- Email: [email protected]
A multi-level incentive model for service organizations
A multi-level incentive model for service organizations R. L. Shell* and R. G. AIIgeier? *University of Cincinnati, Cincinnati, OH 45221-0072, USA; i-Institute of Advanced Manufacturing Sciences, Inc, Cincinnati, OH 45216, USA Incentive programmes for production employees have been used for many years. However, more recently, there has been an increasing desire to use incentive programmes in service applications, particularly with the advent of electronic performance monitoring. This paper summarizes the methodology for developing a model that incorporates the parameters of quality (customer and supervisor evaluations), quantity of production, sales and individual goals. The model was used to compute incentive bonus payments for customer service employees who performed above standard. Employees were motivated by the incentive bonus, and thus productivity was increased. The model could be applied to any service organization. The model was developed to assist a customer service department where performance standards, quality measures or customer feedback were not firmly established. It measured objectively employee performance and rewarded the employee with individually calculated incentive bonuses through an electronically monitored system. The internal working of the model is a combination of the above parameters. Customized software was used to collect and tabulate customer feedback regarding a recent inquiry to the customer service department. This made up the first part of the quality measure for the department. The second part of the quality parameter was a regular evaluation by the supervisor. The quantity of production was measured by the number of inquiries handled per unit time. Sales were also calculated on a perunit-time basis. The individual goals were the result of regular employee-supervisor meetings where past and future employee performance was discussed and mutually agreed upon. The five model parameters of quality (customer and supervisor), quantity, sales and individual goals are combined through the use of a linear equation. It was found that this form of employee evaluation worked very well in an electronic monitoring environment and was accepted by the participating work group. Overall, monthly productivity gains exceeded 15%.
Keywords : Performance assessment, incentive model, service applications, bonus, productivity
Introduction The US service industries continue to grow with a large number of workers using computers to perform many of their tasks. It has been estimated that about 60 million individuals now use video display terminals at work (Hales, 1989). Present-day computer technology permits easy and low-cost electronic monitoring of workers who are using a video display terminal. Most monitoring systems provide detailed descriptive statistics concerning activity times associated with work output. All of these data, however, are quantitative and in most cases will not provide a complete profile of the work output. For example, conventional electronic monitoring does not evaluate quality of work. There has been considerable controversy over the merits of Vol 23 No 1 February 1992
electronic monitoring - some strongly in favour, others strongly opposed. Computer-aided monitoring is likely to become more common in the workplace, and thus central to understanding the employees' response to work (Chalykoff and Kochan, 1989). The value of monitoring, as summarized by Smith (1988), is given as follows: 'It seems clear that electronic monitoring of employee performance is a necessary element in a competitive, productive work system . . . Proper application of electronic monitoring can enhance job design by building intrinsic motivation into work activities. This is typically accomplished by increasing worker control over the task through worker participation in setting goals and establishing work
0003-6870/92/01 0043-06 $03.00© 1992Butterworth-HeinemannLtd
43
A multi-level incentive model for service organizations standards and by providing feedback that assists the worker in gaining control over task activities, leading to satisfaction from successful performance.' In most service organizations there are enormous amounts of data which can be collected concerning the quantity of production output. Less common, however, are data about the quality of service performed in the business. The motivation of employees in most organizations is improved as financial rewards are increased. Satisfaction of the customer and correctly applying company policies to meet customer needs should also be criteria used for employee performance evaluation. With each interaction, the customer service representative has considerable influence on the customer and consequently a direct impact on the revenue generated and/or lost. An investigation into the motivating factors of customer service representatives was conducted. This investigation revealed that the data already being electronically collected needed to be transformed into production and quality indicators for their department. The analysis of the results of this investigation was followed by the installation of a Quality Control/ Incentive Programme for the department. This programme included the parameters of quality (customer and peer), quantity of production (transactions per hour), sales revenue, and personal goals and objectives. Since there were no standards for performance evaluation, quality measures, customer feedback and production, a programme to measure and monitor these parameters objectively had to be developed.
Objective
the literature or from business to business (Tuzcu, 1983). Some companies worldwide use this type of system to obtain a real measure of how the company is performing by combining several parameters for different parts of the business. Using a single incentive parameter can bias the calculation of incentive bonuses compared with the true value of the effort, and cause the worker to sub-optimize his/her positive contribution to the business. Incentive programmes should be implemented only to measure and reward for productivity gains. Productivity, therefore, is not another interpretation of the wage structure of the business, but a measurement of how well the capabilities of an employee help the company to meet its defined goals. The financial rewards should be linked consistently and fairly to productivity gains of the individual, department or entire company. Before implementing a successful incentive programme, the company should follow a defined sequence of events (Smith, 1984). • Select a team of qualified employees comprising management, engineering, supervision and general employee representatives. • Establish a clear and concise definition of the objectives of the system. A means to document the 'as is' of the organization and a means of evaluating the effectiveness of the system are included in this step. • Select the appropriate parameters for inclusion in the incentive calculation. With this, the determination of how the data will be gathered and the initial formula are established. This is commonly the most difficult task of the entire process.
