Project Management Institute

Your tax dollars at work (really!)

Lombard, Illinois, saved time and public funds by dovetailing outsourcing activities with a process improvement plan.

by Robert S. Lewis

REINVENTING GOVERNMENT can be more than an empty buzzword. Hard to believe? Although the dollar amounts involved in the following case study are relatively small, just imagine their impact if multiplied by the hundreds of thousands of small municipalities across the nation. Here's one town's success story.

Bigger Goals, Same Staffing. As a result of a 1995 strategic planning process, the Village of Lombard, Illinois, desired to increase its Capital Improvement Program (CIP) budget from $6 million to $10 million annually for a period of three years. However, because the increase in spending was only for a short term, no increase in full-time staff would be allowed. All impacts were to be absorbed or outsourced wherever possible. It soon became evident that integrating the additional consultants’ work with staff required an investigation of our existing process.

Prior to the increase, the CIP consisted of one major (more than $1 million) construction project, three large projects (less than $1 million, more than $500,000), 10 medium projects (less than $500,000, more than $100,000), and 15 small projects (less than $100,000 in constructed value). The budget increase added one major project, two larger projects, three-to-five medium projects, and 10 small projects per year. All but the smaller projects had portions that could be outsourced. Given an effective outsourcing process, the 40 percent increase in the CIP budget (workload) could be implemented successfully. In this article, we'll look at just one area of outsourcing process improvement—geotechnical data acquisition—and at the procedure Village staff used to streamline the process of geotechnical data acquisition and the integration of the work with our core competency.

img

(Note: The CIP projects consist of roadway and utility construction (water and sewer mains). The design of this type of project requires substantial knowledge of the in situ ground conditions at the surface and up to 10 feet deep. This acquisition of soil data is called a geotechnical investigation.)

The Problem. One portion of the CIP has always presented a problem for staff. The procurement of geotechnical services has been inconsistent and also inadequate statements of work (SOW) have at times led to cost overruns during construction. The procurement of these services has been time consuming and inefficient for staff—for example, $2,000 of geotechnical services may require up to 20 hours of staff time to procure and administer. A substantial increase in this activity, we realized, would absorb a disproportionate amount of staff resources. The cost effectiveness, efficiency and quality of the delivered service needed to be improved to enable the procurement process to internally serve the Department and not produce a “drag” effect on the entire CIP as the workload increased. The development of a proper SOW was required for every project.

This exhibit shows the work breakdown structure as task required and the corresponding labor-hours for each contract initiated

Exhibit 1. This exhibit shows the work breakdown structure as task required and the corresponding labor-hours for each contract initiated.

This fishbone diagram illustrates how problems with geotechnical procurement can lead to trouble for projects. The top portion of the diagram represents internal causes and the lower portion represents external causes of problems

Exhibit 2. This fishbone diagram illustrates how problems with geotechnical procurement can lead to trouble for projects. The top portion of the diagram represents internal causes and the lower portion represents external causes of problems.

In addition to the time element, the quality and scope of the provided services was inconsistent. With over 75 firms in the Chicago metro area vying for a contract, firms cycled in and out, and little professional relationship was established before the contract was over. The scope of services provided for identical projects varied with each firm. This inconsistency effected the overall quality of service received. As our contracts must be open to all competent firms, legal requirements were hindering the establishment of a contract with a few good service providers.

The geotechnical procurement process had to be altered to implement an appropriate SOW, integrate the delivered services, improve quality, and fulfill legal requirements. If this could not be done, the 40 percent increase in the workload may not be possible to implement in an on-time, on-budget fashion.

Time Schedule. The implementation of the process improvement began at the end of the 1995 construction season. The procurement process was targeted as the initial area of focus. Data was collected to identify how many work tasks were required and the duration of each task. The process was changed before the 1996 construction season and was ready for implementation by March 1996. The staff reevaluated the procurement process at the end of the 1996 construction season, and other minor adjustments were made to “tweak” the process again.

