What’s this study about?
This is a guideline to help you come up with a formula to estimate the compensation for your new hires or to analyze the compensation of your current employees relative to each other. Especially, in relation to the sales team that compensation is set around the budgeting season and constantly changing depending on the assigned budget.
Where to apply?
This model is originally built to analyze and set the senior sales managers compensation and to compare their compensations against each other using four different criteria. The sales managers’ compensation is changing each year according to the amount of budget they’re responsible for and their previous year performance.
This model can be expanded to all the other employees and teams by changing the variables and criteria.
How does it work?
Coming up with a right compensation plan to offer to a potential candidate has a lot of HR elements to it. The HR team needs to make sure that the offered plan is fair; relative to current employees with the same positions and experience. While company’s HR and financial policies must be taken into account, the compensation should also be competitive within the same industry and geographic location. The economic growth, the size of the labor marker or talent pool (number of applicants for the position can be a good indicator of this element), the federal and state rules and regulations, etc. are the regular players in this process too.
This model works with the historical data to introduce a formula to help with estimating a compensation plan for the new hires and to calculate the normalized compensation for current senior managers.
In this model, the compensation plan is assumed to be a combination of a base salary and commissions paid off the initially planned budgeted revenue.
Choose your criteria
To find the formula, the first step is deciding on the criteria that is important to determining the right compensation. For example, budgeted revenue is definitely the most important factor to calculate a sales employee’s compensation plan. Other factors can be the company’s budget, the ranking of the open position within the organizational structure, experience, education level, number of employee the new hire will be managing, cost of living in proximity of the office location and other qualifications.
This model is using the historical data gathered from the current employees for the following criteria:
Geo Economic Zone Percentage: This factor is in relation to the cost of living in the city that the office is located. This factor comes into play when the company has offices in multiple locations in different states or countries. Obviously, the living expenses are not the same in every location. For example, the living cost is much higher in New York compared to Albuquerque. For the purpose of this model, the cheapest location receives a 100% and other markets get a lower percentage relative to this location. In our example, the understudy company has only two locations: the cheaper one has received 100% and the more expensive location is assigned a 90%. This means 90% of the compensation in a more expensive place should be as much as 100% of the compensation in the cheaper location.
Now that the criteria are determined, it’s time to gather the historical data for the current employees for the determined criteria and create the following table:
The first two columns are the employees ID and job title.
The third column is the total of their current compensation plan.
And the last four columns are the data gathered for the determined criteria explained above.
At this point, we should study to see if there’s any relationship between the current total compensation and the four variables we gathered data for. In fact, we want to find out if there’s any formula that would help us to get to an estimate of the total compensation using these four variables. In other words, we are trying to see if our dependent variable (F(x) = y) is a function of any of our four variables (Xi, i = 1, 2, 3, 4).
Regression analysis is the tool to be used for this purpose. Simply, the current compensation column is selected as the dependent variable (y) and the other four criteria columns as independent variables (Xi).
The result of the regression for our gathered data is this example after running the regression analysis is the following tables:
The R Square (yellow highlighted row), simply, is an indicator that shows how strong the relationship is between dependent and independent variables. The maximum, possible R Square is 1. So, a 0.96 R Square is a demonstration of high co-relation between current compensation and the four selected variables.
If the R Square is not significant, we probably need to re-do the study by choosing other relevant criteria to finally reach a significant R Square in order to increase the credibility and relevancy of our model.
The compensation formula
Using the results from the regression analysis, the desired formula to calculate the total compensation can be developed as a function of the four variables:
F(X) = a0 + a1X1 + a2X2 + a3X3 + a4X4
Total Compensation = a0 + a1 (Budgeted Revenue) + a2 (# of Years) + a3 (Ranking) + a4 (Geo Zone %)
The ai (i = 0, 1, 2, 3 and 4) are the coefficients highlighted purple in the table above. Therefore, total compensation formula can be developed by entering the coefficient values into the above function:
Total Compensation = 127,984 + 0.003565 (Budgeted Revenue) + 1,032 (# of Years) + 11,473 (Ranking) – 26,849 (Geo Zone %)
The resulted formula can be used as an “estimate” to find the offered compensation for the new hires. It’s an estimate because it’s purely calculated based on the relationship between historical data gathered for four variables and current compensation using a statistical method with a lot of room for errors. In reality, there are other factors to be considered that will impact the final decision. But in general, this formula can be used to develop a range of compensation to use for negotiation and as a base to come up with a targeted range.
This formula can also be used to find out how the current employees standing relative to each other by comparing their current compensation with what they should actually get paid or their normalized compensation.
Total compensation formula is used to calculate the normalized compensation or recalculate employees’ compensation. After entering the data gathered for each employee into the formula, the normalized compensation column is populated as it’s illustrated below:
There are two different compensation figures in the above table: Current Total Compensation (what employees are getting paid) and Normalized Compensation (what they should be getting paid according to the result of the regression analysis). To find out if an employee is overpaid or underpaid and to see where everyone is standing relative to others, we can draw following graphs using these two columns:
The red line is the current compensation compared against the blue line, the normalized compensation. For any employees if the red line is above the blue line, they are overpaid and if the red line is below the blue line they are underpaid. The ideal case is where both lines overlap.
To plan the next year compensation and adjust everyone’s relative to each other, the compensation formula and this analysis can be used as one of the many tools.
Again, this model is not perfect, as the analysis is purely based on a statistical analysis and does not take any other factors into account.
This is not an academic paper or a case study. It’s just an illustration of an actual model used to analyze the performance of a sales team with respect to customer retention. Unlike an academic paper, there’s no history, background, literature review, or bibliography to be added here. This study is just an observation and explanation of a model used in an actual job assignment by a financial analyst.
For confidentiality reasons, none of the numbers presented in this illustration are actual or from any existing, operating enterprises. They are solely examples to help with understanding the concept.