Focus on Quality and Process Control
Marian Pace, procurement manager for Punta Gorda, Fla., sees the city’s embrace of “Six Sigma” and “Lean” business principles as a way to use new tools to continue the city’s long-time quest for excellence. “We do this all the time, connecting with customers and improving operations,” said Pace. Implementation of new, sophisticated quality management tools has proven “really valuable” for the city, she adds.
“Six Sigma” is a business improvement methodology that uses vigorous processes to systematically eliminate defects and inefficiencies. Originated by Motorola in the 1980s, the system initially focused on manufacturing processes, but it has also become popular in both corporate and small businesses. Along with its cousin “Lean,” it has now moved into governments such as Punta Gorda.
“Lean” principles were an outgrowth of manufacturing approaches used in Japan, specifically the Toyota Production System. While Six Sigma focuses on quality and statistical control, Lean emphasizes process flow, waste and time.
The City Manager in Punta Gorda hired a business management consultant to train city employees on the use of Lean and Six Sigma principles. In the procurement area, application of the principles centered on the p-card program and on “over processing” of payment transactions.
From training to implementation
The term “Six Sigma” coined by Motorola is from a statistical concept describing how a process performs. The term sigma is a Greek letter used to refer to standard deviation, a measure of dispersion in data. Six Sigma is commonly described as a reduction in the variation in a process so that there are no more than 3.4 defects per million opportunities.
The phases of Six Sigma are known as DMAIC – Define, Measure, Analyze, Improve and Control. Punta Gorda employees were challenged to brainstorm ways to increase revenues for the city and to improve efficiencies and processes by reducing costs, waste and non-value added activities. When such an idea for improvement is put through the Six Sigma process, teams are formed and a plan is developed. Team members create a Charter Statement to identify and define the problem and the anticipated goal. Next, the team members gather information about the problem, flowchart processes, compile and analyze the data, then present suggestions for improvement. After improvements are deployed, the team creates new process maps to depict the new procedures and, on a scheduled basis, audit the project to make sure the new procedures are working properly.
The Six Sigma training covered Failure Modes Effect Analysis (FMEA) whereby each member of the group selects either a process or piece of equipment, identifies possible failures, causes of each failure, the impact of the failure, and how frequently the process/equipment should be monitored to prevent failures. The purpose of this exercise is to help a team quickly identify as well as prevent future failures in a system or process. In the next two weeks, project teams comprising mostly of city employees began meeting to work on the approved projects.
The fire department had one of the projects, looking at ways to reduce nonessential calls for service on 911 calls. Marian Pace in the procurement department led the other project, evaluating the city’s payable process and the possible use of procurement cards to improve it.
“Lean production” first referred to the Toyota Production System and arose out of a Massachusetts Institute of Technology (MIT) five-year study on the future of the automobile and the differences in Western and Japanese automobile production. Despite the heavy manufacturing emphasis, Punta Gorda found that Lean is relevant to governments and internal service functions. Although profit and revenue are not cornerstones for evaluating governments’ organizational performance, time often is an important driver in terms of customer and stakeholder satisfaction. The rigorous way in which value, waste and time are treated in Lean concepts provides a good roadmap for service organizations.
Define the problem
The assembled procurement team defined the problem in writing: “The current process for ordering and payment for goods and services is not efficient or cost-effective with the advancement of technology and current procurement best practices. Program transactions in FY09/10 amounted to 7,200-plus transactions and $4.5 million that may have qualified for the p-card/ePayable program.”
Next, the team worked to define the project scope. Payments considered in-scope of the project were purchases under $25,000 and direct journal payments (those not requiring issuance of purchase orders, which are processed directly by Finance). Larger purchases, purchases identified by formal solicitation and construction were considered out-of-scope because of their complexity.
Add quantified objectives
The city saw a revenue opportunity by adopting best practices. At the time, the city did not have a procurement card program. However, neighboring Hillsborough County had a procurement card contract that included incentives and was available for use by Punta Gorda.
Procurement set out to estimate what the revenue opportunities were from rebates based on past expenditure volume. The team analyzed data from the financial system for the past two fiscal years, including a breakdown of payable transactions among types: blanket purchase orders; field purchase orders under $1,000; normal purchase orders; and journal payments. By far, most of the payables transactions were blanket purchase orders (77 percent).
The team used these distributions and trends to forecast the profile of future transactions, both in terms of number of transactions and possible dollar amount of purchases that might be candidates for the procurement card.