Our purpose is to demonstrate the conceptual model and the methodology of incorporating therein the parameters of quality, quantity, sales and personal goals. The conceptual model to be presented is not restricted to a particular company. With proper analysis of the business objectives, this model can be applied to other computer-based customer services and sales organizations with electronic monitoring capability.
• Determine the method of training the supervisors about how the programme will be implemented after the approval of the parameters. Test data will be gathered and calculations will be thoroughly examined for accuracy.
State of the art
• Evaluate the effectiveness of the system. A formal audit should be conducted to determine the adherence to the defined objective.
In companies where customer service is involved, much of the transaction relating to the customer, once completed, is lost forever. The data pertaining to that transaction have to be clear and concise, for there is usually no chance for starting again or correction at a later date. Incentive programmes for the service sector require both a firm commitment by management and active involvement of management with their employees. They must work as a team member for the mutual benefit of both themselves and employees. Management must have a defined set of objectives and back them with appropriate policies and procedures. Input from involved employees is essential. Electronic monitoring must be used to provide timely data about worker performance in a positive way and not to harass individuals for more output without fair rewards. At the present time, multi-factor wage incentive programmes are used, but are not discussed openly in
44
• Collect the data. The system is implemented and the initial payments are made to employees based on all the above criteria.
The combination of a proper set of parameters, generous incentives and management support can yield a closer and more personal relationship with the customer. This can return savings or increased earnings many times over for the company. On the other hand, poor design, inefficient utilization of the system and improper management can lead to an overall negative impact on the company. This is particularly true for electronically monitored systems that are not used in a fair and positive manner. A negative effect in the company begins with employee resistance to the changes and measurement practices implemented. As a final note, a successful incentive programme should reward the employees for improved performance with an amount of at least 15-25% of their base pay (Shell, 1986; Reed, 1981). Better-performing employees should be rewarded on the upper end of the
Applied Ergonomics
R. L. SHELL AND R. G. ALLGEIER ! Employee
,rJ
repq~rts
I
Manual inputof other parameters
Quality assurance 'BASIC' program
I Employee data =
~j J Other reports
to rate the telephone employees who actually handled their inquiry, question or problem. The customer also had an opportunity to have any additional concerns resolved. This interaction constituted the customer input of the quality portion of the equation to be derived later. The customized program also collected information useful to other departments of the company. Figure 2 shows the flowchart of the BASIC program indicating the reports which are used for other aspects of the business.
J Oth, qua ity
file
J
file
'Lotus~
I
J
ncentive pay file "J 'Lotus'
Figure 1 Interaction of program files within the Quality Control/Incentive Programme
scale and new employees should be instructed how to increase their productivity to earn an incentive bonus. The incentive programme
The Quality Control/Incentive Programme, as installed, utilized customized computer software programs written in both BASIC and LOTUS 123. These were used to combine several files to arrive at the payment of incentives to the individual service representatives (Figure 1). The customized program was used for the collection, tabulation and output of customer's responses to a quality assurance questionnaire. This 'on-line' questionnaire was designed to give the customer a chance
The LOTUS 123 software was utilized to collect and store employee identification, and quality, quantity, revenue and personal goals indices, as well as other supportive information (Figure 3). Within the LOTUS program, the equations were programmed for the actual calculation of the percentage of incentives to be paid. The information collected by the customized BASIC program was automatically translated into LOTUS readable format and electronically transferred. The incentive calculation retrieved the information collected by the customized program and compiled it with the following formula to calculate the incentive points index. IP
= a(SPI) + b ( M P I ) + c ( C P H ) (E) + d(SVI) + e ( P G I ) . . .
(1) where a, b, c, d and e are weighting coefficients adding to 100, and IP = Incentive Points S P I = Survey Points Index (quality rating by the customer) M P I -- Monitor Points Index (quality rating by the supervisor)
( I ) Data entry / (2) Reports (3) Maintenance (4) Exit 4
3 DATA ENTRY Ask questionnaire I I and store datain I I on-line files
I (I) TSR I (2) Confused I (3) Sales I (4) Service
I ,
MAINTENANCE Delete all records from post questionnaires
EXIT stop
within a specified
range
I (61 General I 17) Exit
.o
,I
4'
42
TSR Manipulates data for
CONFUSED Information
SALES Sales information
I 11 II productivity
and quality
which was confusing to customer
4 5
43 SERVICE
Information
pertaining to field
INSTALLER Installer
information
service
4 GENERAL Other company information
Figure 2 Flowchart of the BASIC program
Vol 23 No 1 February 1992
45
A multi-level incentive model for service organizations CPH = Transactions
PGI is an indicator of how the employees rate them-
Per Hour (quantity of production) E = Errors on data entry (as related to data entry) SVI = Sales Volume Index (revenue generated) PGI = Personal Goals (monthly goals set jointly by employee and supervisor)
selves. Each month the supervisor and employee meet to set the targets or goals for the period. This may include the following; number of transactions per hour, sales per hour, etc. These data are collected on a paper form and the rating is based on how many of the goals are met. The maximum for this parameter is 11. The coefficients of Equation (1), being whole numbers, are the actual percentages which each parameter is actually weighted. Again, the coefficients must add to 100.
