This action might not be possible to undo. Are you sure you want to continue?

From Wikipedia, the free encyclopedia Jump to: navigation, search The Fourt–Woodlock equation (sometimes misspelled Fort-Woodlock equation) is a market research tool to describe the total volume of consumer product purchases per year based on households which initially make trial purchases of the product and those households which make a repeat purchase within the first year. Since it includes the effects of initial trial and repeat rates, the equation is useful in new product development. The Fourt–Woodlock equation itself is

The left-hand-side of the equation is the volume of purchases per unit time (usually taken to be one year). On the right-hand-side, the first parentheses describes trial volume, and the second describes repeat volume. HH is the total number of households in the geographic area of projection, and TR ("trial rate") is the percentage of those households which will purchase the product for the first time in a given time period. TU ("trial units") is the number of units purchased on this first purchase occasion. MR is "measured repeat," or the percentage of those who tried the product who will purchase it at least one more time within the first year of the product's launch. RR is the repeats per repeater: the number of repeat purchases within that same year. RU is the number of repeat units purchased on each repeat event. The applied science of product forecasting is used to estimate each term on the right-handside of this equation. Estimating the trial rate is complex and typically requires sophisticated models to predict, while the number of households is usually well known (except in some unusually complicated markets such as China).

References

•

Fourt L.A., Woodlock J.W., 1960. 'Early prediction of market success for new grocery products.' Journal of Marketing 25: 31–38.

**Bass diffusion model
**

From Wikipedia, the free encyclopedia Jump to: navigation, search

a highly influential work that described the different stages of product adoption. The model presents a rationale of how current adopters and potential adopters of a new product interact. It was subsequently reprinted in the December 2004 issue of Management Science.[3] Model formulation [1] Where: • • is the rate of change of the installed base fraction is the installed base fraction . especially new products' sales forecasting and technology forecasting. and the only marketing paper in the list.[1] Prior to this.[3] It was ranked number five. Mathematically. The Bass model has been widely used in forecasting. The model is one of the most cited empirical generalizations in marketing.The Bass Model or Bass Diffusion Model was developed by Frank Bass and it consists of a simple differential equation that describes the process of how new products get adopted in a population. the basic Bass diffusion is a Riccati equation with constant coefficients. Everett Rogers published Diffusion of Innovations. up to May 2013 the paper with title "A New Product Growth for Model Consumer Durables" published in Management Science has 4639 citations in Google Scholar.[2] This model has been widely influential in marketing and management science. Bass contributed some mathematical ideas to the concept. The basic premise of the model is that adopters can be classified as innovators or as imitators and the speed and timing of adoption depends on their degree of innovativeness and the degree of imitation among adopters. In 2004 it was selected as one of the ten most frequently cited papers in the 50-year history of Management Science. Frank Bass published his paper "A new product growth model for consumer durables" in 1969.

Typical values of p and q when time t is measured in years:[4] • • The average value of p has been found to be 0. adoption) ultimate market potential : multiplied by the [1] The time of peak sales [1] Explanation The coefficient p is called the coefficient of innovation. and is often less than 0. internal influence or word-ofmouth effect. with a typical range between 0. external influence or advertising effect.03.5 .e. The coefficient q is called the coefficient of imitation.38.01 The average value of q has been found to be 0.• • is the coefficient of innovation is the coefficient of imitation Sales is the rate of change of installed base (i.3 and 0.

g.Extensions to the model Generalised Bass model (with pricing) Bass found that his model fit the data for almost all product introductions. Although many extensions of the model have been proposed. but that the shape of the curve is always similar. despite a wide range of managerial decision variables.[3] This model was developed in 1994 by Frank Bass. Trichy Krishnan and Dipak Jain: where is a function of percentage change in price and other variables . only one of these reduces to the Bass model under ordinary circumstances. pricing and advertising. This means that decision variables can shift the Bass curve in time. e.