Process Assessment. A team was formed from members of the Engineering Division and administrative staff from the Finance and Public Works departments. The existing procurement process was evaluated. The number of tasks required (work breakdown structure) and task duration was measured by averaging the time required from this type of procurement over the last two years. A complicating factor in the process is the additional steps required to follow state and federal laws on individual projects when this type of funding is present. The reported WBS includes the steps required for procurement that meets these requirements. The team decided the process could not be radically reengineered. Efficiencies must come through customizing and streamlining the individual tasks. Exhibit 1 shows the WBS as tasks required and the corresponding labor-hours for each contract initiated.

The three largest tasks were writing the RFP, selecting the consultant, and administering the working contract. These three tasks comprised 12 of the 20 hours that an average contract required, or 60 percent of the total time. It was felt the most critical part of each project was the preliminary engineering, which set the SOW and identified what was to be done and when. The existing process dedicated only 10 percent of the total time to this activity. The new process needed to increase that percentage and decrease the total time required for the entire process.

While 20 labor-hours may not seem like much of an impact, over the course of an entire year, 30 projects would require this list of tasks and the 20 labor-hours to be applied. As the CIP is expanded, the project count would expand to 40 projects per year.

Exhibit 2 indicates how geotechnical procurement can lead to trouble for projects. The top portion of the diagram represents internal causes and the lower portion represents external causes of problems. Quality in the procurement process was affected by four main causes. The first is the internal process, which can be classified into four subcauses: (1) the task is too time-consuming for staff; (2) the legal requirements are differing or not understood; (3) other departments have inputs and requirements that are not identified; and (4) the Finance Department's need to have some control in the process creates additional layering in procurement.

The second internal cause of poor quality service delivery is poor management. The four contributing subcauses are (1) poor SOW, where the scope requested is insufficient for the data required; (2) low priority, due to competing interests, where the project does not receive the attention needed; (3) poor communications, where staff does not talk to each other or with consultants to integrate the required SOW; and (4) unique project requirements that often go unnoticed or overlooked.

The first external cause that can lead a project into trouble is inadequate service delivery, indicated as “Service” on the Exhibit 2 “fishbone” diagram. This is categorized into four subcauses: (1) late data delivery by the consultant, where the project is nearing completion and data arrives late, requiring change orders to implement; (2) the data that arrives contains useless information or data contrary to the scope requested; (3) the service is inconsistent due to personnel shifts or turnover requiring the consultants’ staff to be briefed over and over again; and (4) poor service delivery—data that arrives so badly written that no one can understand it, or that the results make no sense when compared to the SOW.

The second external cause is procedures that others require us to follow. The leading subcauses here are state or federal requirements that may require MBE (minority business enterprise), DBE (disadvantaged business enterprise), special advertising, or external review requirements. Shifting or changing legal requirements can also complicate our ability to work in an efficient manner, and finally, poor work by the consultants (indicated by the item “Poor RFPs” in Exhibit 2). Oftentimes consultants wish to get special attention or to gain work without the required submittals, or give submittals that are not clear or concise. These subcauses are a problem because not only do they consume additional resources, they often also take control of the timeline away from the local agency. This hinders or prevents proper integration of the delivered services.

This is the new procurement process for consultant selection. The exhibit indicates the new procurement process for each project. These labor-hours will occur between 30 and 40 times per year

Exhibit 3. This is the new procurement process for consultant selection. The exhibit indicates the new procurement process for each project. These labor-hours will occur between 30 and 40 times per year.

The team met and a new process was developed to address the causes of these problems. The problem-solving process involved brain-storming and staff interview techniques. The development of the new process required a strong effort to identify the causes of the problems associated with the existing process and justification as to why it was a problem. The new process was identified and brought to staff with recommendations within 60 days of beginning.

