Sales Forecasting and Market Research

Sales forecasting

Through sales forecasting, companies predict future sales. As a result, sales data is valuable for businesses. Companies use it as a basis for creating budgets and business plans, including those related to cash flow, production, inventory, and financing.

• The production department uses it to make production plans, including how many units to produce and how much material to order.
• The marketing department uses the sales forecast as a target for them to achieve.
• Human resources departments use it for planning recruitment, training, and development.
• The finance department uses forecasting data to plan cash flows, working capital, and capital budgeting.

Sales forecasting method

How to forecast sales? Forecasting methods broadly fall into two categories:

• Qualitative forecasting
• Quantitative forecasting

Qualitative forecasting techniques rely on personal judgment, like the Delphi technique, which uses independent experts’ judgment. Another way is an executive opinion and market surveys.

Quantitative forecasting techniques rely on data and use statistical models such as time series and regression. For example, the time series models could be Autoregressive (AR), Moving Average (MA), and Autoregressive Integrated Moving Average (ARIMA). Meanwhile, the regression model can be simple regression, multiple regression, and panel data regression.

Two approaches to forecasting. First, the company directly applies the above method to forecast the company’s sales. Second, the company uses the above techniques to predict market demand and then uses the results to derive sales forecasts.

Let’s take an example for the second case. A company uses a time series model to forecast market demand. The model predicts market demand will increase by 10% to USD100 million next year. Meanwhile, management targets the market share to remain unchanged at 15%. The company’s sales are estimated at USD15 million next year from these two data.

Estimating market demand using quantitative models must be matched to historical data patterns, which could be:

• Seasonal demand – demand varies within a year due to seasonal effects such as hotel demand in the tourism sector.
• Random demand – demand does not form a specific pattern.
• Cyclical demand – demand follows/reverses economic cycles.

Benefits and limitations of sales forecasting

Benefits

• Improved cash flow and working capital
• Better inventory control
• More efficient operations and production facilities
• Help secure external resources, including raw materials and finance

Limitations

• Bias in predictions, resulting in the inaccuracy of the output
• Errors in decision-making and business planning due to prediction errors

Market research

Market research is about systematically investigating a market. Companies collect information about markets, customers, competitors, and marketing strategy effectiveness through research. The company uses this information to:

• Identify and define external opportunities and challenges
• Predict future trends
• Identify what the market needs and wants
• Design and evaluate marketing strategies and tactics
• Monitor marketing performance
• Reduce risks in the marketing area

Research in marketing can include:

• Market research
• Product research
• Sales research
• Promotion research
• Marketing environment research
• Competition research

Market research provides valuable information about the following:

• Conditions and changes in the marketing environment
• Market size and growth potential
• Market structure and competition
• Consumer characteristics
• Company and competitor sales data
• Competitive strategy in the market
• Company and product positioning

Reasons for doing market research

• Reducing risks associated with marketing strategies, including when launching new products
• Help improve the marketing strategy
• Evaluate opportunities and challenges in the market
• Explore conditions and predict future trend changes
• Assist in decision-making in marketing planning and resource allocation
• Obtain input to develop a unique selling proposition and competitive advantage
• Explore activities and strategies used by competitors
• Assess customer reactions by testing the product on them

Types of market research

Qualitative research vs. quantitative research

Market research is divided into two based on the type of data or information we use:

• Qualitative research
• Quantitative research

Qualitative research relies on non-numeric data such as text, video, or audio recordings. It is usually used to understand motivations, attitudes, opinions, ideas, behaviors, and perceptions through focus groups, expert panels, or in-depth interviews with credible individuals.

Characteristics of qualitative research

• Use non-numeric information
• Helps to understand behavior and opinions/attitudes more deeply
• Time-consuming and longer to gather information
• Provides richer data because responses can vary widely between individuals
• Relatively cheap because it usually uses fewer samples

Quantitative research relies on numerical or measurable data. This research usually requires a large enough sample size to increase generalizability. To draw conclusions, we can apply statistical techniques.

Characteristics of quantitative research:

• Rely on numeric or quantifiable data
• Requires a larger sample size to increase generalization
• Can use statistical tools to describe and draw conclusions
• Relatively quick to describe and draw conclusions

Primary research vs. secondary research

Market research is divided into two based on how we get the data:

• Primary research
• Secondary research

Primary research or field research relies on data from original sources directly. In other words, it involves collecting primary data. Companies may do it themselves or hire market research agencies who are experts in conducting such research.

