A qualitative forecasting method that obtains forecasts through group consensus

The three primary approaches used in qualitative forecasting are the expert opinion approach, the Delphi method, and the market survey approach.

The expert opinion approach is simple and easy to implement. For example, for many of the stand-alone, one-time activities that take place in a project, an opinion based forecast is all that is either necessary or desirable. The opinion of the person who is most knowledgeable in that field is sought. Furthermore, if a project is brand new, the likes of which have never been seen before and for which no historical data is available, then the only recourse for a project manager is to seek the opinion of an expert to get a forecast or an estimate regarding the concerned event or activity.

The disadvantage of relying on the opinion of a single expert is the inherent element of bias. Further, larger issues in the project may arise where an opinion based forecast of a single expert may be not be adequate. This can occur with forecasts involving such things as the timing of the introduction of a new technology into the market place or a change in public behavior as these could have a significant bearing on the decision to start a project or the timing of market entry. When a new product is introduced it can become a guessing game as to how the market will respond and how and when competitors might respond. Answers to questions such as these may require the opinions of several experts, perhaps across a range of subjects, not simply an opinion from those closest to the job. In such cases, the Delphi method may an appropriate forecasting method.

Devised by the Rand Corporation in the U.S., the Delphi technique is a popular method of qualitative forecasting that generates a view of the future by using the knowledge of experts in particular fields. The name derives from the ancient Greek Oracle of Delphi that was supposed to foretell the future. The steps of the Delphi method are as follows:

    1. Questionnaires are circulated to the team members, who may not be aware of each others' identities, and each is invited to make his own prediction of future progress in a particular field. As far as possible, projections must be quantified and the questions must be framed accordingly: e.g., what proportion of all households do you expect to have a personal computer by the year 2010?
    2. After the first round of replies, the results are analyzed statistically (giving the distribution of responses) and the results re-circulated; panel members are asked to reconsider their views in the light of the new statistics. If their view lies outside the inter-quartile range, they must either revise their opinion or give their reasons for their extreme view; this will be seen by the other panel members. This process can be repeated for a third or fourth round until a consensus of opinion is obtained.

Results of Delphi studies are given in the form of timescales and probability levels for the feature being forecast. Some large corporations have used the method for assessing long term trends and the development strategies that may be open. Research by the Rand Corporation indicates that with current technologies and trends, the Delphi panel does tend to move towards a consensus view which is generally correct, but there tends to be less accuracy when forecasting new developments. On occasions, no consensus view is obtained after several rounds.

The market survey approach is the third qualitative approach that can be used to generate forecasts of project events. This approach involves surveying past customers or potential customers about any plans they may be considering the future. The project organization's marketing staff is perhaps the ideal source to obtain such information because of their direct contact with customers. In addition, the marketing staff, along with the procurement staff, which is in direct contact with suppliers, can also provide market intelligence reports regarding competitors who are contemplating new projects or new technologies.

Qualitative forecasting techniques are based on human judgment. It revolves around the knowledge of customer journey, market research, experts’ opinion, or senior industry leaders’ experience in the field. It is more subjective, intuitive, and experiential than other forecasting techniques. Read ahead to learn more about the most common qualitative techniques used in businesses.

In our previous post – ‘fundamentals of business forecasting’, we talked about two types of forecasting techniques, i.e., qualitative, and quantitative. Here, we elaborate on the types, applications, and limitations of qualitative forecasting techniques.

Businesses employ various qualitative forecasting techniques, individually or in combination, to generate better strategies and direct business tasks such as production, demand planning, marketing, and sales.

Types of qualitative forecasting techniques

Qualitative forecasting techniques are subjective, intuitive, and experiential where the industry experts apply their business knowledge and experience. The commonly used qualitative forecasting techniques are Delphi, Market research, Panel consensus, and Visionary forecast.

A qualitative forecasting method that obtains forecasts through group consensus
Qualitative forecasting techniques

Delphi Technique:

The main objective of the Delphi technique is to construct consensus forecasts from a group in a structured manner. The technique involves a repeated cycle of questionnaires presented to a panel of select experts. The independent answers provided by the expert team are then forwarded to the facilitator to analyze and summarize the opinions.

Highlights of Delphi method

A few of the highlights of this method include

  • Freedom of expression: The anonymity of the experts allows them to express their opinions without any social pressure.
  • Right to change opinions: The experts are free to change their opinions at any time without the fear of criticism.
  • Regular feedback: The experts are informed about the other participants’ opinions in the study after each round and allowed to comment on the responses. This helps to revise their own forecasts and opinions.
  • Quantitative results: Though qualitative in nature, it involves some form of statistical aggregation, paving the way for quantitative analysis as well.

Steps involved in the Delphi technique

It is a simple technique, typically involving the following steps:

  • A facilitator is appointed to manage the entire process.
  • A panel of industry experts are selected.
  • Forecasting challenges are set and distributed to the experts.
  • The initial forecasts are collected, compiled, and summarized.
  • The feedback is then returned to the experts who can revise their responses based on the feedback.
  • These steps are repeated until a satisfactory level of consensus is reached.
  • Final forecasts are constructed by summarizing the experts’ opinions.

These repeated rounds of questionnaires allow a consensus to be reached.

A qualitative forecasting method that obtains forecasts through group consensus
A visual representation of the Delphi method

When to use the Delphi technique?

