Student name: Aldair Rivero Córdoba

Control number: E15021499

Subject: Operations Management 1

Topic: Demand forecast

Professor: Zamudio Poistan Angel

Date: 12/April/2018

1
 Index
Content
Introduction.......................................................................................................................................3
2.1 Strategic Importance of Forecasting............................................................................................4
2.2 Demand Characteristics.............................................................................................................12
2.3 Qualitative Methods...................................................................................................................15
APPLICATIONS.........................................................................................................................16
2.3.1 Consultation with the sales force............................................................................................17
Sales Force Consultation: Its Advantages and Disadvantages..........................................18
2.3.2 Executive opinion jury.............................................................................................................19
Executive opinion on forecasts................................................................................................19
2.3.3 Delphi method.........................................................................................................................21
2.3.4 market research......................................................................................................................24
Objectives of market research.................................................................................................25
Benefits of good market research:...........................................................................................26
Steps to conduct market research..........................................................................................27
2.3.5 Life cycle analogy....................................................................................................................28
2.4 quantitative methods.................................................................................................................28
CHARACTERISTICS OF THE QUANTITATIVE METHOD................................................30
2.4.1 Time series..............................................................................................................................31
Components of a time series.......................................................................................................31
2.4.1.1 Simple approach...................................................................................................................32
2.4.1.2 Moving averages..................................................................................................................33
2.4.1.3 Exponential smoothing.........................................................................................................34
2.4.2 Causal Relationships................................................................................................................35
2.4.2.1 Simple regression.................................................................................................................36
2.4.2.2 Multiple regression..............................................................................................................37
2.5 Forecast in the service sector.....................................................................................................39
2.6 Forecast for start-up companies................................................................................................41
2.7 Use of software in forecasts.......................................................................................................42
Conclusion........................................................................................................................................44
Literature.........................................................................................................................................44

2
 Introduction
Good planning is based on data, that is, forecasts, which can be sales, prices,
consumption or any other variable of interest. This blog explores different
techniques that allow forecasting and recognizing and projecting trends and
seasonality, or relating the variable to another to make the projection.

One of the first steps in any planning and budgeting process is to develop a good
forecast, primarily of demand and sales. The planning of production, purchasing,
labor requirements, financial resources, etc. depend on this forecast, which is why,
without a doubt, it becomes very valuable and decisive information for achieving
the goals of any organization.

Preparing a forecast can be simple or complex, depending on different factors.

Demand (or what is to be forecast) is often affected by cyclical and seasonal
factors, and other times it is influenced by elements that are very difficult to
predict. In any case, experience and market knowledge are vital, but sometimes
this is not enough to make a good forecast.

For the sales department of any company, it is essential to know these methods: to
be able to select an appropriate forecasting method, use it and draw conclusions
from it.

3
 2.1 Strategic Importance of Forecasting

It is the application of statistical techniques to estimate what will happen with the
company's existing sales within a given time, which may be of products or services
. Why are forecasts needed in the company:
The basic objective of a forecast is to reduce the range of uncertainty within which
decisions are made that affect the future of the business and with it all the parties
involved.

Making forecasts in the different production systems that exist in the world today
is essential, since by having accurate estimates, adequate planning of production,
human resources, company finances, deliveries, purchases, etc. can be carried out.
Forecasts are used in the process of establishing both long- and short-term
objectives, thus constituting the basis for the development of plans, at a general
level and in the different areas or units. Plans based on these forecasts will not only
address them but will also establish strategies and actions that can counteract,
correct or promote them.

4
 Data manipulation and cleansing. It is possible to have too much or too little data,
irrelevant data, outdated data, etc., all of which will require some processing to obtain the
necessary and appropriate data.
Construction and evaluation of the model. It involves using the data in a forecasting model
that is adequate in terms of minimizing the forecast error.
Application of the model (the actual forecast). It consists of the actual model forecasts
that are generated once the data have been collected and perhaps reduced to only the
appropriate data, once a suitable forecasting model has been chosen.
Prognosis evaluation. It involves comparing forecast values with actual historical values.
Examination of error patterns often leads the analyst to modify the forecasting procedure.

Classification of forecasting models:

Short term:
They are used to design immediate strategies and are employed by middle management
and first-line management.

Medium term:
Short and long term, useful for decisions at all levels.

Long-term forecasts:
Required to set the general direction of the organization, they are generally made for
senior management to use in strategic planning processes.
Depending on their attention to detail, they are classified as:

5
 Micro forecasts:
They involve small details and interest the middle and front-line levels.

Macro forecasts:
They are carried out on a large scale and are of interest to senior management.

According to the intensity of data use, they are classified as: Qualitative

forecasts:
They are based on the judgment of individuals or groups of individuals, they can be
presented in numerical form but are generally not based on historical data series.

