Market Assessment
Market assessment is the systematic use (i.e. analysis) of hard data and judgments
about customers, organisations, competition, and industry context to support
strategic marketing decisions. It involves analyzing the structure/dynamics of a
market which the organisation competes in, or wishes to compete in.
Formal assessment tools help to make a judgment of the market and how it behaves
(i.e. to do market assessment), and are used for the following five reasons:
1. Formal assessment tools enhance decision making capabilities, improving
decision efficiency and effectiveness.
2. Formal assessment tools improve problem/opportunity recognition by
collecting, integrating, organizing and presenting knowledge.
3. Formal assessment tools avoid the limitations of cognitive constraint that
result from busy schedules and human bias.
4. Formal assessment tools force you to use a problem-solving approach by
focusing on what objective to achieve.
5. Formal assessment tools force you to list alternative actions.
There are three main assessment methods to help managers make better decisions:
1. Prediction methods: used to identify market trends/opportunities.
2. Response methods: used to understand relationships in the market.
3. Decision methods: used to compare the impact of alternative decisions.
The Concept of a Decision Calculus
The problem with management science models is that managers practically never
use them, mainly due to the following four reasons:
1. Good models are hard to find: convincing models that successfully incorporate
managerial input are difficult to build.
2. Good parameterization is even harder, since good quality data can be
difficult/expensive to gather.
3. Managers don’t understand the models: managers prefer intuitive models and
reject those they don’t understand, because they carry responsibility for
outcomes.
4. Most models are incomplete: incompleteness is a serious danger if a model is
used for optimization.
1
,If managers are to use a model, design it around a decision calculus: a model-based
set of procedures for processing data and judgements to assist a manager in his
decision making. After all, a model is meant to guide the manager’s decision making,
not to replace the manager. A decision calculus should therefore meet the following
requirements, based on the aforementioned pitfalls:
1. Simplicity: a user can easily obtain a basic understanding of the model and
how it should be used.
2. Robustness: a user should find it difficult to make the model give implausible
answers, by constraining the output to a realistic range of values.
3. Manageability: a user should be able to make the model behave the way he
wants it to.
4. Adaptivity: the model should be capable of being updated as new information
becomes available.
5. Completeness (on important issues): since completeness is in conflict with
simplicity, make sure to incorporate subjective (managerial) input to get the
best of both worlds.
6. Communicable: outputs should be easy to interpret.
2
, Assessing Market Developments: Prediction Methods
Marketing forecasts are foundational to a marketing plan, allowing managers to
strategically plan the future and guide decision making. Not all methods are
sophisticated, but an educated ‘best’ guess is more valuable than no forecast - and
thus no planning - at all. Forecasts are almost always wrong. However, making
predictions is not about getting the future right, it is about reducing uncertainty.
In order to manage repetitive business activities (of tactical nature) and identify new
market opportunities or problems (of strategic nature), organisations predict factors
such as future sales, consumer behavior and competitor behavior. Marketing
managers benefit from these predictions by gaining insights into future trends,
improved targeting, increased customer retention, more proactive planning, more
precise budgeting and better inventory management.
Qualitative vs. Quantitative
Prediction methods are either qualitative methods (based on expert or managerial
intuition) or quantitative methods (based on statistical models). Both methods have
their strengths and weaknesses:
- Experts suffer from human bias, overconfidence, boredom, social pressure due
to organizational politics, fatigue and emotions, where models do not.
- Experts do not consistently integrate evidence, where models do.
- Experts diagnose (i.e. identify new variables) and predict, where models only
predict.
- Experts can provide (subjective) evaluations of variables that are difficult to
measure objectively, where models can not.
- Models are dependent on data availability, while experts are capable of dealing
with subjectivity and a lack of data.
- Models are consistent, but also rigid. Experts are flexible in adapting to
change, but also inconsistent.
- Experts have highly organized, domain-specific knowledge that allows them to
respond to ‘broken leg’ cues that are very diagnostic but so rare that they are
difficult to anticipate and include into a statistical model.
Experts and models are both complements and substitutes: both take into account
much of the same decision relevant information, but where one decision input is
weak the other is stronger and vice versa. In general, models outperform experts
during stable markets but not during changing markets. A model that combines
qualitative and quantitative prediction methods always outperforms one of its parts.
3