SALES FORECASTING
WHAT IS A SALES FORECAST?
A sales forecast
predicts the value of sales over a period of time. It becomes the basis of
marketing mix and sales planning.
A
short-term sales forecast (say for a period of one year) when linked
to the sales budget helps in the preparation of an overall budget for the firm
as a whole. The short-term sales forecast in effect also provides the essential
financial dimension to sales in terms of expected sales revenue and expenses
required. Also, it helps in assessing the cash inflow and outflow needs and
their sources.
A long-term sales forecast (say
for a period of 5 years or so) on the other hand, focuses on capital
budgeting needs and process of the firm. It provides for changing the marketing
strategy of the firm, if needed, and includes reference to emerging product
market needs, new market segments to be catered, review of distribution network
and promotional programmes, organisation of sales force, and marketing set up.
The long-term sales forecast triggers the task of aligning the production,
procurement, financial and other functional needs of the firm with the
finalised sales forecast.
HOW TO PREPARE A SALES FORECAST?
The preparation of a
sales forecast requires (a) the availability of historical information on the
product and industry sales, (b) identification of Product Sales Determinants, (c) prediction regarding the behaviour of market forces for the
period under forecast, (d) use of appropriate techniques for forecasting, (e)
judgement of executives preparing the sales forecast, and (f) the firm's market
share objectives. These sales forecasting requirements are discussed below.
Information Needs for a Sales Forecast
Use of reliable,
up-to-date and relevant information is the most critical aspect of sales
forecasting. The information required for a sales forecast should cover:
1. An assessment of the total market size
2. An appreciation of the market trends
3. Innovations which may have an impact on the market
4. Market trends in foreign countries where the market pattern
is in advance of the domestic market
5. An evaluation of the market share obtained
6. An evaluation of competitive strengths
7. The criteria on which purchase decisions are likely to be
made
8. Assessment of elements at work in the market which will
influence sales
9. The influence in the market of competitors
10. The level of sales needed by the company to obtain optimum
use of resources
11. The image of the Company in the market
12. The marketing strategy of the company to capitalise on its
strengths and overcome its weaknesses
13. An evaluation of the market share which can be obtained
14. Assessment of factors within the company which will
influence sales levels
15. Planned distribution and sales promotion activities by the
company
APPROACHES TO SALES FORECASTING
There are two general
approaches to sales forecasting at the level of the firm-the breakdown
approach (also called top-down approach), and the market build-up
approach.
Breakdown Approach
Under this approach,
the head of the marketing function initially develops a general economic and
market sales potential for a specific period. The firm's sales
potential is then derived from it. The example of a colour television receiver
company developing its sales forecast relates to the use of the breakdown
approach.
Market Build-up Approach
In this approach the
task of sales forecasting begins by first estimating the sales at the product,
product lines, customer groups or geographical area level. The estimates of the
different product, product lines, customer groups or geographical areas are
then aggregated and reviewed in the light of the firm's objectives, available
resources, as well as competitors activities before the sales forecast is finalised.
For example a leading automobile engine manufacturing company determines the
sales forecast of its diesel engines by using both a breakdown and a market
build up approach. In the first instance, it ties an econometric model and an
estimate of the company's market share to derive the company forecast. Under
the second approach the company initiates the process by undertaking a detailed-study
of the needs of each of its diesel vehicle customers. This study includes an
analysis of market factors such as the vehicle manufacturer's present engine
inventory and back orders as well as the vehicle manufacturer's marketing
programme. The resulting forecast is prepared by vehicle manufacturer, model,
and month-wise. These individual manufacturer forecasts are aggregated to
produce a company sales forecast which is then compared with the company
forecast arrived at under the first approach, and finalised.
While both the
approaches have their own usefulness, the breakdown approach is less time
consuming and costly when it can use aggregate data made available by others.
It may, however, lack the advantages of greater realism and reliability which
result from the use of market build-up approach. Combination of both the
approaches though time consuming seems ideal and worth the effort expended.
METHODS OF FORECASTING
Let us now consider
various methods used for preparing the sales forecast. These methods are
commonly grouped into 5 categories: executive judgement, surveys, time series
analysis, 'correlation anti regression methods and market tests.