The individual terms of this equation represent a number between 0 and 1. These terms are determined by dividing the individual average by the maximum allowable for that term.
SPI is the numerical average of all responses to the
The equation for calculating incentives had to be restricted to as few terms as possible. This was important to ensure that all parameters were meaningful. Too many parameters add unnecessary complexity and too few parameters do not reflect all of the vital components of the system. It was found that including too many terms increased the amount of 'noise' in the system and reduced the significance of the individual terms. Also, parameters with weighting coefficients of 5% or less had little significance for the entire program.
customer evaluation. Each question has a rating of 1 to 10. These are averaged to show a composite rating for that customer. The maximum for this term is therefore 10.
MPI is a numerical representation of the supervisor evaluation. They rate the employee on a scale of 0 to 1. These data are collected on a paper form and the composite percentage is manually entered into the computer for tabulation. The maximum for this term is 1.
The actual equation used to calculate incentive points is as follows:
CPH is the number of transactions per hour actually taken by a representative minus the number of transactions transferred to other representatives minus the number of outgoing transactions made. This information is automatically collected through the electonic monitoring system. The maximum for this term was determined by analyzing historical data. The range of this parameter is 0 to 25.
SPI [SPI(=~,)]
E is the number of on-line errors in entering customer work orders. This could be a wrong service code, a request for the wrong service, etc. This is reported to management in an error report which identifies the customer representative and the date of transaction. This factor is a multiplicative factor because the data are electronically entered into the terminal as the telephone conversation continues.
The incentive earnings percentage was based on a 15-40 concept, where the average performer earns 15% and the top performer earns the maximum bonus of 40%. The 40% upper limit was set by agreement between management and employees to reward a top performer for exceptional performance.
is an upward change in service, or new service. This is automatically reported to management, and the representative also receives a percentage of the sale. Survey
The calculation of individual incentives was determined by averaging the values for all employees and inserting these averages into Equation (2). To detert
Earned Pouible Monitor monitor monitor point= pointl point= index
Timedle=t Logged houri houri ~-
Nome ID Super
Calls Index
Contoined in o LOTUS file. Then ore outomotiually transferred to this file using the BASIC program
Compiled in Manually collected on the monitoring form. t "~ the BASIC program and Earned and polsible points are monually outomotically input Qnd the index tron=forred to this file I " i s colculoted.
/ I i
t
t
Colculoted from the two prt~ious column=.
1
t
I i
Compiled in individual repoct=. Automatically troneforred to this I file f ""
Overall time
individual reoor~. Automatically tronsferred to this file
1
Totol In Out Trons calls
ICPH(=~,)]
Incentive earnings (IE) are calculated from the incentive points (IP) and will be explained later.
SVI is the number of sales generated each hour. A sale
dl,of Point=
MPI IMPI(=.) ]
Calls per
hour Calculated from ~total coils ~ond logged hours
Soles volume units
Soles ~ volume ..~ index
Au~"rnoticaily collected and manually input
# of
=14=of ob=encN tord~
Compiled form Colouloted payroll and from soles volume monually entered to and logged hours this file
i
i
Add error
Gcab and objective=
Manually Input from tolly
Manually input from tally
sheet
sheet
Figure 3 LOTUS worksheet for collection of all inputs
46
Applied Ergonomics
R. L. S H E L L A N D R. G. ALLGEIER
mine the slope and Y-intercept for the line used to determine the individual incentive earnings, the simultaneous solution of two equations and two unknowns was obtained: 100 -- 4 0 m + b , = 15m + b .
I E M o a t h AveraS e
.
and . . . ( 3 ) . . (4)
The slope and Y-intercept are recalculated each month to adjust for any change in productivity of the department. As the slope increases, an employee has to perform better in order to begin earning incentive payment. The conversion from incentive points to incentive earning was obtained through the following equation: I E = m(IP) - b
...(5)
where I P is calculated from Equation (2), and b and m are calculated from the simultaneous solution of Equations (3) and (4). If I E is less than 0, there was no bonus paid and the employee receives base wages for the period. An example is now presented to explain the concept. Assume that earning 100 points will reward the employee with 40% of his/her base pay as the incentive earning. Also, the average for all parameters will yield 15% earnings. These are the two points used to determine the earnings line. For illustrative purposes, an average of 65 points is presented in Figure 4. As the average productivity of the department increases, the slope also increases. This is demonstrated by moving the average from 65 to 75 incentive points (Figure 5). The above analysis is also true for the decrease in the average output of the employees. If the average drops, there is a possibility of more employees earning an incentive bonus. Prior to Quality Control/Incentive Programme installation, there was an electronic monitor report produced and used for the indication of individual output. This report tracked information relating only to number of transactions handled (calls per minute) and the duration of those transactions. There was no information collected regarding quality of transactions with the customer, or a rating scale which the supervisor used to evaluate employee performance. Furthermore, sales, which were an important part of the transaction, were not credited to the representative. Employees were told only of their quantitative output
with no opportunity to provide feedback. The Quality Control/Incentive Programme was the initial step towards rewarding the workers for good performance, quality improvement and increased production.