Primary research can be through:

• Interview (face-to-face or telephone)
• Surveys (field, online, or phone)
• Focus group
• Hall test
• Observation

Advantages of primary research

• Can be customized according to needs
• More accurate data
• Can get more in-depth information
• Confidentiality and data privacy
• More up-to-date data

Disadvantages of primary research

• Expensive to collect
• Requires more resources
• Time-consuming to collect data and information
• Bias, for example, due to errors in sampling

Secondary research relies on data from non-original sources. In other words, researchers use other people’s data or secondary data. We also call it desk research.

Data and information can come from:

• External research publications
• Academic journal
• Company report
• Government statistics and publications
• Trade association reports
• Newspapers and magazines
• Internet (feedback from customers)

Advantages of secondary research

• Cheap and fast because the data already exists
• Saving time
• Widely available

Disadvantages of secondary research

• Unable to dig deeper into topics or information
• Inaccurate data
• Outdated data
• May contain bias

Primary research methods

Survey

Surveys rely on questionnaires to ask respondents. Surveys can be face-to-face, telephone, online, or post-like interviews.

The most straightforward survey might involve visiting respondents’ homes or workplaces. Alternatively, there is also the mall intercept, where people stop at the mall and are surveyed about a topic.

Survey questionnaires can be open or closed. Open questions are usually for qualitative research to dig for more profound information. While closed questions are generally for quantitative analysis where respondents have been given alternative answers to choose from.

Advantages of the survey

• More flexibility in digging for information
• Generate qualitative and quantitative data
• More structured than an interview
• Requires less interviewer training

Disadvantages of the survey

• Difficult for participants to tell the truth about controversial questions
• Low response rate
• Requires large sample sizes
• Time consuming and costly

Focus groups

Focus groups bring together a small group of people in a location to answer questions and discuss specific topics. Participants are encouraged to express their views and opinions freely on the chosen subject. This method is important for gaining insight into participants’ experiences and perspectives.

Advantages of the focus group

• High opportunity to explore qualitative information such as opinions, feelings, and attitudes
• Explore topics in depth

Disadvantages of the focus group

• Difficult to extract qualitative information
• Expensive and time-consuming

Consumer panel

Consumer panels are similar to focus groups. But, in a focus group, the group is disbanded after the interview, and another group is selected. In contrast, in a consumer panel, the same group is asked for their opinion at a certain point in time after some change has been introduced.

Consumer panels could be more accurate. This is because the panel asked the same people to provide better insight into how they thought and felt when the situation was changed.

Interview

Interviews use questions and answers to obtain information. It can be through face-to-face interviews, where the interviewer meets and communicates directly with the interviewee. Or, it could be over the phone and online.

Advantages of the interview

• Communication can be two-way and reciprocal
• Researchers have the opportunity to encourage respondents to answer and find out more information
• Errors and misunderstandings can be resolved immediately

Disadvantages of the interview

• Interviews can be expensive and time-consuming
• Interviewers need training and communication skills
• Interviews are unpopular with consumers because they interfere with the privacy

Hall test

Test halls gather respondents to a specific location, usually, a hall, to respond to stimuli or to test a product or advertisement and ask for their opinion.

Advantages of hall test

• Opportunity to collect qualitative information
• Can be done without disturbing people’s busy lives

Disadvantages of hall test

• Bias because respondents may feel obligated to provide a favorable opinion
• Low representation because the respondents are only those who are willing
• Fewer opportunities for extensive questions

Observation

Under this method, the researcher observes how people behave. It may involve the direct observation and does not involve direct contact with the respondent. For example, they observe customer behavior in a store, such as how long they shop and which shelf they look for first. Or it retrieves data from a recording device such as a security camera.

Advantages of observation

• Avoid being subjective like when interviewing or surveying
• Done without disturbing customer activity
• Relatively easy to do

Disadvantages of observation

• Need a long time
• Distortion when customers know they are being observed
• Just answering what happened but not why

Experiment

Experimenters use scientific procedures to make discoveries, test hypotheses or demonstrate known facts. For example, a company introduces a limited product to test consumer reactions. It may involve changes in features, packaging, advertising, and pricing settings.

Experiments can be:

• Laboratory experiments by inviting participants to a specific artificial setting and asking them to taste the product or try it on their own.
• Field experiment by selecting a specific geographic area and launching a product in that location to see how people react.

Questionnaire

A questionnaire is printed or written questions to be asked to respondents. It includes closed questions and open questions. Closed-end questions present alternative answers to respondents and ask them to choose. Meanwhile, open-ended questions allow respondents to answer according to their own way.

