The Competitive Implications Of Relevant-Set/ Response .

2y ago
14 Views
3 Downloads
7.12 MB
16 Pages
Last View : 13d ago
Last Download : 2m ago
Upload by : Rafael Ruffin
Transcription

JOHN R. HAUSER and BIRGER WERNERFELT*Consumers choose from a relevant set of brands. Advertising encourages consumers to consider a brand as relevant. Price and other variables influence consumerchoice among relevant brands. The authors examine how the explicit considerationof competitive response affects managerial recommendations and explore the interaction between price and advertising decisions. They consider coses in which managers do and in which they do not anticipate competitive actions.The Competitive Implications of Relevant-Set/Response AnalysisAdvertising response analysis has been a key component of marketing decision support systems for morethan 30 years (Benjamin and Maitland 1958; Benjamin,Jully, and Maitland 196(); Little 1966; Lodish 1986). Senior managers at General Electric (Lillis and Mclvor1985), Kodak and General Motors (Barabba 1985), Carnation {Struse 1985), and Mattel (Hatch 1980) describehow such analyses have improved advertising decisions.Ten years ago. Little (1979a) reported that more than 400 million was being spent annually on data and ex(leriments tor market response analysis. We expect thattigure has more than doubled today with the advent ofautomated data collection by supermarket scanners andsplit-cable television advertising tests. Academic interestis high as a variety of researchers investigate how to estimate response curves accurately (see, among others,Assmus, Farley, and Lehmann 1984; Chakravarti,Mitchell, and Staelin 1981; Little 1975; Mahajan, Mullerand Sharma 1984; Rao and Sabavala 1986).The normal application is to prescribe how a targetfirm should act. The impact of competitive actions isincluded in some models and, in practice, judgmentssometimes are made about how competitors will respond. Econometric models (Lambin 1970, 1976) modelcompetitive response explicitly, but rarely do formal advertising response models attempt to take competitive reaction into account. For example, Hauser and Shugan(1983), Kumar and Sudharshan (1988), and Rao and Sabavala (1986) use one form of response analysis that doesnot model competitive reaction explicitly.We demonstrate how explicit consideration of competitive response can influence managerial recommendations. In some cases, we find recommendations to berobust to competitive considerations. In those cases simpler models suffice. In other cases the recommendationsmust be changed. For example, we illustrate cases inwhich a noncompetitive model says to decrease advertising, but competitive considerations indicate that thenet effect is to increase advertising. In those cases, tomake an informed choice, the decision support user mustbe aware of the impact of competitive considerations.Competitive efforts are complex. Though it is mathematically feasible to analyze these effects in very general models, such analyses obscure intuitive understandings with complicated mathematical conditions. Wetherefore choose a two-stage exposition. We begin witha set of expositional assumptions and present the basicresults. We then relax those assumptions and show how(if) they affect the results. Where they do, we believethe reader can better interpret the complicated results oncethe basic results are understood.In some cases we illustrate the results with a specificfunctional form. In such illustrations we select parameters thai can be justified empirically. These illustrationsgive the reader a feel for the magnitude of the results.However, unless otherwise noted, the results apply forthe more general class of advertising response functions.We begin with a section on expositional assumptions,then present the relevant-set/response model in competitive and in noncompetitive forms. Our first set ofresults pertains to the advertising equilibria and our second set of results pertains to the effect of advertising onprice equilibria. Finally, we illustrate whether or not theexpositional assumptions affect our results.*Jijhn R. Hauser is Kirin Professor of Marketing and Birger Weriicrfclt is Asstxriatc Pnitcssor of Marketing, MIT Management School.391Journal of Marketing ResearchVol. XXVI (November 1989), 391-405