The New Process

To solve the inconsistency of the delivered geotechnical services, staff decided a yearly contract comprising the entire annual project volume ($165,000/year) could be awarded to a single firm. All interested firms could submit an RFP and one firm would be selected. The proposals would contain a menu of services that may be required for the whole year, with unit prices for defined services and hourly rates for personnel. Individual projects would be ordered by describing the tasks required, negotiating the labor-hours and estimating the total cost, and issuing a notice to proceed. The team contacted several vendors about the proposed process and they all felt the dollar volume would be worth the effort of a formal proposal process, which would require a 20-to-30 labor-hour effort from each firm submitting on the proposal process.

Reader Service Number 5042

This is also the new procurement process for consultant selection but the labor-hours occur only once per year. (In Exhibit 3 they occurred between 30 and 40 times per year.)

Exhibit 4. This is also the new procurement process for consultant selection but the labor-hours occur only once per year. (In Exhibit 3 they occurred between 30 and 40 times per year.)

This table shows the comparison between and consequent improvement in critical components before and after changes in the procurement process

Exhibit 5. This table shows the comparison between and consequent improvement in critical components before and after changes in the procurement process.

The selected firm would be able to establish a working relationship with us and we would know whom to contact for which service. The consultant, with the contract dollar volume, could afford to assign a person to monitor our concerns about the quality of the delivered service. We would also get consistent staffing and additional senior personnel that would be more concerned with our problems. This would create the ability to integrate members of the consultant's staff with ours.

The initial process was established through one contract that authorized an upset amount that was the estimated annual cost of these services. So the new process would have two components. The first part is the selection of a consultant and the second part is each individual project. Exhibit 3 shows the new procurement process for each project, with labor-hours occurring between 30 and 40 times per year. Exhibit 4 shows the new procurement process for consultant selection, with labor-hours occurring only once per year.

The net result of the new process is a 68 percent reduction in total time required from the existing procurement process. Given a 10-project increase, the existing procurement process would consume 800 labor-hours compared to the new process using 250 labor-hours; this equates to a 69 percent reduction in labor-hour effort. The emphasis also changes from a focus on procurement of services to a focus on the statement of work.

Process Improvement

Exhibit 5 shows the comparison between and consequent improvement in critical components before and after changes in the procurement process. The critical preliminary engineering task improves from 11 percent of total time to 32 percent. The total hours remain at 80; however, with a corresponding reduction in other requirements, additional time could be spent here at the discretion of the project manager. The consultant selection process has been significantly streamlined from 21 percent to 8 percent of total time, and even more impressive, a reduction from 160 labor-hours to 20 labor-hours per year, an 87.5 percent reduction in required hours.

Total hours required to run the projects (indicated as “Implement” in Exhibit 5) were reduced from 160 hours to 80 hours, with the savings in time occurring by streamlining the administrative process. This was achieved by several methods. A standardized process was created to use with standardized formwork. As many of the projects are similar, a standard checksheet was prepared for scope selection and required administrative duties. Also a database from all prior projects was created from which information about labor-hours, fee multipliers and total costs was compiled.

Existing Process Costs. Here's how the existing process costs were calculated:

The existing resource use is 30 projects per year times 20 hours per project, which equals 600 hours per year. Total yearly staff resources are 16,640 hours per year.

This process is 600/16,640, or about 4 percent (3.6 percent) of total yearly staff hours.

Staff cost per labor-hour averages $20 per hour, so staff cost of contract on an annual basis is $20 per hour times 600 hours per year equals $12,000 (for calculating for overhead and benefits, use a 2.5 multiplier equaling $30,000).

Proposed resource use under the accelerated program without the streamlined process is 40 projects per year times 20 hours per project, equaling 800 hours per year.

Annual staff allocation would be 800/ 16,640, or about 5 percent (4.8 percent) of total yearly staff hours. The annual salary costs would be $16,000 ($40,000 with benefits).

