Usporedite metode

	Reference Price Modeling	Brand-Switching Markov Model
Područje	Marketing	Marketing
Obitelj	Regression model	Regression model
Godina nastanka	1995	1992
Tvorac	Gurumurthy Kalyanaram & Russell S. Winer (synthesis); prospect-theory pricing tradition	Stochastic-choice marketing tradition; codified in Lilien, Kotler & Moorthy
Vrsta	Behavioral price-response model with reference dependence	Discrete-time Markov chain of brand purchasing
Temeljni izvor	Kalyanaram, G., & Winer, R. S. (1995). Empirical Generalizations from Reference Price Research. Marketing Science, 14(3 Supplement), G161-G169.	Lilien, G. L., Kotler, P., & Moorthy, K. S. (1992). Marketing Models. Prentice Hall. ISBN: 9780135456415
Drugi nazivi	Reference Price Effects, Sticker-Shock Model, Asymmetric Price Response Model, Prospect-Theoretic Pricing Model	Brand Loyalty Markov Chain, Brand-Switching Matrix Model, Stochastic Brand-Choice Model, Markov Brand-Switching Analysis
Srodne	3	3
Sažetak	Reference price models capture the behavioral reality that consumers judge a price not in absolute terms but relative to an internal benchmark — a reference price they have formed from past prices. When the observed price falls below the reference the shopper perceives a gain; when it rises above, a loss, an unpleasant 'sticker shock.' Drawing on prospect theory, these models enter gains and losses as separate terms and let losses weigh more heavily than equivalent gains, an asymmetry known as loss aversion. Kalyanaram and Winer's 1995 synthesis crystallized three robust empirical generalizations: consumers use reference prices, they form them largely from past prices, and they respond more strongly to losses than to gains. The reference price itself is usually constructed by exponentially smoothing past prices, the same smoothing logic Guadagni and Little used to build loyalty variables, and the gain and loss terms are embedded in a brand-choice logit or demand model estimated on scanner panel data. The result is a richer, behaviorally grounded picture of how price changes move demand than a single symmetric price coefficient allows.	The brand-switching Markov model treats a consumer's sequence of brand purchases as a Markov chain, in which the probability of buying a given brand next depends only on the brand bought last. Its central object is the brand-to-brand transition matrix, whose rows record, for buyers of each brand, the probabilities of staying loyal or switching to each competitor on the next purchase occasion. Estimated from panel purchase histories by simple frequency counts, the matrix can be propagated forward to forecast how shares evolve and solved for its steady-state distribution to predict long-run equilibrium market shares. The diagonal of the matrix measures repeat-purchase loyalty while the off-diagonals measure switching, giving managers a structural picture of competitive churn. The model is the classic stochastic-choice representation of brand dynamics and a conceptual precursor to the loyalty variables used in scanner-panel logit models. It is most useful where purchases are frequent, the brand set is stable, and the first-order memory assumption is approximately satisfied.