JOURNAL OF MARKETING RESEARCH, NOVEMBER 1989392EXPOSITIONALASSUMPTIONSThe following six assumptions are made in the firsttwo analytic sections of our article. Subsequent sectionsconsider relaxation of the first three.Duopoly Versus OligopolyIn the works we have cited and in most managerialpractice, competitive response is not considered explicitly. With one firm, competitive effects are not relevant;with two firms they are. When we go from two firms tothree firms the equations become exponentially morecomplex—that is, there are 2" relevant sets for an «-firmanalysis. Analytic solutions eventually cease to be feasible. Simulation of any particular market is feasible, butsuch simulations require many additional assumptions andmay or may not yield clear intuitive results. We believethe major insights come when we go from one firm totwo firms. In a subsequent section we return to the casewith more than two firms.It is beyond the scope of our article to consider product lines in which each firm offers a number of brands,including flankers, and sizes. However, we believe muchof the basic insight from single-brand firms will proveto be extendable to complex product lines.One Variable at a TimeWe look first at the competitive equilibria for advertising and then at the competitive equilibria for price.The technical conditions for these equilibria are easy tograsp and the intuition is transparent. In a subsequentsection we discuss the more complicated conditions thatapply when advertising and price are set simultaneouslyand discuss what is necessary when they are set sequentially. (Equations are given in Appendix B.)brandy is considered. This model is in common use, sayby Silk and Urban (1978). However, its use must beconsidered conditional on a market structure definition.For example, consider the coffee market. Probabilisticindependence is unlikely to apply to the entire market.Users of Maxwell House ground caffeinated coffee maybe more likely to consider other ground caffeinated coffees as relevant than they are to consider instant decaffeinated coffees as relevant. However, the assumptionmight apply if we condition the analysis on the submarket of ground caffeinated coffees. Indeed, one techniqueto test market structure uses a related form of choiceindependence, the aggregate constant ratio model, to define substructures within markets (Urban, Johnson, andHauser 1984). Our perspective is that the competitors ofgreatest interest are those in the same submarket; hence,to a first order, we believe probabilistic independence isrelevant for this article. (Most applications of Assessoror Defender are preceded by at least a qualitative concern for market structure.)To illustrate that probabilistic independence applies toat least one submarket. Table I reports data on relevantsets for plastic wraps. The data were available for thefour largest brands. As the number of consumers considering zero brands is unknown, we normalized the data.Table 1 can be considered a chi squared contingency table where four variables (consideration of each brand)can take on one of two states (consider or not consider).The resulting chi squared statistic has 11 degrees of freedom, that is, 16 consideration sets minus four marginalTable 1TESTS OF INDEPENDENCE ASSUMPTION FOR PLASTICWRAPS SUBMARKETResponse IndependenceWe model advertising response as a function of a firm'sown advertising spending. This practice is common asillustrated in most of the references cited before and istypical of many applications such as the Assessor (Silkand Urban 1978) and Defender (Hauser and Gaskin 1984)models. However, competitive advertising, say that ofColgate, will affect consideration of a firm's brand, sayCrest. Our expositional assumption is based on the realization that in most cases a firm's own advertising hasa greater effect on whether consumers consider that fum'sbrand relevant than does competitive advertising.In a subsequent section we consider the case in whicha firm's response function depends on competitive advertising and argue that the main qualitative results arenot changed. A supplemental appendix (available fromthe authors) rederives all results for the more general case.(See also Appendix B.) We do not present it here because the technical conditions obscure the basic intuition.Probabilistic IndependenceWe analyze explicitly the case in which the consideration probability of brand i is independent of 10101010Predicted*" Observednumbernumberofofconsumers 00No dataNo data'1 indicates the brand was considered and 0 indicates it was not.The identity of the major national brands is disguised for confidentiality. Brand D is the store brand. Prediction under the assumption of independence, hor example,the predicted percentage for relevant set I is the pnxiuct of the marginal consideration probabilities for brands A, B. C. and D. The marginal probabilities are determined from the data.

