6.5 How to Use the Pyramid

---

6.5 How to Use the Pyramid

Use the pyramid to augment your implementation strategy from a scheduling perspective. Unless your target state is represented on the project calendar, I do not know how you can keep everyone's eye on the ball, including your own. Does that mean that from a scheduling perspective the project proceeds from the bottom of the pyramid to the top? Yes, it does.

The first step is to recognize that we have multiple pyramids that, in concert, cover the entire project. We just counted five pyramids for our vendor management initiative. Each pyramid should have its own independent schedule. It should be linked only to other pyramids for dependencies. Two of our pyramids were Web sites built for customers. It stands to reason that they will have to be installed in the data centers at some point, and thus have dependencies on the infrastructure that will include the LAN, WAN, and DR. We will not link them to the call center end user, however, because the "point" of the Web site pyramid is someone else (i.e., the customer/beneficiary). Having said all this, let us go back to the call center end user pyramid and revisit the schedule (see Exhibit 5).

Exhibit 5: Vendor Management Implementation Pyramid Revisited

The call center pyramid depicts the following scheduling strategy:

• The two facilities have to be prepped for the end user. This includes office construction, such as cubicles and conference rooms. Computer facilities will also be required. Deliverables include physical security, raised floors, conditioned power, racks, data cabling, and so forth.

• The infrastructure comes next. We need switches, routers, firewalls, and multiplexers for the local area network (LAN) and WAN. The physical computing system, such as servers and tape backups, must be installed in the racks. Then the operating systems, antivirus and monitoring agents, and other system utilities will be installed.

• Our third layer will be the applications used in the call center. This may be purchased and customized, or built from the ground up. Databases, development or support tools, and Web servers may all be required, depending on system design.

• A team of call center and data center personnel will be hired, trained, and assigned to tasks and seats in the two facilities. After some period of orientation and practice, the sites will be open for business.

How do we know all this? The past few pages showed a transition from a technology-oriented schedule to an operations-oriented implementation plan. This change resulted from a variety of sources, including prior experience. Quite naturally, other team members provided significant input, including the client who drove the vision, the sales manager who in conjunction with the client crafted the SOW after winning the request for proposal (RFP) battle, and subject matter experts. Hopefully, the steps documented in previous chapters now make more sense than ever. One must truly understand the ins and outs of projects this complex before attempting to build the "real" schedule. As previously discussed, there is probably just one person on the project team who worries about everything coming together, and that is the project manager.

The next step is to revisit the schedule and change it in a way that better defines the direct path to target state (see Exhibit 6). We are still pushing a pencil around, and the computer is still turned off. For simplicity's sake, let us stay focused on just one site - the main call center/data center facility.

Exhibit 6: Second Draft Vendor Management Project Plan

Month	1	2	3	4	5	6	7	8	9
Data center	Prepare data center facility (power, raised floors, etc.)		Install WAN, data center LAN, platform		Install and test call center application software, data bases, etc.			Test call center applications	Operations dress rehearsal before production turn up
Call center	Design office space	Order, receive, and install cubicles, office fittings, and furniture			Install LAN, workstations, and phones in call center work space
Staffing	Complete model	Recruit and hire to staffing model
Operations	Complete SOW		Develop teams and train in operational requirements			Application training for users

Not that we are finished scheduling the call center pyramid, but we have gotten on the right track compared with earlier in this chapter when practically all we saw was disconnects. Now, we can actually locate seminal events on the schedule, including sequences as well as significant points in time relative to other pyramid components. Although we have not drawn any arrows yet, the dependencies are becoming more obvious. For instance, when you look across the "staffing" line and go up one row, you can see that we have physical workspace available as new hires come on board. If you look at the "operations" row beneath "staffing," you can see that team building and training commences after the recruiting-hiring process is under way.

Also, the applications training component of operations begins as we are completing the installation and testing of the applications. In other words, users are trained in software modules as they come on line. With this project, we did not have the luxury of waiting until everything was done in one area before moving forward with the next steps. We were also aware that in today's workplace, you cannot identify new hires in January, and then tell them they cannot come to work (and get paid) until July. That may work for movie stars, but the rest of us do not live like that. Therefore, because we were bullied into an aggressive schedule, events were phased or staggered as their precursors began showing signs of life.

There is a lot more to be done with the pyramid. Before taking that path, it is important to go through the same exercise with the other pyramids. Again, we are looking for two things: first and foremost is the coherent organization of getting from each pyramid's base to its point, just as we did with the call center pyramid; second, we need to understand dependencies between pyramids as well. The quality, timing, and risk associated with these events need to be identified and scheduled as far in advance as possible.

It was previously mentioned that the two Web applications (reporting and procurement) would eventually have dependencies on the data center and the backup (DR) site. Certainly this is a requirement for final testing before going live. Initial development and testing need not wait for the facilities or the wide area network (WAN) to be available, however, assuming that alternate work facilities are available and convenient while the production environment remains under construction.

---