Exhibit 6, showing the percent of total cost with the existing method, graphically indicates where the larger process costs are occurring. This exhibit assumes the increased project count is 40 projects per year, to better compare with the new method, which is calculated at 40 projects per year.

New Process Costs. Here's how the new process costs were calculated:

Here is where the larger process costs are occurring in the existing procurement method. This exhibit assumes the increased project count is 40 projects per year, to better compare with the new method, which is calculated at 40 projects per year

Exhibit 6. Here is where the larger process costs are occurring in the existing procurement method. This exhibit assumes the increased project count is 40 projects per year, to better compare with the new method, which is calculated at 40 projects per year.

This chart indicates where the larger process costs are occurring under the new method. The cost savings in straight salary dollars is $11,000 per year. Considering the effect of benefits, the total cost savings is estimated to be $27,500 per year

Exhibit 7. This chart indicates where the larger process costs are occurring under the new method. The cost savings in straight salary dollars is $11,000 per year. Considering the effect of benefits, the total cost savings is estimated to be $27,500 per year.

The accelerated program use is 40 projects per year times 5.5 hours per project plus 30 hours (for selection process) equals 250 staff-hours per year.

The annual staff allocation is 250/16,640, or 1.5 percent of total yearly labor-hours.

Staff cost is averaging $20 per hour, so staff cost of contract on an annual basis is $20 per hour times 250 hours per year, equaling $5,000 (for overhead and benefits use a 2.5 multiplier @ $12,500).

Exhibit 7, illustrating percent of total cost with the new method, indicates where the larger process costs are occurring. The cost savings in straight salary dollars is $11,000 per year. Considering the effect of benefits, the total cost savings is estimated at $27,500 per year.

Year-End Review

The program was successfully implemented and the data received from this service is now an asset to the staff. The delivered services have helped on numerous occasions solve difficult constructions problems in the field and in the office. The professional relationships that have been established have proven invaluable. The staff, the consultant and the agency have integrated the process, shared information, and increased the quality and value of the outsourced service over that of prior or internal efforts previously undertaken.

Staff has decided to roll the contract over for an additional year and select consultants every two years, with an option to assess the rollover at one year being mutually agreeable. In addition to this step, it was decided that at the next selection process we should prequalify firms to reduce submittals from 75 to about 30. This reduction would occur because this last selection process revealed that many firms do not carry professional liability insurance coverage or appreciable umbrella coverage, exposing the agency to unnecessary risk.

A fourth change would be to allow only those firms that have a state DOT qualified laboratory to be prequalified to submit. This step would increase the quality and timelines of the service. Having only 30 of the best-qualified firms submitting will reduce the evaluation time required by half and will ensure the capability of all firms that do submit. Our only problem should now be trying to figure out which firm is better. The separation of the best from the better in the newly revised process is a happy problem to have, when previously we suffered from the how-come-this-isn't-working syndrome.

THIS PROCESS HAS MADE staff aware of the positive benefits of a streamlining initiative. Perhaps the most frequent staff comment has centered on the realization that the revision process could have achieved nothing. The staff is becoming aware that working together in a team situation helps create changes without blame and that it's all right to take some risks of failure into a project.

This initial success and the subsequent incremental changes have created a positive environment, fostering a commitment to continuous improvement, at least for this process. For the project managers, the challenge will be to parlay this success into improvements in other areas. ■

Robert S. Lewis (e-mail at wilmette@wilmette.com) has 19 years of experience in civil engineering/project management. He is currently the Village Engineer for Wilmette, Ill. The outsourcing experiences in the story occurred while he was Village Engineer for Lombard, Ill. He is a registered professional engineer in Illinois, Virginia and Florida.

This material has been reproduced with the permission of the copyright owner. Unauthorized reproduction of this material is strictly prohibited. For permission to reproduce this material, please contact PMI.

PM Network • October 1998

Advertisement

Advertisement

Related Content

Advertisement