393RELEVANT-SET/RESPONSE ANALYSISprobabilities determined from the data, minus one degree of freedom for the normalization. As shown, thechi squared test does not reject the independence assutnption at the .10 level. (Observed x" - 14.8, x' U 1d.f. I 17.3) A similar test in the automatic dishwashingdetergent category also did not reject the assumption.Focus on Long-Term ResponseAdvertising is a complex phenomenon that affectsconsumer response in myriad ways. Many analyticalmodels have been proposed in marketing (see Blattbergand Jeuland 1981: Bultez and Naert 1975; Kalish 1985;Little 1979b; Lodish 1986; Mahajan, Muller, and Sharma1984; Parsons 1975; Sasieni 1971; Simon 1982; Tengand Thompson 1983) and in economics (Dorfman andSteiner 1954; Gould 1970; Milgrom and Roberts 1986;Nelson 1970; Nerlove and Arrow 1962; Schmalensee1978; Telser 1962; Vidale and Wolfe 1957). These models;irc often dynamic, concerned with the speed of response, decay, or carryover from one period to the next.In our analysis, we are concerned with the long-term,steady-state response—that is, we model what happensif firms hold their spending constant long enough for salesto stabilize, ln such a steady state the effect on sales oftransient temporal phenomena such as carryover, speedof response, and forgetting dampens out and a constantsales level is obtained. For example, when spending isconstant, the previous period's carryover does not vary.See Little (1979b) for discussion and illustrations with avariety of analytical models. For a general treatment, seeFeinberg (1988) and Sasieni (1971).'In this steady state of constant advertising spendingwe examine the relationship, the response function, beiwecn the steady-state sales and the stabilized advertising spending levels. We believe this focus on the longterm response is appropriate for strategic decisions involving competition. Future articles can extend the strategic insights to advertising tactics when markets areconstantly in flux.Decomposition of Advertising EffectsAdvertising has many effects. Two of the most common are to influence consumers to consider brands andto influence consumers to purchase brands from the settif brands being considered. Though Nedungani (1988)lias demonstrated in the laboratory that these effects canbe separated, most real advertising campaigns containelements of both forms of advertising, (lie was able tomodify consideration probabilities without affectingpreference for condiments, fast food outlets, and alcoholmixers.) See also discussion by Ehrenberg (1988), whoiirgues that much advertising is directed at awareness ratherthan ptxsitioning.In our analyses we take the role of an engineer who'In the case of time-varying but periodic strategies sueh as pulsing,we are concerned with the resptinse to average spending rather than(he details of how the budget is allocated over time.separates forces into their components so that the forcescan be understood. Following Hauser and Shugan (1983),Kumar and Sudharshan (1988). and Wemerfelt (1985),we analyze advertising through its effects on (1) consideration sets and (2) choice within consideration sets. Anyreal campaign can be built up from these components.Note that this is a logical distinction for the purpose ofanalysis. It does not require that two separate campaignsbe developed or that the effects be separated. An actualcampaign can exploit synergies between the purposes justas an engineer designs a single truss to support a bridgeagainst vertical forces (gravity) and horizontal forces(wind).For example, such logical decomposition is similar inspirit to that recommended by advertising mcxlels suchas the hierarchy of effects (Lavidge and Steiner I96I)and AIDA (attention, interest, desire, action; Strong1925), which have been used by advertisers and agenciesfor more than 60 years. See discussion by Aaker andMyers (1987, p. 105). It is also similar in spirit to thatused by Arnold et al. (1987) to decompose the overallspending and the "quality" of advertising.Hence, we focus on the component of advertising thataffects consideration sets. We take as exogenous thecomponent of advertising that affects choice within consideration sets. We illustrate how changes in the latteraffect recommended spending on the former.RELEVANT-SET/RESPONSEDUOPOLYMODEL IN AWe now use the assumptions of the preceding sectionto describe the relevant-set/response model. This modelis used widely by major market research firms' and fonnsthe basis of models such as Assessor (Silk and Urban1978) and Defender (Hauser and Gaskin 1984).The Relevant SetThe relevant-set/response model is applied when othervariables, say price, positioning, or product features, areanalyzed separately. A brand is said to be in a consumer's relevant set if it is evaluated seriously. For example.Silk and Urban (1978) operationalize a relevant set asthose brands a consumer has used, has on hand at home,would seriously consider using, or would definitely notuse. It is related to Howard and Sheth's (1969) conceptof an evoked set, but it includes brands consumers haveevaluated and rejected.The relevant set is a much stronger requirement thanawareness. For example, unaided awareness refers to thosebrands a consumer can name without prompting by aninterviewer. Silk and Urban (1978) report that if 95% of Personal communication with Karl Irons of SAMl/Burke. StephenNeedel of A. C. Nielsen & Co. and Steven Gaskin and Steven Cohenof Information Resources, inc. (IRI) indicates a widespread use ofresponse analysis within these, the world's first, third, and fourth largestmarket research rirms. IRl and SAMI/Burke report an active use t)frelevant-set analysis. A. C. Nielsen is concerned mostly with the salesresponse.