Quantitative forecasts:
They use significant amounts of prior data as a basis for prediction.
They can be:
Simple (non-formal): They project past data into the future without explaining future
trends.
Causal (explanatory): they attempt to explain the functional relationships between the
variable to be estimated (dependent variable) and the variable or variables that explain
the changes (independent variables).

6
 Quantitative
forecasting
techniques-Time series forecasting-Exponential
smoothing method-Holt method

Areas where forecasts are usedUses

of forecasts in the companyAll
organizations that plan the conditions of their future, which they do not know for certain,
use forecasts in their different functional areas. Some use cases for forecasting in the
enterprise are:
In the marketing area, we predict how the market will grow, what our own and our
competitors' market share will be, what the price trend will be, what new products will
shake up the market...

In the production area, forecasts are made regarding the cost and availability of raw
materials, the cost and availability of labor, when maintenance will be required for
equipment, and what plant capacity will be necessary to meet demand.

In the financial area, forecasts are made of the reference interest rate for loans, the level
of uncollectible accounts, and how much capital will be required to expand own capacity.

7
In human resources, forecasts are required on the number of employees, staff turnover,
absenteeism trends, training
needs. At the strategic level, forecasts are made on economic factors, price changes,
costs, growth of product lines:
-SalesIn
order to know the income that will be obtained from sales in a certain period of time.

-OperationsIn
the operating budget to know how much will be spent on administrative, sales and
financing aspects.

-MarketingIn
market studies for the incorporation of new products or services.

-ManagementIn
reference to the company's growth in relation to the sector

8
 We extend historical values into the future, where measurements are not yet
available. Forecasting is typically done to optimize areas such as inventory levels,
production capacity, or staffing levels.
There are two main structural variables that define a time series forecast:
The exponential smoothing method can be considered as an evolution of the
weighted moving average method. In this case, the average of a time series is
calculated with a self-correction mechanism that seeks to adjust the forecasts in
the opposite direction to past deviations through a correction that is affected by a
smoothing coefficient.
The period, which represents the level of aggregation. The most common periods
are months, weeks and days in the supply chain (for inventory optimization).

The horizon, which represents the number of periods in advance that must be forecast.
Example of application of a Simple Exponential Smoothing forecastExponential estimation
model that directly smooths the trend by obtaining the difference between successive
values (of exponential smoothing), to forecast into the future for n periods. It allows to
reduce the effect of randomness (using the difference between the averages calculated in

9
 two successive periods). The demand estimate or trend to be forecast is updated. A
prognosis with a delayed response to growth is avoided.

In January a car dealer estimated sales of 142 cars for the following month. In February,
actual sales were 153 cars. Using an exponential smoothing constant of 0.20, budget sales
for the month of March.
With this idea, exponential smoothing can be used to update the trend estimate, leading
to double exponential smoothing, represented by the following set of equations:

ST = α dT + (1- α) (ST-1 + BT-1)

BT = β (ST - ST-1) + (1- β) BT-1

10
 FT+K = ST + k

BT Where:
ST = ordered estimate of the period BT
= slope estimate K
= number of periods FT
=

forecast Application Holt

method Time series forecasts Smoothing method
.

2.2 Demand Characteristics

For many business planners and leaders, reading market demand is the aspect that most
concerns them given the importance of the topic in terms of sales and financial continuity
of the business. We all want to satisfy customer needs, we all want to have happy
customers and we all want the business to prosper commercially and financially.
The market is subject to various factors of different kinds, some of which are controllable
by the company and others not so much. Some are easily predictable and others are
impossible to predict, such as the case of the influenza pandemic that we suffered last
year in Mexico.

Demand behavior: In another of my articles I described the importance of recording data

historically in order to know what is happening with market demand and determine if the
peaks and valleys it presents have an explanation, be it sales promotions, price increases,
market shortages by competitors, etc. Assuming that there is good information and
consistency in the demand data, the next step is its analysis.

11
 The following aspects of the claim must be assessed here:
Seasonality: If I sell beach items, it is expected that the demand for my products will
increase in some months due to the events that happen in those months of the year, and
it is expected that over time, this behavior will repeat itself continuously. Of vital
importance at this point is the calculation of the seasonal index, to help establish a
projection of demand in a particular period of the year. For example, the demand for
swimsuits could be 100 units per month, but in the month of July the average rises to 175
and in the month of September it drops to 35, the seasonal index for the month of July
will be 1.75 and for September 0.35, meaning that in the month of highest sales, sales
increase by 75% above the average and in the lowest month they will be 65% below the
average. If you analyze this information, it is gold dust for your forecasts, since you will
know when to increase your inventories and when to decrease them and gain financial
points in your organization.