. The Bass model is a special case of the Gamma/shifted Gompertz distribution (G/SG). Norton and Bass extended the model in 1987 for sales of products with continuous repeat purchasing. when p=0. The second special case reduces to the logistic distribution.Successive generations Technology products succeed one another in generations. The Bass diffusion model is used to estimate the size and growth rate of these social networks. Use in online social networks The rapid. • • The first special case occurs when q=0. The formulation for three generations is as follows:[3] where • • • • is the incremental number of ultimate adopters of the ith generation product is the average (continuous) repeat buying rate among adopters of the ith generation product is the time since the introduction of the ith generation product • It has been found that the p and q terms are generally the same between successive generations. recent (as of early 2007) growth in online social networks (and other virtual communities) has led to an increased use of the Bass diffusion model. when the model reduces to the Exponential distribution. Relationship with other s-curves There are two special cases of the Bass diffusion model.

For the following case study. Controlling has shown significant.On Forecasting New Product Sales. The Dirichlet Model of Buying Behavior links market penetration to market share and can thus be used to estimate the impact of marketing activities. . While that product advantage is quite useful to determine what might be seen as a fair price. Regarding the development of market share over time. but in the end. let us assume we are in an innovation-driven industry. July 2013 Forecasting sales for new products is among the most difficult tasks in planning. the FourtWoodlock model differentiates between product trials and repeat purchases. somehow implicitly. It assumes market shares to be constant over the relevant period of time. some say the do not forecast. there is not even a finished product to present to customers and get their feedback. There is nothing to extrapolate. At the same time. The current sales forecasts for product innovations have been done based on model assumptions and market research. sales forecasts for product innovations are inevitable. For development products that are advanced enough to be presented to potential customers. and in the early stages. who doesn’t calculate sales forecasts explicitly will do it somewhere. Experience. unexplained discrepancies between planning and actual sales developments. often in a less thoughtful and therefore less careful way. its connection to the achievable market share is obvious as a fact but not trivial in numbers. Some innovators say they cannot forecast. which regularly develops new products. but appears sufficiently tested to be considered valid at least for the peak market share to be reached by a product. whereas the Bass diffusion model describes the early phases of a product lifecycle in more mathematical terms. There is a variety of methods available to forecast product innovation sales. Artificial Intelligence and Statistics Posted on 18. conjoint analysis offers a tool to gauge perceived product advantage.

Aligned. should have a good chance of being true for for future new products. Unfortunately. For high repeat purchase rates and a vanishing share of imitators. both models are more and more similar. therefore. the market share curves from past data could. for example.claiming to model the share of innovators and imitators among customers. we will do just that. The problem becomes apparent in this quote attributed to mathematician John von Neumann: “With four parameters I can fit an elephant. The company will have detailed data about its own product launches in the target market and in similar markets. That is relatively close to what an experienced individual expert asked for a judgement would probably do. taken from textbooks or guessed. they still contain quite a few free parameters. these parameters have to be derived from market research. A totally different approach from these models would be to go by experience. Individuals tend to introduce various kinds of bias in their judgements. and the trick will be to find the proper elephant. reliable answers. with or without realizing it. this approach can be thwarted by a . The problem with all of these models is that in spite of all their assumptions.” In fact. What has been true for previous product launches. Obviously. and with five I can make him wiggle his trunk. a realistic forecast will have to be based on more than the experience of a few products and. On the other hand. by one’s own company or by others. while having personal experience with the introduction of new products will help in managing such a process. which makes personal experience invaluable for asking the right questions but highly problematic for getting unbiased. In practical use. look like this: The simplest way to estimate new product sales would be to find a suitable model product in past data and assume the sales of the new product will develop in roughly the same way. individual managers. there is usually plenty of “quantitative experience” available. scaled to peak values and visualized in a forecast tool. and market research should be able to provide market volumes and market shares for competitors’ past innovations. introducing a significant degree of arbitrariness into the forecasts.

