Methodology & System

The system is based on five different dimensions of a market. Main system outputs are a combination of two or more of these dimensions. For example, a product’s market size over time or a product’s market size for a particular market segment in a particular geography are system outputs.

Time

The system can track and forecast markets over a wide time horizon and with a user-defined granularity of time periods. The current implementation contains 20 quarters of historic market data and 20 quarters of current and/or forecasted market data. Since all market allocations performed by the system can be performed in each time period, the addition of more time periods adds significantly to the computing requirements.

Product

A product is a distinct item, often a service, offered by companies in a particular industry. A complete accounting of product categories and products for a given industry is necessary in order to calculate the market size of the entire industry.

Products are generally grouped into product categories. Market sizes by product or product categories for a given sub-region and a given time period are calculated by the system.

Market Segments

It is often useful to divide markets into market segments in order to analyze buying behavior and target products to specific types of customers. This system provides a method of dividing and allocating markets to specific market segments.

Psychographic information defining market segments and the buying habits of each segment is sometimes available. This system provides a means to accommodate such data . It also accommodates the allocation of a market into segments external to the system.

Geography

An important component of this system is the ability to allocate regional markets to an arbitrary number of levels of sub-regions. Regions are broadly defined and might be a nation such as Japan, or an area of a nation such as the Western U.S., or an international region such as South America or Europe.

Sub-regions are multi-level and typically determined by the available demographic data and government defined boundaries. The area hierarchy defines the sub-regional in a particular region. Customized sub-regions can be aggregated from the defined sub-regions.

Service Provider / Vendor

In any industry, there are companies or government entities that comprise that industry. The companies operate in a highly competitive environment or a monopolistic environment for particular products and particular geographies. These entities are referred to as service providers or vendors.

An area hierarchy is a collection of a region and a number of levels of sub-regions, for the particular industry and geography being processed by the system. An area hierarchy describes the geographic parent-child database relationships used in the calculations. For example, one area hierarchy for the U.S. market has the entire nation as the region and three levels of sub-regions as follows:

• Nation
• State
• Metro Areas (MSA)¹
• County

Furthermore, once the system calculations are complete, post-processing can also include a roll up of data into any geographically based region or sub-region that is desired as long as the sub-regions can be uniquely mapped into the desired region. An example of the usefulness of this feature is the customized territories and regions many companies or even government entities have for administrative and other business purposes. This feature produces customized market data particular to the unique regions defined. Post-processing can also include mapping of these regions.

The system requires an extensive annotation facilities, permitting the creation, categorization, searching, filtering, and flexible reportage of notes covering virtually any data point stored in the database. The annotation system allows for the consistent use of input data by providing the user with a readily accessible reference to the source of data and its inherent assumptions.

Allocation Fractions - Market Segments

The segmentation of any market into constituent market segments is a classic approach to understanding and targeting customers. Most available market research data is not segmented. This system is capable of accommodating any market segmentation scheme.

In the simplest case of two segments, the second segment is by default the remainder of the market. For example, in the case where a market is divided into business and residential segments, if the business segment fraction is 40%, then the residential segment is 60%. Market segmentation can be very complex with many different categories of customers such as urban seniors or rural families.

The Allocation of Markets to Market Segments

The allocation of markets to the market segments can be done at the region or sub-regional level and is product-specific. It is generally done in advance of allocation to providers and can be calculated separately for each time period.

The allocation fraction for the segment can be assigned by the industry analyst or calculated using market factors such as the propensity to buy by product by market segment. The allocation fraction can be calculated as follows:

where pop_k is the total population for the factors in the calculation (e.g. number on business in a certain industry), pop_i is the population for the particular bracket within that factor and rating_i is a ranking of importance of the bracket. t is required to adjust for relative scale across market factors. For example, one business may carry more weight in terms of buying power than one household for a particular product.

where S_T is the sum of all allocation fractions for all n segments for a given product. AF_s(i) is the allocation fraction for the market segment i. AF_s(i) can be assigned by an industry analyst or calculated by the system. The sum of all allocation fractions will generally be equal one² for any product and region.

