Simulated test marketing 1. INTRODUCTION According to the officially definition of American Marketing Association (1987) “marketing research help to identify and define marketing opportunities and problems; to generate, refine and evaluate marketing action, to monitor marketing performance; and to improve understanding of the marketing process”. Marketing research can be based on secondary data (information that already exist having been collected for another purpose) or on primary data (information collected for the specific purpose at hand). Secondary data can be obtained more quickly and at a lower cost than primary data but the data couldn’t be relevant or no so accurate, they could be not current or impartial. Data can be collected with quantitative or qualitative surveys. 2. QUANTITATIVE SURVEYS The term quantitative research covers a different range of techniques that can be used to quantify categories, evaluations, customer’s opinions and attitudes in the market. To develop a quantitative research it is important draw conclusions about a large groups of consumers by studying a small sample of the total consumer population. The sample is the segment of the population selected to represent the population as a whole. If well chosen sample of less than 1 per cent of a population can give good reliability.[i] The quantitative research can be of different types : Product Development (to determine if a new product idea fits a market need) , Pricing Research (to indicate the potential premium price that the product will support), Segmentation Research (to create a map of customer attributes that the marketer can use to divide a large market into identifiable customer cluster), Customer Satisfaction (to gauge how well they are doing in the marketplace), Advertising Research Measures (to monitor the communication impact both of your own advertising and that of a competitor) ; Test marketing (to understand the consumer reaction to new product initiatives and to forecast likely sales volume). 3. TEST MARKETING Test markets are used as a final confirmatory "go/no go" step before any large scale product introduction occurs, such as a regional or national launch. The difference with a product test is that in a market-test all elements of the marketing mix can be involved: packaging, pricing, product, promotion. In test marketing, sampling issues deal more with the number of stores or markets than respondents. It is possible to develop the test in different stores of the same chain, in the same market, or to realize "mini-launches" in limited geography area, or develop an electronic test markets (e.g., BehaviorScan from IRI) that provides store distribution services and media delivery (at the household level). Marketers monitor both retail movement and household purchasing via scanner panels to diagnose product performance (electronic test marketing firms provide household panel data for their designated test markets). The customers that will purchase the product will not have perception that it is a test then the collect data will be reliable. The timing to develop a marketing test depends upon objectives, it could be from 2-3 months to 6-12 months. This typology of quantitative survey is very expensive and it takes a long time, furthermore it can give ideas to competitors (and they could try to confound the test ) and it doesn’t explain the “reason why” of the sales data. 4. SIMULATED TEST MARKETING (STM) Created in the 1960s from the Yankelovich, Skelly and White Inc[ii]. STM has supplanted traditional in-market testing as the predominant method used by packaged goods manufacturers to evaluate new products, line extensions, and other new business opportunities prior to large-scale in-market introduction. According to Joseph Willke President, ACNielsen BASES (2003) it is an alternative to test marketing, which is slower, more expensive, and less secure. Simulated Test Market is a type of laboratory experiment that aims to imitate real life, where respondents are selected, interviewed and then observed making or discussing their purchases. It can bring to get mathematical models used to forecast factors such as awareness, trial, sales volumes, impact on other products etc. 4.1 METODOLOGY Simulated test marketing technology has evolved over time through a combination of methodologies for generating market response and mathematical models that simulate the marketing environment. By the 1970's, half a dozen systems or models were being used by research firms to simulate test markets, for example ASSESSOR, LITMUS, BASES, DESIGNOR, and Simulator ESP. The BaseII uses information from SAMI-Burke’s huge data base of new product introductions to obtain the estimates for the Awareness-Trial-Repeat progression. Key regularities are summarized into “category norms” that estimate the awareness in the target market generated by a given marketing plan, in “concept tables” that convert buying intention data and awareness data from consumers in the target market to an actual trial estimate and in “after-use tables” (from product tests) . The customer data are structured to state buying intentions, to present like/dislike scores, and to perceive price/value are converted into an estimate of the repeat purchase rate using these tables. BASES claims that”90% of their forecasts are within 20% of actual volume and over half are within 10%”. ASSESSOR has been developed by Silk and Urban at Sloan (MIT) and it has been described in detail in 1978 and earlier. ASSESSOR is a simulated test marketing procedure that compress the time and space of a traditional test market into a laboratory situation. Customers from the target population are exposed to advertising for the new brand through the “folder” method. Respondents then “shop” at a mock store set up, where the new brand and its competitors are available. They are given an amount of money to spend on their purchase, after a typical inter-purchase period there is a re-call those buying the new product and they get a chance to repeat purchase. Thus, trial and repeat, and hence long term share, are estimated given 100% awareness and availability. This estimate is adjusted (down) for the planned awareness and availability. ASSESSOR claims that average error is 21.5%; adjusting for actual introduction conditions, the error falls to 11.6%. Over the years, many aspects of the various models have converged. The