Delphi technique is recommended when the problem at hand needs collective and subjective decisions. It can be used to:

  • predict trends in automation, sales, etc.
  • forecast outcomes related to economic development such as improved education, employment opportunities, and income.
  • forecast outcomes of an epidemic/pandemic disease (for example, COVID-19 leads to loss of lives, hampers economic status of the affected nation, brings financial fragility to small businesses).

Limitations of Delphi method

Delphi is time-consuming and laborious for both facilitators and experts. It may lead to loss of consistency or distraction between rounds as it demands long-term and continuous commitment.

Market Research

Market research is a popular qualitative forecasting method used in business. It forecasts future demand through consumer surveys and questionnaires. Businesses apply this technique to gain valuable insights from consumers when, for example, introducing a new product into the market.

It differs from the Delphi technique because public opinions are considered here to glean insights, and not just those of experts. Market research methodologies also change according to evolving market challenges.

Highlights of market research method

Some of the highlights of market research are that it

  • Helps to introduce new products, identify potential markets, select appropriate marketing techniques and promotion measures
  • Introduces consumer-oriented marketing policies
  • Assists companies in studying marketing problems and arriving at solutions
  • Reduces the gap between the producers and consumers

Steps involved in market research

The typical steps involved in market research are:

  • Defining the problem statement
  • Identifying the representative sample
  • Collecting the required data
  • Analyzing the results
  • Creating a research report
  • Arriving at decisions

A qualitative forecasting method that obtains forecasts through group consensus
Typical Market Research process

When to conduct market research?

When business professionals need to estimate the size of the market, define potential customers, understand the reasons behind low sales or low profitability, or support company growth. Market research helps to understand customers’ pain points, target future markets, control inventory for low-volume products, and forecast margins.

Limitations of market research method

The results of Market research may not be always accurate because the consumers’ behaviors are uncertain. Additionally, this methodology demands extensive research, and qualified & experienced specialists, both of which require significant time and money to obtain.

Panel consensus method

Panel consensus is a qualitative forecasting technique that brings all the internal experts of an organization together for an open discussion about a product or service. Anyone may speak up, and the meeting will end when a consensus is reached. The accuracy of the forecast is ex post facto verified against actual sales data.

It is also known by other terminologies such as forecasting averaging, model averaging, ensemble averaging, or expert aggregation. This method differs from Delphi or market research in involving only the experts internal to the organization.

Highlights of panel consensus

  • It involves experts from different company verticals.
  • The forecast period and frequency of the exercise can be varied as per the need.
  • Different forecasting models are deployed as per data availability.
  • Department-specific forecasts can be created.

Steps involved in consensus forecasting

The process typically involves

  • Defining the problem statement
  • Providing everyone a chance to suggest solutions (Through brainstorming or round-robin)
  • Discussing the solutions
  • Deciding on the best solutions
  • Acting upon the decision
  • Evaluating the decision

A qualitative forecasting method that obtains forecasts through group consensus
Consensus forecasting

When to use the Panel consensus method?

Panel consensus process can be undertaken periodically for a set of narrowly defined variables. The forecast acts as an estimator of risk or ‘less likely’ scenarios.

It can be used for

  • incorporating the most recent economic or political developments
  • detecting the trends in the market at an early stage (determine the annual gross domestic product of a country, or the quantity of a company’s product that is likely to sell in a year).

Limitations of panel consensus method

The drawback of this method is that it relies on experience and personal opinion. It may lead to bias in the final forecasts as this is a ‘top-down’ method. The cost is high, and it cannot be used granularly for each product, but can be used to forecast an overall demand.

Visionary forecasting

In contrast to using group think, visionary forecasting is usually based on the individual opinion or judgement of an experienced and respected individual in the field. In this method, a set of future scenarios is usually determined by the “visionary” based on past events. It is therefore based on subjective guesswork and somewhat non-scientific.

Highlights of visionary forecast

Visionary forecasting is characterized by “vision” of the expert being consulted. It incorporates intuitive judgment, opinions, and subjective probability estimates. The company’s senior professionals are encouraged to predict new product development by observing past events and developments.

Steps involved in visionary forecasting

Visionary forecasting is a fairly straightforward process that typically involves

  • Studying the market environment
  • Observing past events
  • Estimating sales of a new product
  • Predicting development of new products

A qualitative forecasting method that obtains forecasts through group consensus
Visionary forecasting

When to use visionary forecast?

The visionary forecasting technique is best used when there is a lack of historical data. It is used to estimate new trends in new and unique market conditions.

Limitations of visionary forecast

An individual’s judgment may lead to confirmation bias. A decision-maker may seek evidence that only validates their beliefs while ignoring the evidence that support differing conclusions.

Combining qualitative forecasting techniques

It should be evident by now that no forecasting technique is appropriate for every industry or situation. Studies suggest that combining the individual forecast methodologies brings more accuracy and acts as a compensatory mechanism to overcome the deficiency in any particular technique. Choosing complementary methods eliminates the shortcomings of one technique by substituting it with the advantages of another.

Conclusion

Qualitative forecasting techniques help organizations to reduce ambiguity in data, predict changes in sales patterns, and customer behaviors. It gives the team required flexibility when using soft data.

Qualitative forecasting techniques, individually or in combination, enables organizations to analyze the current scenario, recognize changing trends in the market, and to connect that information to potential future strategies.