The trend: if sales of my product are increasing, when graphing the data, this has to be
clearly shown. There will be months in which demand does not show increasing behavior,
but in the general picture the trend may be bullish, or on the contrary, it may be stable or
in decline. A graphic analysis of this can give us an excellent indicator of what is happening
in the behavior. Like seasonality, estimates and calculations of demand trends are very
useful. There are many and varied tools, including simple and double moving averages,
exponential smoothing, simple and double, and linear regression among the most
popular. The challenge here is to define which one to use when trying to predict demand
behavior. What will help define the model to be used is the combination of trend analysis

12
 with the maturity graph of a product in the market. If these products have been on the
market for some time, for example, Coca Cola, a stable behavior in demand with an
upward trend is to be expected, so linear regression will be the most advisable, but in the
case of a fashionable toy, which has just been introduced to the market, a simple moving
average may be the best option, it depends largely on the data and information available.

Random variation: We talked a moment ago about new products on the market or
fashions that are presented in the use of a product, either because the fashionable actor
or actress uses it and as opinion leaders they influence the market and cause atypical
behavior in it, in cases like these there will be ups and downs in its sale, without it being
able to predict exactly, here it is convenient to monitor it and prepare with inventory to
cushion the effect of demand. In demand analysis, it is advisable to separate these items
and treat them differently and not mix them with the rest of the products.

13
 Cyclical variation: this occurs over long periods of time, for example, during municipal,
presidential or any other type of election, there will be products (billboards, posters, etc.)
that present atypical, perhaps incremental, behavior, so, again, the timely recording of
this type of events is of utmost importance, in order not to play guessing games when
predicting demand.

2.3 Qualitative Methods

The qualitative research method is the collection of information based
on the observation of natural behaviors, speeches, open responses for
the subsequent interpretation of meanings.

While quantitative methods provide numerical values from surveys,

experiments, interviews with specific answers to carry out statistical
studies and see how their variables behave. Widely used in sampling.

However, the concept of qualitative method analyses the whole

discourse between subjects and the relationship of meaning for them,
according to cultural, ideological and sociological contexts. If there is a

14
 selection made based on some parameter, it will no longer be
considered qualitative.

Let's say that the qualitative research method does not discover, but
rather constructs knowledge, thanks to the behavior between the people
involved and all their observable conduct.

APPLICATIONS
Among the most popular qualitative research techniques and types of
methodology, we find communication between individuals, as the basis
of all of them.

The types of qualitative research methodology are mainly three:

Participatory observation: the researcher participates in the problem or

situation to be analyzed. He experiences the experiences first-hand and
that is an advantage when it comes to understanding the subjects of the
research.
Non-participatory observation: the researcher does not participate in the
problem or situation. Two examples of this type of observation are:
simulations and case studies. In the first ones, a situation is created and
the participants act. They are observed. And the second practice, carries
out an exhaustive study of a person or company, institution, etc.

Ethnographic research: combines the two previous types of observation.

It is used to extract the maximum amount of data, by applying both
participatory techniques and types of observation in which the
researcher is not involved.

While there may be several techniques for analyzing qualitative

information. We highlight the most common ones:

15
 — Open-ended interviews.

— Group techniques:

Discussion groups are a highlight. Information with different points of

view will be most valued. But there are also techniques to encourage
creativity, such as brainstorming.

— Nominal Group Technique:

One of the most democratic. It makes it possible to reach a rapid

consensus on issues, problems, solutions or projects, avoiding the terms
'losers' and 'winners' among group members.

— Focus Group Technique:

Form of group interview that uses communication between researcher

and participant.

— Delphi technique:

Information on predictions is extracted and based on a panel of experts.

2.3.1 Consultation with the sales force

It is a qualitative technique to forecast demand by consulting each seller for an estimate
of future demand, which must be reviewed, because as you will see in the disadvantages,
this technique usually produces very optimistic or perhaps very conservative estimates.

The adjusted data is combined with data from sellers in other regions, allowing you to
generate forecasts by district, city, country, etc.

16
Sales Force Consultation: Its Advantages and
Disadvantages

The advantages of this technique are:

 It is an inclusive method when requesting estimates from sellers, which can

increase their motivation. Although this is maximized when your opinion is
frequently requested and your goals are not out of reach.
 It allows you to adjust the sales force's goals, making comparisons between each
salesperson.
 The forecast obtained in one area can be combined with that of other areas to
obtain forecasts for larger territories.

And on the downsides:

 Since the company sets goals for the sales force, this technique can generate
unreliable data if the salesperson provides a demand estimate that is lower than
necessary to meet his or her goal.

