Chapter 6: Forecasting Telecommunications Services


Overview

Forecasting is a necessary part of planning. The Future cannot be predicted with certainty, but the use of statistical data analysis helps prepare for what lies ahead. Since telecommunications is a supporting department in many organizations generally its forecasting depends on the entire organizational planning and forecasting. It is important for telecommunications managers to understand the organization forecast, or if possible, to be a part of the forecasting team for the organization. Understanding the organization plan and the direction it wants to move helps telecommunications mangers plan efficiently. While some equipment is easy to get on a short term notice, many components and systems take time (e.g.: a new PBX, trunks, increased voicemail ports). It is better to predict these increases and arrange to get the equipment quickly when the needed.

Reasons to Forecast

Telecommunications managers spend most of their time solving problems or “fighting fires”. Managers can plan efficiently to project future demand and eliminate many of these daily fire drills by:
·        Making sure that the number of circuits and trunks meets demand; controlling the cost of connect vs. reconnect.
·        Arranging the fastest way to get extra hardware capacity when needed
·        Anticipating staffing demands such as the number of call center agents and PBX administrator that will be needed
·        Predicting physical space, which is often the hardest commodity to obtain
·        Calculating telecom budgets

Steps to Complete the Forecast

Forecasting is very similar to any other business project in an organization. The three fundamental steps to completing a forecast for telecommunications are determining a pattern, fining a source to gather data and using the best method to manipulating the data for the accurate outcomes.

Determining the Pattern

Collecting statistical data about your organization’s telecommunications services is the best way to predict the future needs for the organization. You can often denote patterns in the data from which you can predict the needs in the time to come. The four important patterns that underline forecasts are:
·        Trends (e.g. an organization closed on the weekends might see an influx of calls on Mondays)
·        Seasonality (e.g. weather, Christmas and the New Year)
·        Cyclically (e.g. rise or fall of economy over several years)
·        Randomness (e.g. riot and flood)

Finding a Source of Data

There are many ways within telecommunications systems and services to find statistical reports and data for forecasting. Most equipment (such as the PBX) is computerized and can easily statistics. Sources of telecommunications data includes:
·        ACD (Automatic Call Distributor): Management information systems, which produce reports that range
from quarter hours to weeks

·        Switching System (PBX): Collect usage information on number of hour and call seconds for circuit groups
·        Call Accounting System (part of the PBX)

·        Data can be sorted and combined in variety of ways
·        Collects data a month at a time
·        Data doesn’t show uncompleted calls

Toll Statement:
·        The bill – Paper is hard to deal with and it can be incomplete
·        It does not show the total load on the system
·        It negates uncompleted calls, local calls, incoming calls, or call setup time
·        Routers and Other Network Management Systems: Provides transmission and bandwidth statistics
·        Budgeting Information: Current expenses for various types of telecommunications service may be useful to predict future demand

Using the Most Feasible Method

There are several methods to forecasting using the data collected from one of the sources described above. The method used depends on the type of information that you have and the type of information that you are searching for. The method also depends on the degree of accuracy that is required. Much of the data can be collected in a spreadsheet, such as Excel, and can be analyzed by one of the following forecasting methods:

·        Time Series: Fits a trend to plotted data to show upward or downward movement. The disadvantage is that it predicts the future on history.

·        Moving Averages: Smoothing out the variations by averaging the data before and after a given point. The disadvantage of this method is that it gives the same value to all observations. In an effective analysis recent observations should have more value than the older ones.

·        Exponential Smoothing: Much like moving averages, but creating an exponential equation to find the average from less evenly weighted data

·        Regression Analysis:

·        Uses one truth to predict other variables
·        Fact or forecast A is used to predict B
·        The correlation coefficient is the denotation of how closely data is related (0.0 – 1.0)
·        Another part of the organization must prepare a forecast for telecommunications manager to use that forecast for telecommunications needs

·        Judgmental Forecasting: Changing the forecast as you go along to keep from getting too far off track

·        The Delphi Method: This forecasting method is better than simple guess but it is very risky. Data is gathered through following steps:
1.     Asking the opinion of the various department leaders about the direction that the organization going
2.     Weighing their answers considering knowledge, optimism or pessimism
3.     Averaging those weighted answers