Executive Judgement
It is an efficient
method of sales forecasting. Based on the past performance, insights gained and
intuition of the executive(s), this method of sales forecasting works out
fairly well particularly when the market is stable. However, this method
generally suffers from difficulty in realistically reflecting changes in the
market. Sales force composite method and jury of executive opinion are the two
popular forms of this method of sales forecasting.
Surveys
A second way of sales
forecasting is by surveying the customers, salesforce, experts, etc. and
ascertaining their predictions. Customer
surveys can provide information relating to type
and quantity of products which customers intend purchasing. Salesforce surveys can provide estimates
of overall territory off-take, company's share and the share of the major
competitors. Dealers survey may also form part of the salesforce survey
if a firm so desires. Expert surveys provide sales forecast as the experts and industry consultants
look at it. They bring in an outsider's view to the company's internal forecast
and help many a times by adding new dimensions for consideration of management.
Time Series Analysis
Using the historical
sales data, this method tries to discover a pattern or patterns in the firm's
sales volume over time. The identification of the patterns helps in sales
forecasting.
Time series analysis
helps locate the trend, seasonal, cyclical and random factor changes associated
with the past sales data. In this way, it improves the prediction from the past
sales data. Experience reveals that time series analysis for sales forecasting
are quite accurate for short and medium term forecasts and more so when demand
is stable or follows the past behaviour.
Some of the popular
techniques of time series analysis are: moving averages, exponential smoothing,
time series extrapolation, and Box-Jenkins technique.
Correlation and Regression Methods
These methods attempt
at examining the relationship between past sales and one or more variables such
as population, per capita income or gross national product, etc. The use of
regression analysis is done in order to determine whether any relationship
exists between the past sales, and changes in one or more economic, competitive
or internal variables to a firm. The accuracy of forecasts made by using
correlation and regression methods is generally better than the other methods.
Typical forecasting applications of these methods are sales forecasts by
product class. Though the correlation method helps in identifying the
association between the factors, it does not explain any cause and effect
relationship between them.
Some more advanced
forecasting methods explaining cause-effect relationship besides regression
method include econometric model, input-output model and life-cycle analysis
method. The life-cycle analysis method is used for forecasting of new product's
growth rate based on s-curves. The phases of product acceptance by the
various groups of customers such as innovators, early adopters, early majority,
late majority, and laggards are central to the analysis.
Market Tests
Market tests are
basically used for developing one time forecasts particularly relating to new
products. A market test provides data about consumers' actual purchases and
responsiveness to the various elements of the marketing mix. On the basis of
the response received to a sample market test and providing for the factor of a
typical market characteristic as well as learning from the market test, product
sales forecast is prepared.
Substantial
fluctuation that one finds in reality from market to market limit the accuracy
of sales forecasts made by this method, unless the market test is designed
systematically.
Combining Forecasts and Using Judgment
Experience brings out
that the forecasts resulting from the use of multiple methods in a combined way
greatly surpass most individual methods of sales forecasts. Research also
supports the combined use of quantitative and qualitative methods of sales
forecasting in a given situation rather than using either of the two.
Application of judgment to quantitatively arrived forecasts should be done in a
structured manner with a view to adding insights and realism to the forecasts
so arrived at, since a forecast is a prediction and needs the subjective
perception too.
Several studies have
shown how combining forecasts by using one or the other methods can improve
accuracy of the forecasts. The methods which can be used for combining
forecasts are: (i) a simple average of two or more forecasts, and (ii) by
assigning historical or subjective weights to such forecasts which more closely
reflect the changing reality. In short, being aware of the conditions under
which some forecasting methods work better than, others enables the firm to
prepare for different alternative forecasts. By monitoring which alternative
works better, the firm can learn to achieve its goals more effectively.
COMPUTERISED
SALES FORECASTING
The rapid developments
in computer hardware and software has made it possible for managers to make
sophisticated forecasts with the help of computers. The greatest advantage of
this is that managers can introduce subjective inputs into the forecast and
immediately test their effects.
Specifically, the last
few years have seen sophisticated forecasting models being rewritten using
Spread Sheet software programmes for personal computers. Lotus 1-2-3 and
Microcast programmes are now available at reasonably affordable prices.
Developments in the computer field especially in computer artificial
intelligence systems have also enabled the development of expert systems
models i.e., the model that the experts use in making a decision. These are
of great use when judgement is an important part of the forecast. In future, we
are going to see greater use of computers in sales forecasting.
0 comments:
Post a Comment