Evaluation The data used for determining the effectiveness of the programme were obtained from the document used for incentive bonus payments. For complete analysis of this system, a baseline of data in all measurement areas was required. The parameters of 'Transactions Per Hour' and 'Sales Per Hour' were determined from historical data. The Survey Points and Monitor Points history could not be verified since these parameters were developed as part of the implementation of this programme. The data were normalized term-by-term to a number between 0 and 1 by dividing the monthly average by the maximum for the period. The maximum values are presented in Table 1. An example of the ranges of the values for a given month in the department is shown in Table 2. Included in this table, on the far right, is an indication of the percentage of the base wage an employee would earn. This means that an employee with a base wage of $6.00/hour and earning a 25% bonus would effectively be working for $7.50/hour for the entire month. Overall productivity gains exceeded 15%.
Table I Maximum equation values for department and time used Parameter Survey points Monitor points Transactions per hour Sales per hour
Personal goals
Time used (Months)
10.0 1.0 20.0 17.0 3.5 3.0 3.5 11.0
22 22 1 21 1 3 17 22
Conclusions In this investigation, the employees were subjected to an environment that was uncertain in the minute-to!
40
40
3O
.~ 3o
c
Maximum
b 2O c
c
~o I
I0
I
20
I
30
I
40 50 60 Incentivepoints
I 70
80
90
I00
Figure 4 Incentive calculation plot showing 65 points average earning. The slope of the line is 0.71 and the Y intercept is - 3 1 . This indicates the need to earn at least 44 points to begin earning an incentive bonus
Vol 23 No 1 February 1992
I I0
I 20
I 30
I
,r"
40 50 60 Incentivepoints
I
70
I
I
80
90
I00
Figure 5 Incentive calculation plot showing 75 points average earning and change in earning slope. The slope of the line is 1.0 and the Y-intercept is -60. This indicates the need to earn at least 60 points to begin earning an incentive bonus.
47
A multi-level incentive model for service organizations Table 2 Typical month of collected data for department ID
SPI
MPI
CPH
SVI
PGI
IP
IE
37 49 61 48 43 11 35 39 55 101 45 99 14 15 52 60 33 50 107 95 10 77
9.70 10.00 9.50 9.30 8.30 8.60 8.60 8.60 9.80 8.80 9.20 9.30 9.30 9.50 8.80 8.60 9.30 9.30 9.60 8.80 9.30 9.30
0.94 0.96 0.91 0.92 0.76 0.92 0.92 0.88 0.94 0.92 0.88 0.91 0.76 0.93 0.89 0.87 0.78 0.95 0.80 0.90 0.78 0.85
15.76 14.50 16.61 14.54 18.49 16.64 15.19 16.66 12.68 14.18 16.19 14.26 14.99 13.99 15.46 14.81 15.09 13.19 13.31 14.02 15.17 1.06
2.60 2.61 1.74 2.45 1.85 1.79 2.50 2.06 1.73 2.10 1.76 1.94 2.16 1.93 1.60 1.86 1.88 0.97 1.10 1.33 1.03 0.80
5.00 5.00 1.00 4.50 5.00 4.00 0.50 3.00 2.00 4.50 1.50 0.75 2.00 0.75 0.75 0.75 0.25 1.50 2.00 0.00 3.25 0.50
92 84 82 81 79 79 78 79 77 77 75 74 74 74 72 72 71 71 69 67 62 51
32 25 23 21 20 20 19 19 18 18 16 16 15 15 13 13 12 12 10 9 3 0
0.88
14.40
1.81
2.20
Average 9.16
74.55 15.86
minute activities. Randomness in routing the telephone inquiry distributed any activity to any individual representative. With the new system each employee was constantly aware that they controlled the 'cash register' of the company. Also, they accepted the responsibility for treating the customer accurately and fairly in all responses. Management was doing its part to ensure
48
that this was happening by installing a system which combined the parameters of quality, quantity, production, sales and personal goals to reward the representative for performing tasks according to the objectives of the department. This incentive system has been working for over two years with increased productivity and no adverse effects on the operation of the department. This investigation has shown that the proper use of electronic monitoring along with a fair and equitable incentive system can be advantageous to both management and the employees. In addition, a higher quality of service was provided to the customer along with increased overall productivity. References
Chalykoff, J., and Kochan, T. A. 1989, 'Computer-Aided Monitoring: Its Influence on Employee Job Satisfaction and Turnover.' MIT Industrial Liaison Program Report, 4-4-89. Massachusetts Institute of Technology, Cambridge, MA. Hales, S. W. 1989, Occupat Health and Safety, 58.7, 21. Safe VDTs. Reed, S. K. 1981, 'A New Look At Wage Incentives And Their Effect On Productivity.' Spring Annual Industrial Engineering Conference Proceedings, 17-20 May, Institute of Industrial Engineers, Detroit. Shell, R. L. 1986, 'Work Measurement: Principles and Pracrice.' Industrial Engineering and Management Press, Norcross, Atlantic GA. Smith, C. 1984, Indust Enging, 16. 1, 82-86. Awareness, analysis and improvement are keys to white collar productivity. Smith, M. J. 1988, Bull Human Factors Soc, 31. 2, 1-3. Electronic performance monitoring at the workplace: Part of a new industrial revolution. Tuzcu, E. 1983, 'Group Incentives In Office Operations.' Fall Industrial Engineering Conference Proceedings, 13-16 Nov, Institute of Industrial Engineers, Toronto.
