A Model for Calculating Training ROI and the Value of Human Capital
The model is the product of a careful sifting of many ROI models and processes. Each model has its particular strengths and weaknesses. Kearsley (1986) presents a number of approaches for calculating ROI, all with their own form of outcomes. What we present here is an operational model that draws primarily from the work of Schneider and Wright (1990), Schneider, Monetta and Wright (1992), Robinson and Robinson (1990, 1992), Spencer (1984) and Stolovitch and Keeps (1995). The figure on the next page graphically portrays the model and its steps.
The model consists of seven major steps. Below, we describe each step and provide additional explanations and examples as necessary.
To perform this step requires conducting a front-end analysis (Harless, 1970; Rossett, 1987, 1992; Stolovitch and Keeps, 1997). In brief, front-end analysis consists of the following steps:
1. Determine purpose of training request (mandated; new system; performance improvement).
2. Identify desired performance and, if relevant, actual performance.
3. Identify feelings related to the desired performance.
4. Identify causes for not achieving desired performance if this is a performance improvement issue.
5. Identify appropriate solutions that are economically relevant, feasible and acceptable to both organization and targeted performers.
Through careful investigation, the front-end analysis identifies the magnitude, urgency and value of the performance gap as well as the factors affecting it (environmental; skills/knowledge; emotional/political). It also proposes a basket of intervention solutions that are most appropriate, economic, feasible and acceptable to achieve desired performance. If training is one of the suitable interventions, the step proceeds to verify the potential for improvement in monetary units (e.g. dollars).
1.1 Certain types of training result in tangible, easily verifiable outcomes. Examples are productivity gains, increased sales, decreased errors. These improvements can often be measured in the short term. Pre and post measures of performance with assigned value per unit of gain can establish the potential for improved performance. Example: Reduction of 10% in invoicing errors can result in $800,000 annual savings in investigations, late payments, dispute resolution, customer retention, etc.
Another method for calculating potential for improved performance is by comparing exemplary with typical or average performance. Gilbert (1996) suggests that the potential for improved performance (PIP) is the ratio of the worth of what the exemplary performer produces (Wex) to that of the typical performer (Wt). Hence,
Wex PIP = __ Wt If an exemplary performer makes $1,000,000 in sales and a typical performer $500,000, PIP = 2.0. If total sales is $80,000,000, the potential for improved performance is $160,000,000 or twice current performance.
1.2 Intangible, less readily counted and measured outcomes of training provide a special challenge for calculating potential for improved performance. Examples include listening skills, counseling techniques, systemic thinking. Although less directly measurable, results of improved performance in these areas can produce significant organizational gains. Changes, however, generally take place and can only be measured over the long term.
Schneider et al. (1990, 1992) have proposed a form of survey methodology for estimating the value of improved performance in the so called "softer" areas. It works as follows:
- Define the major job competency requirements of the persons targeted for training.
Effective use of human resources; budget administration; management and evaluation of work; planning; work coordination; problem-solving. The competencies listed should include those which are the subject of the proposed training content.
- Assign a value to each competency expressed as a percentage of time each competency is required in the job.
Effective use of human resources 25%
Budget administration 10%
Management and evaluation of work 10%
Planning 15%
Work coordination 25%
Problem-solving 15%
Total 100%
This can be achieved through survey and estimation
- Within the targeted competency area for training, break this down further into performance requirements. This can be done with supervisors of the target trainee group, the trainee group itself and, if relevant, subordinates and customers.
- Create a 0-9 scale for each performance requirement. Based on supervisor and trainee estimates, assign a desired performance value, 0 being the lowest, 9 the highest. Assign a current performance estimate as well.
Develop and monitor long term work objectives...Desired, 8...Current, 2.
- Calculate the difference between desired and current for each required performance. This represents the estimated gap in performance.
- Total all the desired and current performance requirement estimates as well as the gaps.
Desired total score for all 6 performance requirements 48
Current total score for all 6 performance requirements 24
Performance gap 24
This represents a potential for improved performance of 200%.
- Convert the potential for improved performance to a dollar value.
Average salary per trainee is $60,000/year.
Planning represents 15% of $60,000 or $9,000/year.
300 trainees = $2,700,000/year salary for planning.
Current performance is at 50% of required.
Potential for improved performance = $1,350,000/year.
The seven-step calculation of potential for intangible improvement is based on estimates and on base salaries. As estimates are not perfectly accurate, using a base salary produces a conservative potential for improved performance figure. The survey procedure can be rigorously conducted to increase estimate validity. Trainee fully loaded costs may be used in place of base salary (leading to much higher potential for improved performance figures and closer to actual impact on productivity).
Step 2. Calculate estimated training costs.
Training costs include:
- Training development costs: human resources* ; direct travel and non-travel; media production; consultant fees, licenses.
- Training implementation costs: training facilities and equipment, instructors; trainees; replacement of trainees; lost opportunity; travel; materials; communications; administration.
- Training course maintenance costs. If a course is to be delivered over several years, updating and revisions are usually required at an annual percentage of initial development costs. This can vary from as little as 5% for stable content to 50%+ for highly volatile material.
Step 3. Calculate the worth analysis.
This step essentially verifies the worth of doing training by comparing costs against potential outcomes. Worth analysis is performed as follows:
- Estimate the highest number of annual deficiencies or improvements.
- Estimate the lowest number as well.
Current annual number of deficiencies 20,000
40,000
low high
- Estimate the low and high cost of each deficiency or improvement.
Current annual cost per deficiency $15.00
$25.00
low high
- Calculate current annual cost of deficiencies or improvements both low and high.