394JOURNAL OF MARKETING RESEARCH, NOVEMBER 1989the consumers are aware (unaided) of a brand, then it isin the relevant set of only about 50% of the consumers.At 70% unaided awareness, the relevant-set percentagedrops to 10%. It disappears almost entirely if unaidedawareness is below 60%.There are many potential explanations for the relevant-set phenomenon. For example, we have offered atheory of the consideration-set formation process (Hauser and Wemerfeit 1989). For the purposes of this articlewe need only the empirical facts that (1) consumers consider a number of brands that may be more than onebrand but less than the total number available and (2)the probability of consideration is a function of advertising. See, for example. Brown and Wildt (1987),Campbell (1969). Gr0nhaug (1973/74), Hauser (1978),Hauser, Urban, and Roberts (1983), Jarvis and Wilcox(1973), Silk and Urban (1978), and Urban (1975).Figure 1ILLUSTRATION OF HOW CONSIDERATION ADVERTISINGAFFECTS RELEVANT SETSThe Response ModelResponse analysis is based on the notion that there issome relationship, a response function, that enables amanager to predict steady-state effects as a function ofthe advertising spending by both firms in the market.Sueh response functions are estimated by judgment, experimentation, and/or econometrics. There is much debate on how best to estimate response functions (see areview in Rao and Sabavala 1986), but there is little debate that such functions exist and are relevant.The relevant-set/response model isolates the portionof advertising that affects the probability that a consumerwill consider a brand as relevant. It usually is applied inconjunction with other procedures such as pre-test-market simulation in which other aspects of advertising, saypositioning, are measured exogenously. These aspectsare taken as given and the relevant-set/response modelis used to decide how much a firm wants to invest inorder to get consumers to consider its brand as relevant.Basically, if firm / invests k, dollars per annum onadvertising, that firm's brand will be in the relevant setof AXk,) consumers, where -4,(A,) usually is stated as afraction. (Our expositional assumptions of response independence and probabilistic independence enable us torepresent -4,(A:,) without explicitly noting any functionaldependence on kj or A/kj) for / T j . We consider dependence on kj subsequently.)Suppose firm I achieves Ai as a relevant-set fractionand firm 2 achieves AT. Then, as illustrated in Figure 1,A1A2 of the consumers consider both brands, A ( 1 - A3)consider just brand 1, A2(I - A,) consider just brand 2,and (1 — A])(l — A2) consider neither. In other words,it is as though firm I competes in a product/price duopoly for A1A2 percent of the market and a product/pricemonopoly for A (i — A2) percent of the market. Salesare a combination of these "duopoly" and "monopoly"sales.In particular, if .9 were the sales firm 1 would realizeif both brands were in all consumers' relevant sets andS, were the sales firm I would realize if only brand IBrand 1 'sAdvertisingA,1 -A.Duopoly"Monopoly"forBrand 1"Monopoly"forBrand 2UntappedMarketA21-A2Brand 2's Advertisingwere in the relevant sets, the sales by firm 1 are givenby(1)sales of fimi I A1A2 si Sales of firm 2 are given by a similar equation.The conditional sales, .v, and 5,, are considered to befunctions of other marketing and product design decisions.' Let pi and p be the prices of brands 1 and 2,respectively, let c, and c be the production costs, andlet V and V2 represent other marketing and design variables such as positioning, advertising, promotion, andproduct features. Then 5, is a function ofp, and p2 andof V, and V2, whereas 5, is a function of only / , and v,.Without loss of generality, we define the v) such that .v,and Sj are increasing functions of Vj. As long as the twobrands are substitutable. Sj Sj.Under the stated conditions and definitions, the profit,TTi, of firm 1 is given by(2)IT, -(ptThis, and the equivalent equation for 1T2, is the relevantset/response model. Note that the profit of brand 1 isexplicitly dependent on the price and the relevant-set advertising of brand 2. As discussed in the preceding section, advertising is decomposed into its effect, A:,, onrelevant sets and its effect, v,, on choice within relevantsets. As the purpose of our article is to concentrate onthe implications of the re levant-set/response model, weconsider v, and vs to be set exogenously. Naturally, weconsider what happens to the advertising and price equilibria when V and V2 are changed.Because the products are considered to be differen- One potential generalization would be additive errorswith s, and S,.j , .