relatively simple models will be able to describe the data with sufficient accuracy – if they use the right parameters. and they define a well-founded set of basic shapes. at best. all published sales forecast models use market research data from actual products to verify their validity and to tune their parameters.or sshaped. which sales of new products generally follow. which are more difficult to forecast. Try explaining to your top-level management that you have reached a conclusion with a tool without knowing how the tool came to that conclusion.variety of factors influencing the success of new products: The varying influence of all these factors will make it. we need a more complex system. Based on an analysis of the available data. there is no need to model a new product exclusively from past data without any further assumptions. The uptake curve to the peak value will be either r. In this case. Besides. On the other hand. and in fact. we have selected the following set of parameters to structure the forecast: • • Peak market share Time from product launch to peak market share . we are simply using model parameters to structure the information we will derive from recent market data from our own markets. difficult to find and verify a suitable model in past data. judging from one model product yields no error bars or other indication of the forecast’s trustworthiness. AI tools like a neural network could be used for such a task: It could be trained to link resulting sales or market share curves to a set of input parameters specifying the mentioned influences. Apparently. The question is to what extent that historical data actually relates to the products and markets we want to forecast. after which they will either decrease or gradually level off. and can usually be well fitted by adjusting the parameters of the Bass model. to get a reasonable forecast. Often. The disadvantage of a neural network in this context is that the way it reaches a certain conclusion remains largely intransparent. depending on market characteristics. All the more analytical models and forecast tools cited above have their justification. The term “learn” indicates artificial intelligence. which should be able to learn from all the experience stored in past data. The development after the peak is usually of lesser importance and depends strongly on future competitor innovations. Generally. which will not help the acceptance of the forecast. New product market shares will usually rise to a certain peak value in a certain time.

everything else being equal. The more market research data is available. In that case. we approximate it as a function of product profile. As the screenshot from the case study tool shows. either as simple numbers by scoring or by their similarity to the new product to be forecasted. mostly linear. forecast parameters and forecasted market share development: If the parameters were discrete numbers. On the other hand. marketing effort and the other influencing factors simultaneously.• • • Bass model innovation parameter p Bass model imitation parameter q Post-peak change rate per time period The list may look slightly different depending on the market looked at. the parameters are continuous functions of all the influencing factors. which was an interesting research topic in the 1990s. even if research data is missing in certain dimensions (right . we have to take into account the influencing factors displayed in the graph above. linear dependencies should be sufficient. order to market. the more detailed the functions can be. For example. which we approximate using simple. Tuning these parameters to the full set of the available data would lead to the average product. this graph would describe a Bayesian network. These influencing factors can be quantified. however. This leads us to the following structure of influence factors. For most parameters. In our case. dependencies. the multidimensional field leads to reasonable results. These approximations are done jointly in a multidimensional numerical optimization. forecasting could take the form of a probabilistic expert system like SPIRIT. however. rather than calculating peak market share as a function of product profile scoring.

hand side graph). In addition. Implemented in a planning tool. confidence intervals can be derived in the fitting process. quantitative market share forecast implemented in an interactive model that can be used for all new products in the markets analyzed. forecasts from that model could look as follows: . leading to a wellfounded.

depending on the confidence decision makers in the company have in different theories. it can also be combined with more theory-based approaches.Forecast numbers will depend on the values of the different influence factors selected for the respective product innovation. Holm Gero Hümmler Uncertainty Managers Consulting GmbH . Besides this pure form. marketed products. This approach presented can be implemented for a multitude of different markets and products. While using known and well-researched models to structure the problem. Dr. provided there is sufficient market research data available. the actual information used to put numbers in the forecast stems entirely from from sales data on actual.

Sign up to vote on this title

UsefulNot usefulnew product

new product

- Pop REPORT.docx
- AJCEBVol4No2Wong
- Unit - I - Production_Mgt Introduction
- This Article Was Downloaded by
- Opportunity in the Indian Pick and Carry Cranes Market_Feedback OTS_2012
- 65667_1970-1974
- Henry Car Rental_Q3
- Production Control
- AIRPORT FORECASTING
- 7317173 Demand Forecasting Lecture
- Demand Forecasting Visit Us @ Management.umakant.info
- Jbfa Paper
- Aula1forecast
- Westpack JUN 21 Weekly Commentary
- Forecast Value Added Analysis
- Forecast
- MI0031- Technology Management - FALL 2010 - Set 2
- Modelling and Forecasting of Hydrological Variables Using Artificial Neural Networks the Kafue River Subbasin
- Sol Forecast
- Golderr and Tellis 1993
- SESSION_2_EXERCISES
- Exercise ToPrint on Forecasting
- Draft Transmission Planning Guideline Mar 2012
- Module 2.pdf
- Report
- 5, 6 & 7 - Demand Management
- Europe Sugarcane Harvester Market Report 2016
- Full Text 01
- FORECASTING RUNOFF TRIANGLES
- Smith Kilburn Forecasting Eruptions JVGR 2010
- Modeli New Product