Applied Ergonomics
Keywords : Performance assessment, incentive model, service applications, bonus, productivity
Introduction The US service industries continue to grow with a large number of workers using computers to perform many of their tasks. It has been estimated that about 60 million individuals now use video display terminals at work (Hales, 1989). Present-day computer technology permits easy and low-cost electronic monitoring of workers who are using a video display terminal. Most monitoring systems provide detailed descriptive statistics concerning activity times associated with work output. All of these data, however, are quantitative and in most cases will not provide a complete profile of the work output. For example, conventional electronic monitoring does not evaluate quality of work. There has been considerable controversy over the merits of Vol 23 No 1 February 1992
electronic monitoring - some strongly in favour, others strongly opposed. Computer-aided monitoring is likely to become more common in the workplace, and thus central to understanding the employees' response to work (Chalykoff and Kochan, 1989). The value of monitoring, as summarized by Smith (1988), is given as follows: 'It seems clear that electronic monitoring of employee performance is a necessary element in a competitive, productive work system . . . Proper application of electronic monitoring can enhance job design by building intrinsic motivation into work activities. This is typically accomplished by increasing worker control over the task through worker participation in setting goals and establishing work
0003-6870/92/01 0043-06 $03.00© 1992Butterworth-HeinemannLtd
43
A multi-level incentive model for service organizations standards and by providing feedback that assists the worker in gaining control over task activities, leading to satisfaction from successful performance.' In most service organizations there are enormous amounts of data which can be collected concerning the quantity of production output. Less common, however, are data about the quality of service performed in the business. The motivation of employees in most organizations is improved as financial rewards are increased. Satisfaction of the customer and correctly applying company policies to meet customer needs should also be criteria used for employee performance evaluation. With each interaction, the customer service representative has considerable influence on the customer and consequently a direct impact on the revenue generated and/or lost. An investigation into the motivating factors of customer service representatives was conducted. This investigation revealed that the data already being electronically collected needed to be transformed into production and quality indicators for their department. The analysis of the results of this investigation was followed by the installation of a Quality Control/ Incentive Programme for the department. This programme included the parameters of quality (customer and peer), quantity of production (transactions per hour), sales revenue, and personal goals and objectives. Since there were no standards for performance evaluation, quality measures, customer feedback and production, a programme to measure and monitor these parameters objectively had to be developed.
Objective
the literature or from business to business (Tuzcu, 1983). Some companies worldwide use this type of system to obtain a real measure of how the company is performing by combining several parameters for different parts of the business. Using a single incentive parameter can bias the calculation of incentive bonuses compared with the true value of the effort, and cause the worker to sub-optimize his/her positive contribution to the business. Incentive programmes should be implemented only to measure and reward for productivity gains. Productivity, therefore, is not another interpretation of the wage structure of the business, but a measurement of how well the capabilities of an employee help the company to meet its defined goals. The financial rewards should be linked consistently and fairly to productivity gains of the individual, department or entire company. Before implementing a successful incentive programme, the company should follow a defined sequence of events (Smith, 1984). • Select a team of qualified employees comprising management, engineering, supervision and general employee representatives. • Establish a clear and concise definition of the objectives of the system. A means to document the 'as is' of the organization and a means of evaluating the effectiveness of the system are included in this step. • Select the appropriate parameters for inclusion in the incentive calculation. With this, the determination of how the data will be gathered and the initial formula are established. This is commonly the most difficult task of the entire process.
Our purpose is to demonstrate the conceptual model and the methodology of incorporating therein the parameters of quality, quantity, sales and personal goals. The conceptual model to be presented is not restricted to a particular company. With proper analysis of the business objectives, this model can be applied to other computer-based customer services and sales organizations with electronic monitoring capability.
• Determine the method of training the supervisors about how the programme will be implemented after the approval of the parameters. Test data will be gathered and calculations will be thoroughly examined for accuracy.
State of the art
• Evaluate the effectiveness of the system. A formal audit should be conducted to determine the adherence to the defined objective.
