Analyzing and Forecasting Church Revenue: A Case Study Project

This project analyzes the revenue forecasting of St. Elizabeth Seton Catholic Church, aiming to predict future revenue trends. The analysis employs various forecasting methods, including moving average, exponential smoothing, and seasonally adjusted methods, to determine the most suitable approach. The project evaluates the performance of each method using Mean Absolute Percentage Error (MAPE) and compares their ability to capture seasonal patterns. While the seasonal method accurately reflects monthly variations, its dependency on current revenue data limits its predictive capabilities. Consequently, the project recommends using exponential smoothing for predicting future revenue, although it acknowledges the limitations in accurately capturing monthly patterns and distant future values. The study provides insights into the application of forecasting techniques in financial decision-making for non-profit organizations.

Service Analytics

Introduction
In this particular research the case study of St. Elizabeth Seton catholic church is
analysed where the problem is to predict the revenue of church in the future months.
The vital decisions of the church is dependent on its revenue and hence to take certain
financial decision the pattern of the revenue generation must be known. Hence,
different forecasting methods are applied with the provided data of case study and
their relevancy to prediction are evaluated.

Case background
St. Elizabeth Seton catholic church was found in 1976 at Daphne in the state Alabama. As the church
grows in popularity the expenses and cash inflow to the church are also dynamically changes with
time. Thus the church needs to take some difficult financial decisions to stabilize the net cash outflows
which was planned by the church committee. Thus the church needs to forecast the only income
source which is offertory revenue in each month to make financially viable decisions. This
responsibility is given to Megan an important member of church who needs to predict the offertory
revenue in coming months as accurate as possible.

Objective of research
The primary objective is to predict the future trend of the revenue
generation in months after September 2005 for the church. Thus to
predict the monthly revenues different forecasting techniques are
applied such as moving average method, exponential smoothing and
other advanced forecasting methods.

Descriptive statistics
Descriptive statistics first 12 months
Mean 103752.2725
Standard Error 8329.036045
Median 93535.13
Mode #N/A
Standard Deviation 28852.62722
Sample Variance 832474097.3
Kurtosis 4.802159965
Skewness 2.038509209
Range 105002.33
Minimum 76936.52
Maximum 181938.85
Sum 1245027.27
Count 12

Descriptive statistics
Descriptive statistics 2nd 12 months
Mean 97240.82
Standard Error 6654.166
Median 94384.46
Mode #N/A
Standard Deviation 23050.71
Sample Variance 5.31E+08
Kurtosis 4.943079
Skewness 1.980098
Range 82967.09
Minimum 77038.89
Maximum 160006
Sum 1166890
Count 12

Descriptive statistics
Descriptive statistics 4th 12 months
Mean 100948.5
Standard Error 6280.455
Median 98303.69
Mode #N/A
Standard Deviation 21756.13
Sample Variance 4.73E+08
Kurtosis 2.908834
Skewness 1.505644
Range 76811.04
Minimum 79139.66
Maximum 155950.7
Sum 1211382
Count 12

Descriptive statistics
Descriptive statistics 4th period of 12 months
Mean 104833.4
Standard Error 6889.265
Median 100494.6
Mode #N/A
Standard Deviation 23865.11
Sample Variance 5.7E+08
Kurtosis 0.837461
Skewness 1.086748
Range 77645.6
Minimum 81039.41
Maximum 158685
Sum 1258001
Count 12

Line chart of Actual Offertory
revenue
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
0
20000
40000
60000
80000
100000
120000
140000
160000
180000
200000
Line chart of Actual offertory income
OFFERTORY
Period (in months)
OFFERTORY

Forecasting Part 1:
In the part 1 of forecasting the two forecasting methods are applied which are moving average and
simple exponential smoothing method.
The moving average forecasting formula is given by,
N period Moving Average Forecast (for next period) = (Sum of Actuals for N number of previous
periods)(León-Castro et al. 2018).
Here, N = 12 months is assumed

Forecasting Part 1:
The second model which is applied to forecast is the exponential
smoothing model which is given by,
Simple Exponential Smoothing Forecast (for next period) = Previous
Period Forecast + (β * (Previous Period Actual - Previous Period
Forecast))(Karmaker 2017).
Here , β =damping factor = 0.05 is assumed.
Mean Absolute %Error = Average of: ((|Actual - Forecast|) / Actual ))
for all periods where Actuals and Forecasts exist.

Forecasting Part 1:
The results are compared by line charts with the actual data and by a
statistical measure known as percentage mean absolute error.
Mean Absolute %Error = Average of: ((|Actual - Forecast|) / Actual ))
for all periods where Actuals and Forecasts exist.
Mean Absolute %Error of moving average method = 17.18%
Mean Absolute %Error of exponential smoothing method = 16.71%