About the Study

Pricing House Cleaning Assignments

HCA Statistical Analysis with Practical Applications

2. Statistical Analysis of the HCA Pricing Formula–About the Study

2.1.	The House Cleaning Alliance (HCA) has developed statistical models to predict the number of man-minutes required for a team of professional house cleaners to clean an occupied single family home on a recurring basis. The pricing models have been derived based on multivariable linear regressions. The factors which were found to be significant in predicting cleaning time can be categorized as: House Attributes Usage Factors Execution Factors
2.2.	House Attributes include physical characteristics which could affect man-minutes, including the size of the residence measured in square feet, any basement area which requires cleaning, number of toilets, and the number of showers in use. These factors are objective and are generally relatively easy to quantify for any house or apartment.
2.3.	*Usage Factors* include those which can significantly affect the time required to clean a home. Some, like number of inhabitants, can be counted, while others are more subjective in nature and relatively more difficult to measure. These include Lifestyle factor, Floor factor, and Pet factor. The precise definitions of these are discussed below.
2.4.	*Execution Factors* include quantifiable attributes associated with a house cleaning company’s unique operating profile, like standard team size, as well as psychological factors which were found to impact cleaning times, like whether or not the workers have more or less work than they would like that day, time of day for the cleaning assignment, and whether, for a specific service date, the client is present during the cleaning process. The analysis has also considered the effect of variability among the efficiency of distinct Team Leaders.

3. Benefits and Limitations of Applying the HCA Pricing Model

3.1.	The HCA pricing model has a few major advantages. Its linear form (in terms of each coefficient) makes it easy to understand, calculate, and implement. Yet, it not only takes into account the physical characteristics of the houses, but also considers other important factors which impact cleaning times. Regression analysis on these variables allows us to study the effect of each factor on man-minutes in both absolute and relative terms. From the business point of view, the results allow a new company to predict more precise man-minutes and prices, and provide a basis for any company to reevaluate alternative operating profiles through a rigorous cost-benefit analysis, as another means of increasing profit margins.
3.2.	In terms of limitations, although the Usage Factors help us predict cleaning times, their subjective nature can limit the model’s accuracy. This limitation can be compounded if any confusion exists about their definitions, or when the data is collected by multiple observers.
3.3.	For the model to be useful, the person performing in-home quotes must have a thorough understanding of how to define and consistently score all variables, especially the Usage Factors.
3.4.	The model can predict with 90% confidence the number of minutes required to clean a home, within a range of 27%. For someone inexperienced in providing in-home quotes, this will undoubtedly be significantly more accurate than he or she could perform without the model. For someone experienced in providing in-home quotes, studying the model can provoke him or her to re-evaluate those factors which he or she already implicitly or explicitly considers in developing a quote. Finally, for anyone performing in-home quotes (or even quotes by phone), the results can serve as a sanity check to confirm the results of his or her own alternative calculation methods.
3.5.	Finally, it is possible that the predictive quality could have been enhanced had more data been available, particularly regarding the psychological factors.