 Sometimes the estimate given by the sellers can be very high or very discreet. It is
therefore necessary that the estimate be reviewed and adjusted to generate
reliable data or that it be used in conjunction with other techniques such as Delphi
or executive judgment.

17
 2.3.2 Executive opinion jury
Executive opinion polling is a technique that draws on the experience and technical
knowledge of senior management in a company to reach a consensus for forecasting
demand. It is one of the most commonly used when it is necessary to act quickly in the
face of unforeseen events or the launch of new products on the market.

Executive opinion on forecasts

Looking at its advantages, we have to:

 It allows you to make decisions quickly. Useful for sudden changes in demand for a
product, an emergency situation in the company, reactions to the market, etc.

 It is not expensive: It is not an absolute advantage, but it is generally the case.

Well, it depends on the company. Other sources mention that it is an expensive
technique because it takes up executives' time. I am inclined to think (based on my
experience) that it is more expensive to develop a focus group, the Dolphin
method or to do market research. So advantage: It is not expensive.

And among the disadvantages, we can find the following:

 Unfamiliar with field work: This can be a disadvantage, not always, but it can
be. They are senior managers in the company, so… making a decision based
on their knowledge of the business is important, but considering what is
happening in the mission is also important. If you try to calculate the sales
forecast by asking the salesperson, he or she is more likely to give you a
conservative number compared to the marketing manager, who expects
higher figures.

 It can lead to ambiguous conclusions: This can happen when different

knowledge and experiences are mixed at the same table. As? You have
different people in senior management, each with similar but different
knowledge. It is important to choose the objective of the session with the

18
 jury carefully, since each one can begin to offer his or her opinion from his
or her own perspective and experience, which can clash with that of the
other executive.

Once the forecast has been obtained, it is always useful to adjust what was
obtained with a qualitative method with a quantitative one, or vice versa.
This way we can ground the data when we are not sure of the consensus
reached by the company's senior management.

2.3.3 Delphi method

The Delphi method is included within the prospective methods, which study the future, in
relation to the evolution of the factors of the techno-socioeconomic environment and
their interactions.
The first Delphi study was conducted in 1950 by the Rand Corporation for the United
States Air Force and was named Project Delphi. Its objective was the application of expert
opinion to the selection of an optimum American industrial system and the estimation of
the number of bombs required to reduce munitions production to a certain amount.
It is a method of structuring a group communication process that is effective in allowing a
group of individuals, as a whole, to deal with a complex problem. (Linstone and Turoff,
1975)
The Delphi's predictive capacity is based on the systematic use of an intuitive judgment
issued by a group of experts.

19
 *CHARACTERISTICS:
.– Anonymity: During the Delphi no expert knows the identity of the others who
make up the debate group.
– Iteration and controlled feedback: Iteration is achieved by presenting the same
questionnaire several times, which allows the interquartile range to be reduced,
since the experts become familiar with the different points and can modify their
opinion.
– Group response in statistical form: The information presented to the experts is
not only the point of view of the majority but all opinions are presented indicating
the degree of agreement that has been obtained.
– Heterogeneity: Experts from certain branches of activity may participate on the
same basis.

*THE METHOD CONSISTS OF 4 PHASES:

1st) Definition of objectives: In this first phase, the formulation of the problem and a
general objective are proposed, which would be composed of the objective of the study,
the spatial frame of reference and the time horizon for the study.

2nd) Selection of experts: This phase has two dimensions:

– Qualitative dimension: They are selected based on the pre-established objective and
taking into account criteria of experience, position, responsibility, access to information
and availability.
– Quantitative Dimension: Choice of sample size based on average resources and time
available.

20
 Panel formation. The recruitment phase begins, which will lead to the configuration of a
stable panel. When contacting experts, it is advisable to inform them of:
– Objectives of the study
– Selection criteria
– Calendar and maximum duration time
– Expected results and potential uses
– Expected reward (monetary, final report, others)

3rd) Preparation and launching of the questionnaires: The questionnaires are prepared in
such a way as to facilitate responses by respondents. The answers must be quantified and
weighted (year of an event, probability of an event, etc.)

4th) Exploitation of results: The objective of the successive questionnaires is to reduce

the dispersion and specify the average consensual opinion. In the second mailing of the
questionnaire, the experts are informed of the results of the first consultation, and must
provide a new response. The reasons for the differences are extracted and an evaluation
of them is carried out. If necessary, a third wave would be carried out.

*ADVANTAGES OF THE METHOD:

– It allows you to obtain information from points of view on very broad or very
specific topics. The Delphi Exercises are considered “holistic”, covering a very wide
variety of fields.
– The horizon of analysis can be varied.
– It allows the participation of a large number of people, without creating chaos.
– Helps to systematically and objectively explore problems that require qualified
participation and opinion.
– Eliminates or reduces the negative effects of “face-to-face” group meetings.