395RELEVANT-SET/RESPONSE ANALYSIStiated, 5, and 5, are continuous functions of the prices,(ln an undifferentiated market, the lower priced brandwould capture all sales.)Benchmark 1 NoncompetitiveAnalysisTo appreciate better the interactive nature of equation2, consider a noncompetitive formulation in which theresponse function simply scales .v,, that is,(NC 1)IT]— Ai\k])\pft'i)'SirCi,with a similar equation for brand 2. In this model theadvertising response function tells a firm the fraction ofconsumers who will consider its brand, but the firm doesnot take into account the fact that competitive relevantset advertising will affect sales among those consumers.We do not propose NCI as a model of firm behavior,but we do note that some recent published models (e.g.Hauser and Shugan 1983; Kumar and Sudharshan 1988;Rao and Sabavala 1986) use conditions similar to NCI.For our purposes, NCI serves as a benchmark with whichto compare the implications of the competitive analyses.(One might consider an even more restrictive benchmark: consideration-set advertising has no effect. Thatmodel is the same as NCI with the exception that A (A )is deleted. Naturally, it .would imply that relevant-set advertising has no effect and that the price equilibria wouldbe as in standard economic analysis.)Some Technical AssumptionsTo make equations 2 and NCI empirically relevant,we must restrict the properties of the advertising and priceresponse functions, A('), s(*), and 5(«). These technicalassutnptions are more than expositional; they endow themtxiel with behavior that represents real markets,C mcave advertising response. Recall that we focuson the long-tenn effects of relevant-set advertising. Weassume that such (steady-state) advertising spending increases the relevant-set proportion, but at a decreasingrate. Technically, this means that Aj(kj) is nondecreasingand strictly concave for both firms. Though this assumption may seem restrictive, as Little (1979a) andLodish (1986) repc rt S-shaped sales response functions,profit maximization implies that a firm should either opcrate on the concave portion of the response curve or notadvertise at all. Our assumption implies simply that rational firms operate on the concave portion of the respt)nse curve.Empirically, if an advertising response curve is concave, the advertising elasticity of the effect will be lessthan I.O.* Though econometric studies typically report'\f A is the effect and k is the advertising spending, then for a nondecreasing striclly concave function with a non-negative intercept, itmust be Inie that dA/dk A/k. This relation implies ihat ihc elasticityis less than 1.0. Strict concavity is sufficient but not necessary for anelasticity below 1 because a linear curve, A a,, a,k. has an elaslicity of atk/{ai, a,*) which is less than 1.0 whenever a and a, arepositive. (We thank an anonymous reviewer for this insigbt.)elasticities for sales rather than relevant sets, they do giveus some insight on whether it is reasonable to expectelasticities less than 1.0.In a summary of 128 econometric studies, Assmus,Farley, and Lehmann (1984) reported mean short-termelasticities of .22 with an ANOVA grand mean of .70.When the short-term mean was converted to a long-termelasticity, the result was .42. In a similar study of 37European markets, Lambin (1976, p. 98) rept rted a longterm mean elasticity of .23 with all reported long-termelasticities less than 1,0. Thus, at least for sales elasticities, the data do not reject concavity.Concave conditional profit functions. For the duopolyand monopoly price response we need not assume that.V and 5, {sj and 5 ) are concave. We use only the lessrestrictive assumption that firms consider prices in theranges where the implied duopoly and monopoly functions, iTji (p, — r )a'i, TTnii (/?! - fi)5,, are concave.Such an assumption is reasonable and applies to a widerange of important markets. For example, we have shown(Hauser and Wemerfelt 1988) that the assumption applies to any market described by the Defender price andpositioning model. Technical ConditionsOne of the first steps in any decision support analysisor market simulation is to choose a family of parametricfunctional forms for modeling. For example, an analystmight choose an exponential form, a constant elasticityform, a quadratic form, or even a linear fonn. The analyst then estimates the parameters, runs the mtxiel, andprovides managers with recommendations. There is someconfidence in these predictions because the parametersare based on empirical data.However, the seemingly innocent choice of a functional form may have implications for managerial action.One of the purposes of theoretical analysis is to identifyconditions that are necessary and/or sufficient for certain actions. The analyst then can compare his or hermodeling choices with the general conditions. For example, for symmetric firms, condition Al in AppendixA holds for all exponential response functions, but forconstant elasticity functions there is an empirical constraint related to the elasticity being less than .5,We use two technical conditions in our results (statedin Appendix A for easy reference). Both relate to therelative impact on a firm of its own actions and those ofcompetitors. The conditions require, in essence, that afirm's own actions influence marginal responsiveness morethan do competitive actions. Technical readers will recognize the price condition as the standard economic theory price-equilibrium condition. The advertising condition is the direct analog.'The analytical proof is for uniformly distributed tastes. The Defender price response is quasiconcave. It is not concave for all prices,but the maximum profit always occurs on the concave portion of thecurve.