result of this combination is a reliable and valid methodology for forecasting awareness, penetration, share, and volume for new and repositioned products and services. At a minimum, the sales projection requires two concept statements: one for a product (probably competitive) that is already on the market and whose sales history can serve as a control; the other will be for the test product for which a sales projection is sought. Each concept should be complete with the name, positioning, packaging, features and price. Ideally there will be more than one control product and at least one control will not be in national distribution. This allows both the test and control products to be evaluated in regions or markets in which shoppers are not familiar with either. In seeking congruence between measures, e.g., of trial, it is important that there be no extraneous biases favoring one or another concept. Each must be a fair representation of the competition in the marketplace, and for this reason the impact of a familiar brand must be considered. If product is going to be made available for either sampling or in-home use, it needs to reflect fairly what will be the competitive marketplace, as well as congruence in packaging. Consumer input begins in selected stores (e.g. supermarkets) where introduction of concepts and interviewing occurs in the category section. In this way, all candidates for the study are automatically pre-screened as category shoppers. After verification of category usage, brand share and frequency, shoppers are introduced to the concepts. The presentation may be of finished packages on the shelf, mock-ups, a competitive display board or individual (rotated) concept cards. Respondents then state their future purchase interest in the test product and its key competitor (controls), as well as evaluating critical attributes. This purchase interest data can be buttressed by the use of “chip allocations” to represent projected purchases. The usage of the product by the consumer is essential to incorporate concept fulfillment, additionally data on the purchase cycle, attitudinal, behavioural and demographic information is collected to allow profiles of probable triers and rejecters, in particular if adequate amounts of test product are not available for usage in a trial (typically 200-500 packages). For the home-use phase, single samples of either the test or control product(s) are sent home with shoppers who profess that they will “definitely” or “probably” purchase, based on their evaluation of the concept. Multiple samples including both the test and control products may be sent home, particularly if a third wave of interviews, the “sales wave” is to be conducted. The call-back interview for the home use phase is conducted long enough after placement to allow, on average, half of the product to be consumed. The shoppers will have a diary for use in evaluating the product and when contacted by telephone will report on repeat purchase interest, purchase cycle and diagnostic information. Other data collected will be the amount of product used, family/household member usage, occasions for use, anticipated purchase volume, etc. In addition to the first call back interview, further interviews may subsequently occur in one or more waves, designated as “sales wave interviews.” Requestioning as to purchase quantity and interest may occur. But, more importantly, shoppers are offered the opportunity to purchase products at the regular retail price. The interviewer takes orders and fulfills the order with a letter and monetary compensation; or alternatively may ship the products. In order to evaluate the purchase probability and to predicting likely market response it is a right way using a scale superior to traditional 3-, 5-, or 7-point. This typology of behaviour probability scale was developed by Dr. Thomas Juster and published in different forms during the 1960's. It couples word meanings with probability estimates to enhance serious thinking. After this scale has been employed and validated by marketing research companies such as Yankelovich Partners since the early 1960's in numerous consumer behaviour studies including packaged goods, durables, financial products and services, and other categories. An eleven-point purchase intent scale better predicts real world behaviour, especially for mixed and high-involvement decisions. Yet even this 11-point scale overstates the actual purchasing that takes place. People don't do exactly what they say. The relation between saying and doing is the foundation of simulated test marketing. For one thing, the research environment assumes 100% awareness and 100% distribution—in other words, all are aware of it and able to find it easily—which never happens in the real world. Even taking this into consideration, the overstatement problem remains. Each of the eleven points is coupled with a verbal anchor from: "Certain will buy‑‑99 chances in 100" to "No chance will buy‑‑zero chances in 100.” Usually no more than 75% of the people who claim they definitely will buy actually do so. This figure declines as self-reported purchase probability declines, but the ratio is not constant. This leads to a set of adjustments for each level of self-report, which converts questionnaire ratings into estimates of likely behaviour. These adjustments, as an aside, vary by the consumer's (or industrial buyer's) level of "involvement" in a category. The higher their involvement, the more faith we can have in what people say and the lower the need for overstatement adjustment. Needless to say, by taking purchase probabilities and involvement into account, it's possible to produce a reasonably valid estimate of actual sales (i.e., the percentage of consumers who would buy the product at least once). 