*DRAWBACKS:
– Its high cost.
– Its execution time (from the formulation period to obtaining the final results).

21
– It requires massive participation for the results to have statistical significance. But
the group must have a high degree of correspondence with the topics to be
addressed in the exercise.
– A critical part of the method is the questionnaire questions.
– Biases in the correct choice of participants.
– High number of dropouts due to weather.
After completing the 4 phases of this method, a final report is produced, which will
help in decision-making regarding the problem or objectives initially set.

2.3.4 market research

Market research has many advantages if carried out properly. To do this, it is
necessary to use various tools to collect information and make it possible to
understand the results. One of the most commonly used are online surveys, as they
have advantages such as the ability to request more complete information due to
the multiple formats it supports, as well as being more economical compared to
traditional surveys or other collection systems.
We have created the following infographic so that you can learn about all the
advantages of using online surveys as a tool for conducting market research.

22
 Objectives of market research
Market research has 3 different types of objectives.

 Administrative: Assist in the development of the company or business through

proper planning, organization and control of both material and human resources,
in order to meet specific needs within the market at the exact time.

 Social: Satisfying the specific needs of the client through a required good or service,
that is, that the product or service meets the requirements and desires of the client
when it is used.

 Economic: Determine the economic degree of success or failure that a company

may have when it is new to the market, or failing that, when it introduces a new

23
 product or service in order to know with certainty the actions that must be
implemented.

Benefits of good market research:

 There is more and better information to be able to make correct decisions that promote
the growth of the company and make it more efficient.

 It provides real and accurate information that helps solve future problems that may arise.

 You will know the size of the market that must be covered if you sell a product or service.

 Determine the correct sales system according to what the market is asking for, and thus
marketing is more effective.

 It helps to understand how customers' preferences (and tastes) change so that the
company can meet their preferences, purchasing habits, and income level.

 In addition to generating information that helps us know how consumers perceive us.

 You will be able to determine the type of product that should be manufactured or sold
based on the specific needs of consumers.

24
  It serves as a guide for communicating with current and potential customers.

 Market research helps to understand market trends, which is why it is important to carry
it out frequently to get to know customers better.

 It is a great investment for any business, since thanks to it, invaluable information is
obtained, it shows us the path to follow to take the right path and achieve the sales that
are required.

 By properly researching the market, we will undoubtedly be taking a step forward, and
therefore we will be taking advantage of our competitors.

Steps to conduct market research

Planning is vital to the market research process. Knowing what to do in the various
situations that may arise during your research will save you time and trouble.

Consider these tips for your market research process:

 Define the problem. Having your research topic well defined will help you when
formulating your questions. Don't forget that your questions should be aimed at
solving problems and must be adapted to the project you are carrying out. Make
sure your questions are clearly worded and understood by respondents. You can
do a test with a small group to see if the questions you are going to ask are
understandable and will give you the answers you need.

 Define the sample. A representative sample is very important. If you get answers
from the wrong people, your research will be useless. Your representative audience
must be present. You must read: What is a Representative Sample of the National
Population?

 Carry out the data collection. First you will need to develop a data collection
instrument. Failure to answer your survey, or answering it incompletely, will cause
errors in your research. Correct data collection will prevent this.

25
 Analyze the results. Each of the points in the market research process is linked to
another. If all of the above is done well, but there is no correct analysis of the
results, then the decisions you make will not be appropriate. Analyze in depth, do
not leave loose ends, the data is there to give you solutions, not more problems.
Remember that the analysis you make of the data will be reflected in a report,
which must also be written clearly and invite decision-making. What comes next
after conducting market research?

 Make the results report. The report must address the problem and provide
information in a way that is understandable to the interested parties.
Recommendations can also be made in it.

 Making decisions! Act and implement.

2.3.5 Life cycle analogy.

In this case, it is about predicting the evolution over time that the life cycle of a
certain product will have. This cycle can normally be divided into five stages:
development, introduction, growth, maturity and decline. This forecast allows us
to estimate the size of the market, and together with the market share that the
company will have, estimate the quantity of product that will be demanded.
Profits are usually achieved only in the growth and maturity stages. Additionally,
depending on the type of product, it is critical that the development stage is rapid,
particularly in those with a short life cycle or those that quickly become obsolete,
such as technology-intensive products.

A graphical representation of the product life cycle as defined above is presented

below:

26
 2.4 quantitative methods
Quantitative observation relies on numbers to analyze and verify specific data and
information. It is the empirical-analytical research par excellence. Things happen by
cause and effect, starting from quantitative questions. Hence its usefulness in the
most exact sciences such as mathematics, physics or statistics.