396JOURNAL OF MARKETING RESEARCH, NOVEMBER 1989COMPETITIVE REACTIONSCommon usage of response models is to estimate empirically the response functions and then choose the advertising (or price) level to maximize profit. This profitmaximizing action is the recommended managerial action.We begin by deriving the optimality conditions for NC I.They are(NC2)dAm)fdk, I(/7, - c,) ,)]"'.Because A (») is a concave function, a smaller slope, dAjdki, implies a larger optimal advertising level, kf. Hence,in this formulation, the larger the net dollar volume (pj— Cj)Sj, the more a firm advertises. If the margins, pj —Cj, are equal, the firm with the largest unadjusted unitsales, Sj, advertises more. Note that these recommendations are independent of the particular functional formor its parameters, as long as it is concave. In fact, thisis just an example of the Dorfman-Steiner (1954) rule.An equilibrium. In a competitive market this traditional analysis may not be appropriate, if firm 1 changesits advertising, firm 2 may respond. The advertising levelthat was optimal before firm 2's response may not beoptimal after firm 2's response.One way to extend the analysis to the competitive formulation, equation 2, is make firm 2's response endogenous and continue the analysis until an equilibrium isreached—that is, until the optimal advertising levels,kf and k*, are such that neither firm has any further incentive to change them. This equilibrium is called a Nashequilibrium.This concept does not require that firms have knowledge of their competitors' advert is ing-response functions, budget levels, or profits. Firms act to maximizetheir own profits. Competitive actions affect these decisions because they affect a firm's own sales responseto price and to advertising. Such responses are readilyobservable.Conjectural variations. As we have described, eachfuTTi reacts to the current spending levels of competitors.This type of reaction is known as zero conjectural variations (ZCV). Econometric analysis (Gollop and Roberts1979; Iwata 1974) suggests that it is a reasonable descriptive model. At minimum, it provides a first step beyond noncompetitive analyses.An alternative to ZCV is for a firm to anticipate competitive reaction and act accordingly. We begin by analyzing the case of ZCV and then, in a following section,consider the implications of one or both f"irms anticipating the other. We show in that section that the advertising game is like a "prisoner's dilemma," that is. firmshave unilateral incentives to increase advertising to levels higher than the levels that would maximize profitswere they to collude.THE ADVERTISINGEQUIUBRIUMExistence and UniquenessBefore we analyze the equilibria implied by the relevant-set/response model, we want to know whether themarket will stabilize or whether it will decay into an advertising war. If the equilibria exist, we want to knowwhether, for given values of Pj and Vj, the final set ofadvertising levels are unique. If the equilibria exist andare unique, independent of initial conditions, we can talkabout the properties of the equilibria and, if necessary,find them numerically.The following formal result addresses these issues.Formal proofs of this and all subsequent results are available from the authors. Sketches of the proofs are givenin Appendix A.Result I. For fixed prices and for fixed exogenous marketing variables, the advertising equilibriumexists and is unique.Does Competitive Analysis Make a Difference?The competitive formulation, equation 2, is clearlydifferent from the noncompetitive formulation, equationNCI, but it is interesting only if it influences managerialrecommendations. The following formal result shows thatthe formulation does make a difference and, furthermore, if one does not consider competitive reaction,managerial recommendations are consistently low. Thatis, a brand manager who uses the noncompelitive formulation will find that the market will consistently drivespending levels up in relation to those anticipated in amarketing plan.Result 2. The competitive formulation implies advertising spending greater than or equal to that implied by the noncompetitive formulation.Empirical evidence for result 2 is difficult to interpretbecause published studies rarely separate adverti.sing intoits components. However, a meta-analysis by Aaker andCarman (1982) provides some evidence based on overalladvertising. They examine 11 field studies