4.2. FLAWS OR WEAKNESS There are some significant flaws or weaknesses in the current STM process. The conventional STM is strongly dependent on a marketing plan that is largely driven by mass media. For many manufacturers, mass media is not driving their sales. Rather, in-store factors such as packaging, shelf-configuration, in-store promotions, and the whole range of issues subsumed under “category management” are the relevant drivers. STM’s, whether BASES, LITMUS, or any of the others were designed to model a market structure that does not obtain today. The standards and norms developed for STM’s describe a bygone market. In any event, they have always played a lesser role in the sales projection process than commonly thought by clients. Judgment and other factors were far more central to the projection. Rather than being an asset to legitimately exploit in the new millennium, they are components of a rear view mirror, looking back to an earlier marketing environment. The conventional STM requires independent determination of a lot of measures. These must then be assembled together by addition or multiplication, with each measure being adjusted with normative factors to produce a final number that, hopefully, will reflect the real world at some future time a year or two away. Unfortunately, each measure has its own error which must also be added or multiplied to produce the final error. 4.3 FOCUS ON INTERNET USE Recently some marketers have begun to use interesting new high tech approaches to simulated test market research, such as virtual reality and the Internet.[iii] Virtual reality could be the wave of the future for simulated test-marketing .Some Research Company have develop research tools (Simul Shop, VisionDome) to re-creates shopping situations in which researcher can test consumers’ reactions to such factor as packaging, store layout, product positioning, sounds and videos. This tool has several advantages: it is relatively inexpensive, it has flexibility : it is possible to create a large number of simulated surrounding for several different products with several different forms, sizes, colours , sound. Several people can work together via computer and it is possible to use a single approach to evaluate products and programmes worldwide. However this tool has also limitations infact simulated shopping situation never quite match the real life. "Only research firms with true internet panels maintain detailed demographic (and other) information about their panel members and balance their samples so they match U.S. Census statistics. Without this balance, t here is a very high risk survey results will not measure what they are supposed to and lack study-to-study consistency." According to Greg McMahon, senior vice president Synovate (2003) “the average response rate to a web survey is less than 1%, making it even less likely to get a reliable read on the potential of a new product.” 4.4 THE SIMULATED TEST MARKETS’S FUTURE The development of a one-to-one marketing environment challenges all of today’s research techniques, and Simulated Test Markets (STMs) are no exception. We are observing the steady erosion of the mass-marketing world and the emergence of the one-to-one world. Over the next decade, the marketing research industry as a whole (and STMs in particular) will need to develop tools that lead to better and more detailed understanding of small consumer segments, evaluation of individual components of an initiative (rather than the initiative as a whole), meaningful analysis of business performance over shorter time periods, and insights that are unique to individual business partners. To avoid to become less and less reliable STMs will need to forecast entirely at the individual-level, not just trial or repeat probabilities, to allow for different marketing plans for each individual and to estimate different promotional and advertising elasticities for each person.[iv] STMs will need to forecast at the weekly level to assist in production planning and inventory control. STMs will need to employ much larger sample sizes to provide estimates of targets and key segments. This will allow manufacturers to tailor one-to-one marketing plans. STMs will need to optimize marketing plans, first by estimating the contribution of all marketing elements to total sales, then by examining the ROI of each marketing plan element. 5. CONCLUSION Yet because of their small samples and simulated shopping environment many marketers do no think that simulated test markets are as accurate or reliable as larger , real world tests but according to data the products that have had success in the STM have 80% probability to have success in the market[v]. The question is under discussion infact the leading provider of STM research claim that about 90% of our forecasts are within ±20%, manufacturer clients, however, assert that only 52% of STMs were confirmed by in-market results and a whopping 41% had sales lower than predicted.[vi] Simulated test market are used widely often as a pre-test markets. They overcome some of the disadvantages of standard and controlled test markets : are fast and inexpensive, can be run to assess quickly a new product or its marketing programme. If the pre-test results are strongly positive, the product might be introduced without further testing. If the results are very poor , the product might be dropped or substantially redesigned and retested. The resulting forecast, usually delivered with a range of error of ±20 percent, provides the manufacturer with an estimate of the size of the new business opportunity, which can then be used to evaluate its potential profitability before committing to the investment required for a full-scale market launch.[vii] Word Count : 2812 Bibliography guidelines [i] Armstrong, G. Kotler,P. Saunders. J, Wong .V (2002) Principles of Marketing, (Prentice Hall, Financial Times) [ii] Gian Luca Marzocchi,Elisa Montaguti (2003) Le Ricerche per il lancio di nuovi prodotti (Bologna) [iii] Armstrong, G. Kotler,P. Saunders. J, Wong .V (2002) Principles of Marketing, (Prentice Hall, Financial Times) [iv] Joseph Willke President ACNielsen BASES (2003) “The Coming Obsolescence of Current Models and the Characteristics of Models of the Future” [v] Luca Molteni, Gabriele Troilo, (2003) Marketing research , (Mc Graw Hill) [vi] The Soresen In-Store Sales Forecast , (1999) AppendixII, Portland [vii] Jim Miller Senior Vice President, ACNielsen BASES “Global Research: Evaluating new products globally; Using consumer research to predict success (Part 1) “
PERCHE' QUESTO BLOG / THE REASON WHY OF THIS BLOG
Ho creato questo blog per parlare di sociologia e ricerche di mercato, "fare ricerca sul campo" e condividere opinioni e professionalità.
I have done this blog in order to speak about sociology and market research, to do survey and share opinions and skills about this topic.
I have done this blog in order to speak about sociology and market research, to do survey and share opinions and skills about this topic.