In companies where customer service is involved, much of the transaction relating to the customer, once completed, is lost forever. The data pertaining to that transaction have to be clear and concise, for there is usually no chance for starting again or correction at a later date. Incentive programmes for the service sector require both a firm commitment by management and active involvement of management with their employees. They must work as a team member for the mutual benefit of both themselves and employees. Management must have a defined set of objectives and back them with appropriate policies and procedures. Input from involved employees is essential. Electronic monitoring must be used to provide timely data about worker performance in a positive way and not to harass individuals for more output without fair rewards. At the present time, multi-factor wage incentive programmes are used, but are not discussed openly in
44
• Collect the data. The system is implemented and the initial payments are made to employees based on all the above criteria.
The combination of a proper set of parameters, generous incentives and management support can yield a closer and more personal relationship with the customer. This can return savings or increased earnings many times over for the company. On the other hand, poor design, inefficient utilization of the system and improper management can lead to an overall negative impact on the company. This is particularly true for electronically monitored systems that are not used in a fair and positive manner. A negative effect in the company begins with employee resistance to the changes and measurement practices implemented. As a final note, a successful incentive programme should reward the employees for improved performance with an amount of at least 15-25% of their base pay (Shell, 1986; Reed, 1981). Better-performing employees should be rewarded on the upper end of the
Applied Ergonomics
R. L. SHELL AND R. G. ALLGEIER ! Employee
,rJ
repq~rts
I
Manual inputof other parameters
Quality assurance 'BASIC' program
I Employee data =
~j J Other reports
to rate the telephone employees who actually handled their inquiry, question or problem. The customer also had an opportunity to have any additional concerns resolved. This interaction constituted the customer input of the quality portion of the equation to be derived later. The customized program also collected information useful to other departments of the company. Figure 2 shows the flowchart of the BASIC program indicating the reports which are used for other aspects of the business.
J Oth, qua ity
file
J
file
'Lotus~
I
J
ncentive pay file "J 'Lotus'
Figure 1 Interaction of program files within the Quality Control/Incentive Programme
scale and new employees should be instructed how to increase their productivity to earn an incentive bonus. The incentive programme
The Quality Control/Incentive Programme, as installed, utilized customized computer software programs written in both BASIC and LOTUS 123. These were used to combine several files to arrive at the payment of incentives to the individual service representatives (Figure 1). The customized program was used for the collection, tabulation and output of customer's responses to a quality assurance questionnaire. This 'on-line' questionnaire was designed to give the customer a chance
The LOTUS 123 software was utilized to collect and store employee identification, and quality, quantity, revenue and personal goals indices, as well as other supportive information (Figure 3). Within the LOTUS program, the equations were programmed for the actual calculation of the percentage of incentives to be paid. The information collected by the customized BASIC program was automatically translated into LOTUS readable format and electronically transferred. The incentive calculation retrieved the information collected by the customized program and compiled it with the following formula to calculate the incentive points index. IP
= a(SPI) + b ( M P I ) + c ( C P H ) (E) + d(SVI) + e ( P G I ) . . .
(1) where a, b, c, d and e are weighting coefficients adding to 100, and IP = Incentive Points S P I = Survey Points Index (quality rating by the customer) M P I -- Monitor Points Index (quality rating by the supervisor)
( I ) Data entry / (2) Reports (3) Maintenance (4) Exit 4
3 DATA ENTRY Ask questionnaire I I and store datain I I on-line files
I (I) TSR I (2) Confused I (3) Sales I (4) Service
I ,
MAINTENANCE Delete all records from post questionnaires
EXIT stop
within a specified
range
I (61 General I 17) Exit
.o
,I
4'
42
TSR Manipulates data for
CONFUSED Information
SALES Sales information
I 11 II productivity
and quality
which was confusing to customer
4 5
43 SERVICE
Information
pertaining to field
INSTALLER Installer
information
service
4 GENERAL Other company information
Figure 2 Flowchart of the BASIC program
Vol 23 No 1 February 1992
45
A multi-level incentive model for service organizations CPH = Transactions
PGI is an indicator of how the employees rate them-
Per Hour (quantity of production) E = Errors on data entry (as related to data entry) SVI = Sales Volume Index (revenue generated) PGI = Personal Goals (monthly goals set jointly by employee and supervisor)
selves. Each month the supervisor and employee meet to set the targets or goals for the period. This may include the following; number of transactions per hour, sales per hour, etc. These data are collected on a paper form and the rating is based on how many of the goals are met. The maximum for this parameter is 11. The coefficients of Equation (1), being whole numbers, are the actual percentages which each parameter is actually weighted. Again, the coefficients must add to 100.
The individual terms of this equation represent a number between 0 and 1. These terms are determined by dividing the individual average by the maximum allowable for that term.
SPI is the numerical average of all responses to the
The equation for calculating incentives had to be restricted to as few terms as possible. This was important to ensure that all parameters were meaningful. Too many parameters add unnecessary complexity and too few parameters do not reflect all of the vital components of the system. It was found that including too many terms increased the amount of 'noise' in the system and reduced the significance of the individual terms. Also, parameters with weighting coefficients of 5% or less had little significance for the entire program.
customer evaluation. Each question has a rating of 1 to 10. These are averaged to show a composite rating for that customer. The maximum for this term is therefore 10.