20,000 x $15 = $300,000 low cost; 40,000 x $25 = $1,000,000 high cost
- Estimate the range of expected deficiencies corrected or improvements obtained from training.
Low = 20%; High = 40%
- Estimate low and high annual value of training. $300,000 x 20% = $60,000 low annual value $1,000,000 x 40% = $400,000 high annual value
- Multiply the low and high annual values of training by the expected life of the training and divide by the estimated training costs to obtain potential worth of the training effort.
$65,000 $65,000
High worth: $400,000 x 3 years = $1,200,000 = 18.5
$65,000 $65,000 The potential ROI for this training = 2.8 - 18.5:1
This represents a far greater potential return than on most physical capital. It is in line with published results from training when appropriately selected, applied and supported on the job.
Step 4. Train.
This step includes design, development, implementation, support and monitoring/evaluation of training.
Step 5. Calculate the true cost of training.
This step mirrors Step 2, calculate the estimated training costs. However, new actual figures replace previous estimates. As mentioned earlier, the highest costs of training are in the human resources. In Canada, fringe benefits represent an additional 12% - 15% of salary. Fully loaded costs for employees are generally calculated on the basis of salary + benefits + overhead (usually 100% - 150% of salary + benefits) divided by productive annual work hours.
Example:
Salary = $60,000
Benefits = $60,000 x 12% or $7,200
Overhead = $67,200 x 125% or $84,000
Fully loaded cost= $67,000 + $84,000 or $151,000
Cost/day = $151,000 · 230 work days or $656.52/day
Cost/hour = $656.52 · 7.5 or $87.53/hour
The fully loaded cost is approximately 2.5 times the base salary. In the U.S., where benefits run as high as 35% of the base salary, the fully loaded factor for calculating personnel costs is 3.0.
Other expenditures sometimes forgotten in estimates that increase the true cost of training include: equipment maintenance, shipping, handling and storage of materials, correction of errors caught after implementation, trainer training, course publicity, enrollment and tracking. Often, the organization delivers more training sessions than planned due to work schedules that reduce class sizes or unexpected turnover.
Step 6. Calculate organizational return on investment.
This is difficult to carry out in a fully valid manner. The front-end analysis should have led to the selection of a "basket of interventions" of which training is one. If all the interventions have been implemented in an integrated manner, isolating the impact of training alone becomes almost impossible. Ideally, the organization should capture the costs of all interventions and calculate return globally. However, two realistic options are: estimate the percentage of impact likely due from training (establish a range low to high); calculate the value of results against training costs only on projects where training has been the major performance intervention.
6.1 Calculate organizational return on investment for tangible improvement. This usually occurs six months or more after all training has been completed. Robinson and Robinson (1989) suggest, as examples, the following types of performance indicators:
Sales: Size of average sale, sales volume, add-on sales. Non-monetary indicators may include ratio of new accounts to old ones, call-to-close ratio; percentage of objections overcome, items per order.
Supervisory and management: Decreased rejection rates, increased output, reduced absenteeism, reduced tardiness, decreased waste, decreased production costs, reduced cost of new hires, reduced overtime. Non-monetary indicators may include reduced number of grievances, reduced turnover, increased number of employee suggestions adopted, improved climate survey data.
Customer relations: Accuracy of orders and information, size of orders and transactions, adherence to credit procedures, amount of repeat business, number of transactions per day. Non-monetary indicators may include number of complaints, customer satisfaction, number of referrals.
In both front-end analysis and this step, it is always wisest to use indicators with which the organization is already familiar and values. Once improvement data have been gathered, they are compared with the training costs. It is essential to amortize costs properly. Hence, when calculating ROI three months into training, include training development costs proportional to the projected life of the training (estimated in step 3). Similarly, only include implementation costs for those already trained.
6.2. Calculate organizational return on investment for intangible improvement. The procedure for this is almost identical to that of Step 1.2 for calculating improvement in required performance. The difference between pre and post training performance is converted to dollars. This is then compared to appropriately proportional training costs.
Example: Planning (after 6 months of training activities):
- Post training total score for all 6 performance requirements 40
- Pre training total score for all 6 performance requirements 24
- Performance improvement 16
Planning represents 15% of job time.
Average salary per trainee is $60,000.
Planning (15%) represents $9,000.
Pre training performance at 50% of requirement = $4,500.
Post training performance at 83.3% of requirement = $7,500.
Improvement = $3000 per trainee.
ROI = Value
Amortized Cost of Training
= $240, 000 = 2.28
$105,000
The sum of all returns from tangible and intangible training improvements represents the total organizational return on the training investment.
Step 7. Calculate individual increased value of human capital.
Each employee possesses a human capital account, usually established with an initial value equal to his/her salary (although benefits can be included). As the employee´s competencies increase, the value of the account rises. This serves to track the value of the individual human capital asset and is not necessarily given out in salary and bonuses to the employee. However, too great a discrepancy between what is in the account and compensation can result in the employee leaving the organization to obtain more money. On the other hand, not all can be paid to the employee as the company assumes the risk cost associated with the training. The employee´s individual account is increased by the value training has added to his/her current human capital account.
Example: Employee X Human Capital Account
Current value $60,000
Increased post training performance $ 3,000
New value $63,000
In some cases, especially for new hires, the initial account value may be less than the base salary, but is paid out in anticipation of future increased performance capability.
In summary, this section has presented a model for calculating training ROI. As noted at the outset of the section, front-end analysis is required to determine whether or not a need for training exists. Tangible or intangible improvement, this model permits calculation of training ROI. While we may apply the model to intuitively selected training, the first three steps will not be relevant. The results will also remain questionable.
No comments:
Post a Comment