This system calculates a score for any region or sub-region by product. The algorithms sum a series of terms; each term and the overall algorithm have coefficients that can be tuned to bring each market data point closer and closer to the actual market. The algorithms use reliable, external market data inputs, such as demographic or firmagraphic forecasts, to calculate the first market data value. The coefficients and profiles refine that initial value. The system’s coefficients and profiles use qualitative and quantitative inputs from industry expert knowledgeable in the particular market dynamics of the industry and the product being processed by the system.

Geographic allocation fractions are computed using the input data. The algorithm includes the calculation of an effective buying population:

where pop _j,k is the actual population for a particular bracket of a market factor and rating_j is the propensity to buy score for that population, for a particular market factor with n brackets and a particular geographic area k.

The total effective buying population over all market factors can then be calculated as follows:

The allocation fraction for a given sub-region and product can then be calculated.

The Allocation of Markets to Geographic Areas

Markets are allocated using the method described above in a top down fashion i.e. from the largest region to the smallest sub-region. This approach protects the integrity of the external market data at the largest region. This is not to say the system can not calculate a new market size for the entire region by aggregating market sizes from sub-regions.

From the lowest level, new aggregated regions can be customized. This can be done simply if the geographic boundaries of the customized region are coincident with the boundaries of the lower sub-region. If the boundaries are not aligned, then further processing is required.

A second method is used to protect the data integrity. This method is described under the Self Adjusting System section.

Allocation Fractions – Service Providers / Vendors

A similar method can be used to quantify the market share of service providers / vendors active in the specific region or sub-region. In this case, the product profile is replaced with a Service provider profile, the geographic weighting is replaced with a provider weighting and Geo Gain is replaced with a Market Gain.

The Allocation of Markets to Service Providers / Vendors

The algorithm for a given product in a given region or sub-region for a given time period can be calculated using:

where M_c is the current market size for that service provider, M_T is the total market size, M_a is the size of the market available to capture, and rating_i is the rating for a particular market factor such as brand name recognition or advertising budget. l and b are weighting coefficients. l is referred to as the Market Gain. AF_marketshare is the market allocation fraction. Ratings are given using a scale such as 1 to 100.

Forecasting is a method of predicting a future market size. Forecasting can be done manually time period by time period by assigning or calculating a growth rate to the each time period for a given product at the highest level of the hierarchy. The forecasting methodology is independent of this system. A variety of traditional forecasting techniques can be applied to the input data such as regression or S-curve analysis.

Product profiles are also be used for forecasting. In this case, each product is assigned one of a series of profiles that generically describe a set of product growth trends that often vary by the stages of a product life cycle. Product profiles can be displayed as a line or bar chart of growth rates versus time. Examples of product profiles include a declining product that has reached the end of its life cycle and has increasingly smaller growth rates in each subsequent time period

Inter-regional forecasts can be accomplished by holding a single region constant and reallocating all market share assigned to it to all the other regions under study. This is especially helpful when looking at traffic between regions.

The system is self-adjusting – meaning that known or verified sub-regional market data points can be substituted for calculated values. The system then computes the difference between the reported and the calculated values and adjusts all related allocated values to preserve the integrity of the regionally accepted market sizes. This is not to say that the system could not calculate independent regional totals e.g. national product revenues.

The computation of the difference can be used to automatically readjust coefficients in the system so that calculated and reports values can be aligned and the forecasts refined. Since there was many coefficients in the system, a careful analysis of correlation of factors is necessary to accurately refine the forecast.

This self-adjusting feature of the system greatly increases the number of computations needed to process the data.

Once all allocation fractions are calculated, the region’s market sizes can be allocated to the associated segments, sub-regions, and providers. If sub-regional reported market sizes have been entered, then the reallocation described under the previous section applies. Each allocation fraction is then multiplied by the appropriate market size for a specific product and time.

A determination of the order and geographic level for each allocation is made. For example, the market can be segment at the regional level, the allocated to the sub-regions, then allocated to the providers. Any order for the three allocations is possible. At a minimum, one scoring algorithm must be used. Depending upon the application and need, one or more algorithms may not be necessary.