Calculating potential rebates, the team estimated that based on current expenditure estimates, it would have a baseline incentive rate of 1.32 percent, worth $49,000. Punta Gorda set two related goals. First, they wanted to achieve an incentive of 1.38 percent by increasing expenditures through the system. The goal considered the forecasted expenditures during the rebate period and the application of expenditure volume to the rebate triggers of the contract (that were based on combined expenditures in the consortium). Second, procurement concluded that if it could achieve “best in class status,” it eventually would achieve an incentive rate of 1.53 percent, worth $57,000. The “best in class” figure was determined by reviewing research and studies conducted by various third party groups.
Use surveys to measure
Part of the data collection involved customer satisfaction, and vendors were also considered customers or stakeholders here.
Procurement used two survey approaches. First, they selected 100 vendors randomly from their database to survey about current practice with respect to acceptance of procurement cards and electronic payments. Second, they selected every 100th vendor in the database (21 total) for a telephone survey.
Procurement received 26 survey responses: 68 percent accepted credit cards; 12 percent accepted electronic payments; and 8 percent did not but were interested in one or the other. Twelve percent were not interested in either. Procurement used the 68 percent estimate to determine the factor in analyzing the total spend qualifying for a potential rebate for purchase orders, blanket purchase orders and field purchase orders. The 12 percent figure was used for analyzing the total spend on Direct Journal payments, a different process that permitted Finance to liquidate direct accounts payable against pre-paid balances on their credit cards accounts, slightly affecting the timing of rebate receipts.
Measure: Dig into processes
Procurement dove deeper into an analysis of its process, using process maps to visually describe the payable process for purchase orders less than $1,500. (See diagram.) The team decided not to include purchase orders arising out of a solicitation process, which are typically a result of a more complex purchasing process and not the high transaction volume type of financial transaction the project was intended to address.
The process map included times for each step. The team sampled transactions identified through the financial system to compute process times and error rates. Sample data was recorded by a stopwatch to measure relevant steps in processes. Once the average times for the processes were computed, the team assigned the respective time to each position. The team had computed an average hourly rate (inclusive of taxes/benefits). The time was multiplied by the average hourly rate to determine the cost for processing a transaction.
Color coding was used on the process map to identify those steps considered “non-value-added” and steps that represented bottlenecks. In some instances, steps were required in the process for auditing. Each step was evaluated on whether the step added any value to customers (i.e., would the customer be willing to pay for it?). If there was non-agreement, a vote was taken among team members and the majority ruled.
For example, the process of returning the purchase order was considered non-value-added. The steps associated with receiving corrected, conforming goods after rejection of nonconforming shipments were considered both non-value-added and bottlenecks. Moreover, the 8 minutes associated with having to print and scan the purchase order was considered non-value-added. Likewise, much of the manual document handling associated with receiving, acceptance and payment was considered non-value-added and identified as bottlenecks.
As a result of the process analysis, the team concluded that the total average time per transaction was 41 minutes, with 12 minutes attributed to receiving and 20 minutes transporting the paperwork to finance.
The invoicing and payment process was similarly mapped. An accountant was part of the team, critical to fully understanding the steps in the process. Again, manual document printing and handling were typically identified as non-value-added. Errors in documentation and required user action, such as submission of receiving reports, were identified as bottlenecks in the process.
One of the seven quality control tools well-known in the quality profession is the Ishikawa diagram, also known as the fishbone diagram. (See diagram on page 54.) The team used the tool to identify through brainstorming the environmental, people, resources, documentation/materials and process factors that may have been root causes of the problems. The process analyzed Receiving and Receipt Processing. At the first level, Level 1, the root cause was identified as a “manual process.” Level 2 analysis drilled down this initial root cause to identify the two final root causes: 1) automated process not in place; and 2) fear of change to current processes.
In the purchase order process, the team found bottlenecks. There were errors in account numbers or inadequate funds assigned on the requisition. In receiving, there was insufficient attention paid to the city’s receiving procedures, and the process was burdened by excessive manual paper handling. There likewise were errors in another fundamental part of the city’s internal controls process – errors and failures to follow procedures in processing invoices.
The team concluded that departments often were not following the procedures in receiving and invoicing. They were not verifying receipt of supplies and services. They were not examining invoices for accuracy or approving duplicate invoices in accordance with city procedures. In some cases, employees were being delegated authority to prepare receiving documents and to approve invoices even though they personally had nothing to with the receipt of the goods or services. In other cases, departments were not adequately training personnel, publishing the required procedures or enforcing accountability. At its core, the absence of automated processes was adversely affecting the receiving and payment processes. Both processes involved excessive, manual paper handling. Moreover, fear of change of processes was identified as a potential risk factor.