MPI is a numerical representation of the supervisor evaluation. They rate the employee on a scale of 0 to 1. These data are collected on a paper form and the composite percentage is manually entered into the computer for tabulation. The maximum for this term is 1.
The actual equation used to calculate incentive points is as follows:
CPH is the number of transactions per hour actually taken by a representative minus the number of transactions transferred to other representatives minus the number of outgoing transactions made. This information is automatically collected through the electonic monitoring system. The maximum for this term was determined by analyzing historical data. The range of this parameter is 0 to 25.
SPI [SPI(=~,)]
E is the number of on-line errors in entering customer work orders. This could be a wrong service code, a request for the wrong service, etc. This is reported to management in an error report which identifies the customer representative and the date of transaction. This factor is a multiplicative factor because the data are electronically entered into the terminal as the telephone conversation continues.
The incentive earnings percentage was based on a 15-40 concept, where the average performer earns 15% and the top performer earns the maximum bonus of 40%. The 40% upper limit was set by agreement between management and employees to reward a top performer for exceptional performance.
is an upward change in service, or new service. This is automatically reported to management, and the representative also receives a percentage of the sale. Survey
The calculation of individual incentives was determined by averaging the values for all employees and inserting these averages into Equation (2). To detert
Earned Pouible Monitor monitor monitor point= pointl point= index
Timedle=t Logged houri houri ~-
Nome ID Super
Calls Index
Contoined in o LOTUS file. Then ore outomotiually transferred to this file using the BASIC program
Compiled in Manually collected on the monitoring form. t "~ the BASIC program and Earned and polsible points are monually outomotically input Qnd the index tron=forred to this file I " i s colculoted.
/ I i
t
t
Colculoted from the two prt~ious column=.
1
t
I i
Compiled in individual repoct=. Automatically troneforred to this I file f ""
Overall time
individual reoor~. Automatically tronsferred to this file
1
Totol In Out Trons calls
ICPH(=~,)]
Incentive earnings (IE) are calculated from the incentive points (IP) and will be explained later.
SVI is the number of sales generated each hour. A sale
dl,of Point=
MPI IMPI(=.) ]
Calls per
hour Calculated from ~total coils ~ond logged hours
Soles volume units
Soles ~ volume ..~ index
Au~"rnoticaily collected and manually input
# of
=14=of ob=encN tord~
Compiled form Colouloted payroll and from soles volume monually entered to and logged hours this file
i
i
Add error
Gcab and objective=
Manually Input from tolly
Manually input from tally
sheet
sheet
Figure 3 LOTUS worksheet for collection of all inputs
46
Applied Ergonomics
R. L. S H E L L A N D R. G. ALLGEIER
mine the slope and Y-intercept for the line used to determine the individual incentive earnings, the simultaneous solution of two equations and two unknowns was obtained: 100 -- 4 0 m + b , = 15m + b .
I E M o a t h AveraS e
.
and . . . ( 3 ) . . (4)
The slope and Y-intercept are recalculated each month to adjust for any change in productivity of the department. As the slope increases, an employee has to perform better in order to begin earning incentive payment. The conversion from incentive points to incentive earning was obtained through the following equation: I E = m(IP) - b
...(5)
where I P is calculated from Equation (2), and b and m are calculated from the simultaneous solution of Equations (3) and (4). If I E is less than 0, there was no bonus paid and the employee receives base wages for the period. An example is now presented to explain the concept. Assume that earning 100 points will reward the employee with 40% of his/her base pay as the incentive earning. Also, the average for all parameters will yield 15% earnings. These are the two points used to determine the earnings line. For illustrative purposes, an average of 65 points is presented in Figure 4. As the average productivity of the department increases, the slope also increases. This is demonstrated by moving the average from 65 to 75 incentive points (Figure 5). The above analysis is also true for the decrease in the average output of the employees. If the average drops, there is a possibility of more employees earning an incentive bonus. Prior to Quality Control/Incentive Programme installation, there was an electronic monitor report produced and used for the indication of individual output. This report tracked information relating only to number of transactions handled (calls per minute) and the duration of those transactions. There was no information collected regarding quality of transactions with the customer, or a rating scale which the supervisor used to evaluate employee performance. Furthermore, sales, which were an important part of the transaction, were not credited to the representative. Employees were told only of their quantitative output
with no opportunity to provide feedback. The Quality Control/Incentive Programme was the initial step towards rewarding the workers for good performance, quality improvement and increased production.
Evaluation The data used for determining the effectiveness of the programme were obtained from the document used for incentive bonus payments. For complete analysis of this system, a baseline of data in all measurement areas was required. The parameters of 'Transactions Per Hour' and 'Sales Per Hour' were determined from historical data. The Survey Points and Monitor Points history could not be verified since these parameters were developed as part of the implementation of this programme. The data were normalized term-by-term to a number between 0 and 1 by dividing the monthly average by the maximum for the period. The maximum values are presented in Table 1. An example of the ranges of the values for a given month in the department is shown in Table 2. Included in this table, on the far right, is an indication of the percentage of the base wage an employee would earn. This means that an employee with a base wage of $6.00/hour and earning a 25% bonus would effectively be working for $7.50/hour for the entire month. Overall productivity gains exceeded 15%.