Designing a questionnaire requires us to pay attention to the following principles:

• Explain the purpose of the research or survey
• Formulate unambiguous questions
• Use understandable language
• Avoid leading questions
• Develop questions in a logical sequence

Types of questions

The questions in the questionnaire could be:

• Open question
• Closed question

Open questions allow respondents to answer in their own way. And researchers do not provide guidelines for answering.

Open questions are commonly used for qualitative research where the researcher wants more in-depth information. For example, they may want to know opinions or motives.

However, the answers to questions can vary greatly. Moreover, each individual may give a different answer. Thus, it is challenging to classify and therefore draw conclusions.

Examples of open questions:

• What do you think about iPhones?
• What’s the first word that comes to mind when you hear iPhone?
• What factors do you consider when buying an iPhone?

Closed questions provide alternative answers to respondents. The surveyor asked them to give choices for the solutions provided.

Closed question answers might be:

Dichotomous: there are only “Yes/No” answers

Multiple choice: there are two or more possible answers

Likert scale: respondents indicate their level of agreement/disagreement, for example:

1. Strongly agree
2. Agree
3. Doubtful
4. Disagree
5. Strongly Disagree

Semantic differential scale: connects two opposite words, and the respondent chooses the point that represents his opinion

Importance scale: to rate how important something is, for example:

1. Very important
2. Important
3. Rather important
4. Not too important

Rating scale: a scale used in which respondents rate some attribute (e.g., service quality) from ‘very good’ to ‘very bad,’ for example:

1. Very good
2. Good
3. Neutral
4. Bad
5. Very bad

Factors to consider when choosing a research method

Several factors influence which research method is chosen, including:

• How much budget is available: primary research may be required if the budget is available
• How accurate is required: primary research is an option if we want more accurate results
• How quickly information is needed: secondary research is ideal because we don’t want it to take a long time to gather data or information
• How easy is it to get a sample: Primary research may not be done if getting a sample is difficult

Reasons for unreliable market research data

Market research data may not be reliable for the following reasons:

• Questionnaires may have misleading or leading questions
• The interviewer may direct the respondent to provide answers
• The language cannot be understood by the respondent
• Respondents may not fully understand the question
• The interviewer filled out the form himself
• Respondents did not give real answers
• Selected respondents do not represent the population due to sampling errors

Ethical considerations in market research

Some ethical considerations in conducting market research include:

• Damage or harm to the participant mentally or physically.
• Deception by distorting data and lying about statistics
• Plagiarism by quoting other people’s work or research
• Disclosure of privacy and confidentiality of respondent data
• Detachment or denial of the opportunity to fully understand the context being studied

Sampling

Sampling is selecting the target respondents (sample) from the studied population. It’s cheaper than having to research the entire population.

An accurate sample must represent the population. Thus, the conclusions obtained can be used to generalize the population. In other words, the sample will produce similar results if the researcher asks everyone in the population.

• Sample size points to the number of respondents selected from the population in which marketing research is conducted. In general, the larger the sample size, the more accurate the results.

Types of sampling

Sampling methods fall into two groups:

1. Random sampling
2. Non-random sampling

Random sampling

Random sampling is when each sample has an equal chance of being selected. It can be simple random sampling, systematic random sampling, stratified random sampling, cluster random sampling, and multi-stage random sampling.

• Simple random sampling – using random numbers to select a sample.
• Systematic random sampling – every n-th member in the population is selected as a sample. For example, selecting every 10th order member until the required sample size is reached.
• Stratified random sampling – dividing the population into groups (strata), e.g., based on age and occupation, social class, and then randomly selecting a sample from each stratum.
• Cluster random sampling – dividing the population into groups (clusters), for example, based on their locations, and then randomly selecting some of these clusters as a sample.
• Multi-stage random sampling – takes a sample from the population using smaller and smaller groups at each stage and is usually used when the population is geographically dispersed.

Advantages of random sampling

• Representative
• Free from bias

Disadvantages of random sampling

• Time consuming and costly
• Impractical for some methods
• Complete population lists may not be available

Non-random sampling

Non-random sampling is if each sample has a different chance of being selected. Examples are quota sampling, judgmental sampling, convenience sampling, and snowball sampling.

• Quota sampling – dividing the population into mutually exclusive subgroups and then taking a non-random sample unit until it reaches a quota.
• Judgmental sampling – selecting respondents based on what the researcher thinks is appropriate for their research.
• Convenience sampling – selecting anyone available and willing to be a respondent, for example, by stopping passers-by and asking buyers in only one location.
• Snowball sampling – taking the following sample based on what was recommended by the previous respondent.

Advantages of non-random sampling

• More practical for researchers
• Faster and more cost-effective

Disadvantages of non-random sampling

• Bias to generalize results
• Unrepresentative sample