After the in-depth view of the problems (in the measure phase) and the analysis to determine the root causes (in analysis phase), the team moved to identifying solutions. The process required setting priorities among competing solutions, evaluating costs and benefits associated with solutions, planning and managing the improvement project, and anticipating and managing the associated change.
Pace’s team used a solution prioritization matrix to identify the best solutions from among a range of possibilities. The team first identified five criteria for the selection: whether they were cost effective (including considerations of the potential rebate); time to implement; process reduction (time and paper); ease of use/implementation; and the level of acceptance by internal and external stakeholders.
The second step was to use a “nominal group technique” on the spreadsheet to determine the weighting of the criteria. Each of the six team members weighted the criteria from 1-5, and the relative weights were assigned using ordinal averages from the team members. In this case, the team members unanimously agreed that cost effectiveness was the most important criteria: it ended up being 33 percent of the weighting.
The third step was to brainstorm ideas for possible solutions. In this case, the team came up with five ideas. “No change” was one. Another was to execute a “linking agreement” based on the county’s procurement card contract. A third was a “non-linking agreement,” meaning rebate amounts would not be based on the total consortium spend. A fourth possibility was for the city to conduct its own request for proposals. A final alternative was to use the city’s current commercial credit card process in the financial system as a direct journal entry to pay Bank of America.
Each team member then ranked (in each of the criteria) the alternatives in order of preference. An Excel spreadsheet was used to average the ordinal ratings and weight them according to the weights selected by the team. The result? “No change” was the least attractive alternative. The number one alternative was to execute a linking agreement based on the county’s procurement card contract.
The team then developed an implementation plan. The city council would have to approve the approach: the agreement and purchase of software to permit procurement card purchase reconciliation with the city’s financial system. The team further forecasted the steps needed for implementation, the pilot training, a pilot period and the training for broader implementation in the city.
The city wanted implementation by the third quarter of the fiscal year because that period accounted for the highest amount of expenditures in the city. Coordination with the information technology department was required because the payables/procurement card reconciliation system had to be installed in the city’s financial system.
To brief the city council in November 2011, the team prepared a cost-benefit analysis, comparing the licensing costs of the payables/procurement card reconciliation software system to the financial benefits from the expected procurement card rebates. Financially, the decision appeared to be an easy one: procurement card rebates were expected to be more than three times the software installation, licensing, training and maintenance costs.
The team also used a “force field” analysis in its presentation, a tool used in strategic planning and familiar to senior executives. (See diagram above.) A slide was used to capture the qualitative considerations in the decision. Forces in support of moving forward on the project included the rebate and increased efficiency in processing, improvements overall to the city’s internal controls and better payment support for the city’s emergency management program. Forces working against the decision to adopt the program included initial perceptions of risk (from misuse) in procurement card programs, resistance to change, and resource and other barriers to effective implementation of the program.
The city council approved the team’s recommendation. The city executed a contract and started the pilot program. In Punta Gorda’s case, the pilot project included purchases by Procurement, Finance, the City Manager’s Office, Information Technology, Wastewater Collections, and Facilities/Parks, in order to include a variety of various fund types. Pace describes the current phase of the project this way, “We are in the improve phase of the DMAIC cycle, writing standard operating procedures. Soon we will be rolling out the new policies and procedures to departments.” By April 2011, the city was midway through its pilot period.
Next: The Control Phase
The final phase is the control phase. Pace expects that final phase also to involve seeking feedback from customers. That phase will use new policies now being approved during the pilot implementation. The city will identify the monetary limits of programs’ controls and delegated authority. The city will look at the individual needs of departments with respect to procurement cards. Then those departments will be trained. After the pilot, the city plans to roll out the procurement card program incrementally citywide.
Pace says that one of the most valuable tools was the solution prioritization matrix. “The solution matrix helped the team cut through the variety of alternatives and select the actions considered most likely to work. The process maps were valuable as well.” According to Pace, “It was amazing to get in a room and see the purchasing process on the process maps. People would look at a step and ask, ‘What is the value?’ We were able to identify a series of non-value-added steps in our process.”
The Ishikawa (fishbone) diagram also impressed Pace. “There was a real ‘wow’ moment using that tool also. The diagrams visually helped the team find the causes of problems, and they had the side benefit of helping address the fear of change.”
“This was such a success, I’ve been asked to facilitate another project looking at the legal advertising and public notice process,” Pace added. “Don’t get me wrong, time is an issue. We are rolling out our procurement card program and implementing a new e-procurement system. But this improvement system really works.”
Contact Pace at firstname.lastname@example.org for more information about the project.