competitive response explicitly, but rarely do formal ad-vertising response models attempt to take competitive re-action into account. For example, Hauser and Shugan (1983), Kumar and Sudharshan (1988), and Rao and Sa-bavala (1986) use one form of response analysis that does *Jijhn R. Hauser is Kirin Professor of Marketing and Birger Wer-

Related Documents:

May 02, 2018 · D. Program Evaluation ͟The organization has provided a description of the framework for how each program will be evaluated. The framework should include all the elements below: ͟The evaluation methods are cost-effective for the organization ͟Quantitative and qualitative data is being collected (at Basics tier, data collection must have begun)

Silat is a combative art of self-defense and survival rooted from Matay archipelago. It was traced at thé early of Langkasuka Kingdom (2nd century CE) till thé reign of Melaka (Malaysia) Sultanate era (13th century). Silat has now evolved to become part of social culture and tradition with thé appearance of a fine physical and spiritual .

On an exceptional basis, Member States may request UNESCO to provide thé candidates with access to thé platform so they can complète thé form by themselves. Thèse requests must be addressed to esd rize unesco. or by 15 A ril 2021 UNESCO will provide thé nomineewith accessto thé platform via their émail address.

̶The leading indicator of employee engagement is based on the quality of the relationship between employee and supervisor Empower your managers! ̶Help them understand the impact on the organization ̶Share important changes, plan options, tasks, and deadlines ̶Provide key messages and talking points ̶Prepare them to answer employee questions

Dr. Sunita Bharatwal** Dr. Pawan Garga*** Abstract Customer satisfaction is derived from thè functionalities and values, a product or Service can provide. The current study aims to segregate thè dimensions of ordine Service quality and gather insights on its impact on web shopping. The trends of purchases have

Chính Văn.- Còn đức Thế tôn thì tuệ giác cực kỳ trong sạch 8: hiện hành bất nhị 9, đạt đến vô tướng 10, đứng vào chỗ đứng của các đức Thế tôn 11, thể hiện tính bình đẳng của các Ngài, đến chỗ không còn chướng ngại 12, giáo pháp không thể khuynh đảo, tâm thức không bị cản trở, cái được

Le genou de Lucy. Odile Jacob. 1999. Coppens Y. Pré-textes. L’homme préhistorique en morceaux. Eds Odile Jacob. 2011. Costentin J., Delaveau P. Café, thé, chocolat, les bons effets sur le cerveau et pour le corps. Editions Odile Jacob. 2010. Crawford M., Marsh D. The driving force : food in human evolution and the future.

Le genou de Lucy. Odile Jacob. 1999. Coppens Y. Pré-textes. L’homme préhistorique en morceaux. Eds Odile Jacob. 2011. Costentin J., Delaveau P. Café, thé, chocolat, les bons effets sur le cerveau et pour le corps. Editions Odile Jacob. 2010. 3 Crawford M., Marsh D. The driving force : food in human evolution and the future.