The quantitative method is based on empirical-analytical research. It bases its

studies on statistical numbers to provide answers to specific causes and effects.

Quantitative research aims to obtain responses from the population to specific

questions. The business purpose would be to make accurate and effective decisions
that help achieve what we were pursuing. It could be the launch of a new product.
The goal is to be successful with your market positioning. And to do this, it is
necessary to carry out a prior study using the quantitative method, for example. To
then use management tools that help us make those more effective decisions.

27
 CHARACTERISTICS OF THE QUANTITATIVE
METHOD
We identify some of the key elements that best define the concept of quantitative
method. Without them, we would be talking about a different type of research.

 There needs to be a numerical relationship between the variables of the

research problem.

 The data analyzed must always be quantifiable.

 It is descriptive.

 Analyze and predict population behavior.

 It focuses on a cause and an effect, or what is the same: it is based on the
application of a stimulus to obtain a response.

 The results can be applied to general situations.

 It is results-oriented.

 Numbers and data represent the most abstract reality.

 It studies human behavior and the behavior of a sample of the population.

28
 2.4.1 Time series
A time series is a sequence of observations over regularly spaced time intervals. For
example:

 Monthly unemployment rates over the previous five years

 Daily production in a manufacturing plant for one month
 The decade-by-decade population of a state in the previous century

Components of a time series

Trend
The long-term tendency of a series to increase or decrease (uptrend or
downtrend).

Seasonality
The periodic fluctuation in time series within a given period. These
fluctuations form a pattern that tends to repeat itself from one seasonal
period to the next.

Cycles
Large deviations from the trend due to factors other than seasonality. Cycles
usually occur over a long time interval, and the times between successive
peaks or troughs in a cycle are not necessarily equal.

Irregular movement
The movement that remains after accounting for trend, seasonal, and
cyclical movements; random noise or error in a time series.

29
 2.4.1.1 Simple approach
Simple approach: Also called empirical forecasting. One of the simplest methods is
to use the latest data as a forecast for the next period. That is, the demand forecast
for the next period is equal to the demand observed in the current period.

For example, if the actual demand for Wednesday was 35 customers, the demand
for Thursday will be 35 customers. If the actual demand on Thursday is 42
customers, the forecast demand for Friday will be 42 customers.

This method can take into account a demand trend. The observed increase (or
decrease) in demand over the last two periods is used to adjust current demand for
forecasting purposes.

For example: If the demand was 120 units in the last week and 108 units the week
before, the increase in demand was 12 units in one week, so the forecast for the
next week is 12 units. Week will be 120 + 12 = 132 units. If the actual demand for
the following week turns out to be 127 units, then the next forecast will be 127 + 7
= 134 units.

30
 2.4.1.2 Moving averages
Simple Moving Average: Used to estimate the average of a demand time
series and to suppress the effects of random fluctuations. This method is
most useful when demand does not have pronounced trends or seasonal
fluctuations. It simply involves calculating the average demand for the most
recent “n” periods in order to use it as a forecast for the next period. For the
next forecast once demand is known, the oldest demand included in the
previous average is replaced by the most recent demand and then the
average is recalculated.

That is to say:

Ft+1 = Sum of the last n demands / n = Dt + Dt -1 + Dt-2 + ……+ Dt – n + 1

Where:

Dt = actual demand in period t

n = total number of periods included in the average

F t+1 = Forecast for period t+1

Exercise:

Taking data from the factory of a product, prepare a 5-week moving forecast to
estimate how many boxes of the product will be needed by week 11.

If actual demand in week 11 was 55 cases, the 5-week rolling forecast for week 12
would be:

31
 F12 = (55 + 60 + 55 + 61 + 58) / 5 = 57.8

The forecast is 57.8 boxes

What will be the forecast for week 13 at this time?

Answer F13 = 57.8 boxes

2.4.1.3 Exponential smoothing

 Exponential smoothing: This is a very refined weighted moving average
method that allows calculating the average of a time series. It is the most
frequently used formal forecasting method due to its simplicity and the small
amount of data it requires. Unlike the d3 weighted moving average method,
which requires n periods of past demand and n weights, exponential
smoothing requires only three types of data: the forecast for the last period,
the demand for that period, and a smoothing parameter, alpha (a), whose
value fluctuates between 0 and 1.0. To develop a forecast with exponential
smoothing, it will be sufficient to calculate a weighted average of the most
recent demand and the forecast calculated for the last period. The equation
corresponding to this forecast is:

Example:

32
 2.4.2 Causal Relationships
Learning causal relationships, or how we learn why things happen, is among the
most useful skills our minds and those of other animals are capable of exhibiting.
Assuming basic principles of Darwinian evolutionary theses, we raise the possibility
that the mental mechanisms through which some simple forms of causal
knowledge are acquired are shared by different species, including humans. This
approach will lead us to describe some experimental data obtained in the
laboratory that show that animals and humans are successful and wrong in a
similar way when learning causal relationships.

