * AWAGE):

There are no storage costs. Thus, in our activity map, the total cost of the arrow between the Assistant and Mail Express is as follows:

Similarly, you determine these costs equations for each arrow in the map. All these equations have several variables. You place the variables from these equations into a control panel (see Figure 5).

*Our cost model's control panel (Reprinted with permission from Flor, N.
[2000], Web Business Engineering, Addison-Wesley Longman)*

Finally, you program the cost equations into their corresponding cells in the interaction matrix (see Figure 6—the AàM cost equation is depicted):

*Programming the cost equations into the interaction matrix (Reprinted with
permission from Flor, N. [2000], Web Business Engineering, Addison-Wesley Longman)*

### Building a Cost Model, Part 3: Building a Summary Row

The interaction matrix usually depicts a single "run" of a process, such as one mailing. It is often useful to summarize these costs over several time periods, such as the weekly or yearly costs. So, the final step in building a cost model is building a summary section, which is shown in Figure 7 along with the interaction matrix and control panel.

*Summary section—in the upper-right corner—along with the interaction
matrix and the control panel (Reprinted with permission from Flor, N. [2000],
Web Business Engineering, Addison-Wesley Longman)*

We can now use our cost model to help us determine, or *diagnose*, where
to best use the Web to support the process online.