Table I Maximum equation values for department and time used Parameter Survey points Monitor points Transactions per hour Sales per hour
Personal goals
Time used (Months)
10.0 1.0 20.0 17.0 3.5 3.0 3.5 11.0
22 22 1 21 1 3 17 22
Conclusions In this investigation, the employees were subjected to an environment that was uncertain in the minute-to!
40
40
3O
.~ 3o
c
Maximum
b 2O c
c
~o I
I0
I
20
I
30
I
40 50 60 Incentivepoints
I 70
80
90
I00
Figure 4 Incentive calculation plot showing 65 points average earning. The slope of the line is 0.71 and the Y intercept is - 3 1 . This indicates the need to earn at least 44 points to begin earning an incentive bonus
Vol 23 No 1 February 1992
I I0
I 20
I 30
I
,r"
40 50 60 Incentivepoints
I
70
I
I
80
90
I00
Figure 5 Incentive calculation plot showing 75 points average earning and change in earning slope. The slope of the line is 1.0 and the Y-intercept is -60. This indicates the need to earn at least 60 points to begin earning an incentive bonus.
47
A multi-level incentive model for service organizations Table 2 Typical month of collected data for department ID
SPI
MPI
CPH
SVI
PGI
IP
IE
37 49 61 48 43 11 35 39 55 101 45 99 14 15 52 60 33 50 107 95 10 77
9.70 10.00 9.50 9.30 8.30 8.60 8.60 8.60 9.80 8.80 9.20 9.30 9.30 9.50 8.80 8.60 9.30 9.30 9.60 8.80 9.30 9.30
0.94 0.96 0.91 0.92 0.76 0.92 0.92 0.88 0.94 0.92 0.88 0.91 0.76 0.93 0.89 0.87 0.78 0.95 0.80 0.90 0.78 0.85
15.76 14.50 16.61 14.54 18.49 16.64 15.19 16.66 12.68 14.18 16.19 14.26 14.99 13.99 15.46 14.81 15.09 13.19 13.31 14.02 15.17 1.06
2.60 2.61 1.74 2.45 1.85 1.79 2.50 2.06 1.73 2.10 1.76 1.94 2.16 1.93 1.60 1.86 1.88 0.97 1.10 1.33 1.03 0.80
5.00 5.00 1.00 4.50 5.00 4.00 0.50 3.00 2.00 4.50 1.50 0.75 2.00 0.75 0.75 0.75 0.25 1.50 2.00 0.00 3.25 0.50
92 84 82 81 79 79 78 79 77 77 75 74 74 74 72 72 71 71 69 67 62 51
32 25 23 21 20 20 19 19 18 18 16 16 15 15 13 13 12 12 10 9 3 0
0.88
14.40
1.81
2.20
Average 9.16
74.55 15.86
minute activities. Randomness in routing the telephone inquiry distributed any activity to any individual representative. With the new system each employee was constantly aware that they controlled the 'cash register' of the company. Also, they accepted the responsibility for treating the customer accurately and fairly in all responses. Management was doing its part to ensure
48
that this was happening by installing a system which combined the parameters of quality, quantity, production, sales and personal goals to reward the representative for performing tasks according to the objectives of the department. This incentive system has been working for over two years with increased productivity and no adverse effects on the operation of the department. This investigation has shown that the proper use of electronic monitoring along with a fair and equitable incentive system can be advantageous to both management and the employees. In addition, a higher quality of service was provided to the customer along with increased overall productivity. References
Chalykoff, J., and Kochan, T. A. 1989, 'Computer-Aided Monitoring: Its Influence on Employee Job Satisfaction and Turnover.' MIT Industrial Liaison Program Report, 4-4-89. Massachusetts Institute of Technology, Cambridge, MA. Hales, S. W. 1989, Occupat Health and Safety, 58.7, 21. Safe VDTs. Reed, S. K. 1981, 'A New Look At Wage Incentives And Their Effect On Productivity.' Spring Annual Industrial Engineering Conference Proceedings, 17-20 May, Institute of Industrial Engineers, Detroit. Shell, R. L. 1986, 'Work Measurement: Principles and Pracrice.' Industrial Engineering and Management Press, Norcross, Atlantic GA. Smith, C. 1984, Indust Enging, 16. 1, 82-86. Awareness, analysis and improvement are keys to white collar productivity. Smith, M. J. 1988, Bull Human Factors Soc, 31. 2, 1-3. Electronic performance monitoring at the workplace: Part of a new industrial revolution. Tuzcu, E. 1983, 'Group Incentives In Office Operations.' Fall Industrial Engineering Conference Proceedings, 13-16 Nov, Institute of Industrial Engineers, Toronto.
Applied Ergonomics