2.4.2.1 Simple regression

Simple Linear Regression: Linear regression analysis establishes a
relationship between a dependent variable and one or more independent

33
 variables. In simple linear regression there is only one independent variable.
The dependent variable is whatever you want to predict.

This model takes the form: Y = a + bX

Y = dependent variable

X = independent variable

a = Y-axis intercept

b = slope of the line

These constants a and b are calculated as follows:

Once the values of a and b have been calculated, any future value of X can be
substituted to predict the corresponding value of Y.

Simple Linear Regression Example

34
Student enrollment at a regional college has grown steadily over the past six years
as shown below. Use time series regression to forecast student enrollment for the
next three years.

Enrolled Students
Year (x 1000)
1 2,5
2 2,8
3 2,9
4 3,2
5 3,3
6 3,4

35
 2.4.2.2 Multiple regression
Multiple Linear Regression: Multiple linear regression analysis establishes a
relationship between a dependent variable and two or more independent
variables, with the aim of estimating or predicting an average value, based on
the known values. This type of forecast occurs when two or more
independent variables influence a dependent variable. Example: Y = f (x, w, z
……… n).

In multiple linear regression analysis, the following are considered:

 It is assumed that there is a linear relationship between the variables.

 The forecast has a short or long time horizon and is of medium-high
precision.
 A correlation coefficient is determined which indicates that the relationship
being investigated is potentially a useful indicator.

Multiple linear regression analysis has an equation with two additional

independent variables:

36
 It can be extended to any number “m” of independent variables:

2.5 Forecast in the service sector.

Services are all those activities that seek to satisfy the different needs that a client
may have.

Types of services

There are two broad ways to classify services. One of them is classifying it into
public and private services.

 · Public services: are supported by the State, and defend the general interest
of society.
 · Private services: which are financially supported by private initiative and
only defend the rights of the consumer of that service.

And the other is a classification of different categories and a relationship that exists
between two parties: the one who offers the service and the one who needs it as a
user to satisfy a need.

 Maintenance services: These are those that offer preventive maintenance of

items that require care.
 Home service: These are services that the client uses without leaving home
and that are contracted by telephone or the Internet.
 Rental services: These are those that a person hires to satisfy a momentary
or time-bound need.
 Workshop services: These are services offered by individuals for the
maintenance and repair of a necessary item. They usually operate within a
workshop.

Features of the services

37
The characteristics that services have and that distinguish them from products are:

1. Intangibility: This is the most basic characteristic of services, it consists of

the fact that they cannot be seen, tasted, felt, heard or smelled before
purchase. This feature makes a number of actions that might be desirable
difficult: services cannot be inventoried or patented, easily explained or
represented, etc., or even their quality measured before provision.
2. Heterogeneity (or variability): two similar services will never be identical or
the same. This is for several reasons: deliveries of the same service are made
by people to people, at different times and places. By changing just one of
these factors, the service is no longer the same, even by changing only the
state of mind of the person who delivers or receives the service. For this
reason, it is necessary to pay attention to the people who will provide
services on behalf of the company.
3. Inseparability: in services, production and consumption are partially or
totally simultaneous. These functions can often be supplemented by a sales
function. This inseparability also occurs with the person who provides the
service.
4. Perishability: services cannot be stored, due to the simultaneity between
production and consumption. The main consequence of this is that a service
not provided cannot be performed at another time, for example, a flight
with an empty seat on a commercial flight.
5. Lack of ownership: Buyers of services acquire a right to receive a benefit,
use, access or lease of something, but not its ownership. After the service
they only exist as lived experiences.

2.6 Forecast for start-up companies

The sales forecast differs from the company's sales potential. This establishes what
the company's actual sales will be at a given level of the company's marketing

38
 effort, while sales potential assesses what sales are possible at various levels of
marketing effort, assuming certain environmental conditions exist. It is also called
the technique that allows you to calculate sales projections in a fast and reliable
way, using as data sources either inventory transactions or sales invoicing.

It also allows estimating future demand, based on historical information generated

by the movement of products in the Inventory Control module or by sales in the
Billing module.

Advantages:

 Supporting decision-making by Marketing, Sales and Production

Management by providing them with consistent and accurate information,
which is calculated using mathematical forecasting models, historical data
on sales behavior and the judgment of executive representatives from each
department involved in the company.
 Greater security in the management of information related to the
company's sales. Great flexibility in forecasting and creating and comparing
multiple scenarios for the purposes of analyzing projected sales.
 Supports the decisions of the Sales department in an effective and timely
manner by forecasting the product guidelines and demands established
within the Master Production Plan. Generally accepted techniques for
forecasting fall into five categories: executive judgment, surveys, time series
analysis, regression analysis, and market testing.
 The choice of method or methods will depend on the costs involved, the
purpose of the forecast, the reliability and consistency of historical sales
data, the time available to make the forecast, the type of product, the
characteristics of the market, the availability of the necessary information
and the expertise of those responsible for making the forecast.

39
 2.7 Use of software in forecasts.
In many short-term forecasting applications, computers are indispensable.
Companies often have to prepare forecasts for hundreds or even thousands of
products or services repeatedly. For example, a large network of medical service
facilities needs to calculate demand forecasts for each of its services in each
department. This operation involves large volumes of data that must be
manipulated frequently. Analysts have to examine the time series that correspond
to each product or service in order to develop a forecast.

There are two types of computer packages that help determine the forecasting
process efficiently:

1. Statistical packages including regression analysis and other techniques

frequently used in forecasting; and
2. Forecast packages designed specifically for applications of this type.
Computer packages for forecasting.

In the past decade, the development that has had the greatest impact on
forecasting is that of computer software packages designed specifically to deal
directly with different forecasting methods.

Hundreds of statistical and forecasting packages have been developed for both
large and microcomputers (or personal computers, often called PCs). Managers
with PCs on their desks and knowledge of forecasting techniques no longer depend
on a team to make their forecasts. Modern managers are taking advantage of the
ease and availability of complex forecasting methods provided by personal
computers.

In this section, some of the most commonly used statistical and forecasting
computing packages are mentioned. Statistical packages that run on mainframe
computers are often used and include sections that directly deal with various
forecasting methods.

40
 Three of the most popular packages are:

1. Minitab: Features menus and dialog boxes, while maintaining the command
language to add speed and flexibility.
2. Statistical Package for the Social Sciences (SPSS): statistical package for the social
sciences.
3. Statistical Analysis System (SAS): statistical analysis system.
In recent years, a new type of computing package has been developed for
mainframe computers, specifically aimed at the needs of administrators. Some of
the various packages specifically designed to run on microcomputers and currently
on the market are:

1. Autobox 3.0 by Automatic Forecasting Systems, Inc.

2. Business and economic forecasting: decision support system software by John
Wiley.
3. Easy forecaster plus 1 or 2 From the Institute of Business Forecasting.
4. Cast by Scientific System, Inc.
5. Micro – BJ by Stratix
6. Hewlett Packard Forecasting
7. MICCROFIT3 from Oxford University Press.
8. Forecast Pro by Business Forecast Systems, Inc

Conclusion
In conclusion, there are many demand forecasting methods among the most
common and basic are: Moving Average (MA), Weighted Moving Average
(WMA), Simple Exponential Smoothing (SES) and for the topic of sales trends

41
 we have regression analysis, in its linear, logarithmic, exponential and
potential trend versions, finally forecasts can also be generated using the
"Moving Annual Trend" method. Highly recommended for companies with
historical data and certain stability in sales.

Literature
https://prezi.com/iah24lsolcxk/21-importancia-estrategica-del-pronostico/

https://sites.google.com/site/ittijuanaao2016/1-5-estrategias-de-operaciones-en-un-entorno-
global/2-1-caracteristicas-de-la-demanda

https://www.sinnaps.com/blog-gestion-proyectos/metodologia-cualitativa

https://ingenioempresa.com/consulta-fuerza-de-ventas/

https://ingenioempresa.com/jurado-opinion-ejecutiva/

http://www.eoi.es/blogs/nataliasuarez-bustamante/2012/02/11/%C2%BFque-es-el-metodo-
delphi/

https://www.questionpro.com/es/investigacion-de-mercados.html

https://www.gestiondeoperaciones.net/proyeccion-de-demanda/metodo-del-ciclo-de-vida-del-
producto-para-pronosticos-de-ventas/

https://www.sinnaps.com/blog-gestion-proyectos/metodo-cuantitativo

https://support.minitab.com/es-mx/minitab/18/help-and-how-to/modeling-statistics/time-
series/supporting-topics/basics/what-is-a-time-series/

https://pymex.com/pymes/estrategias-de-crecimiento/tecnicas-y-modelos-de-pronostico-para-
nuestra-empresa-parte-3

http://paginaspersonales.deusto.es/matute/psicoteca/articulos/Lopez&Luque03.htm

https://pymex.com/pymes/estrategias-de-crecimiento/tecnicas-y-modelos-de-pronostico-para-
nuestra-empresa-parte-5

42