Analysis of Urban's Perceptor Model for designing and positioning frequently purchased consumer products. The presentation talks about the model description. For analysis and explanation of the model, in depth understanding of the original article (Urban, 1975) and MDS is required.
2. Motive
Develop model for designing and positioning of new frequently
purchased consumer products.
Model structure
Trial and repeat process that produces an estimate of long- run
share for a new brand.
Main idea
Model the distance from the ideal brand to brand trial and repeat
purchase estimates for a new product in an existing market.
Outputs
Understanding of perceptions, preference and purchase process
underlying product positioning decision.
Tool for estimating the market share for alternate new brand
designs.
3. Long Run Market Share Model:
m = ts …………. (1)
m= long run market share,
t= fraction of the target market that ever tries the
new brand (0 ≤ t ≤1)
s= share of purchases of new brand among those
who have ever tried the brand (0 ≤ s ≤1)
Now, Urban further developed this ‘t’ or the
ultimate trial fraction on the basis of the
Trial Model
4. Ultimate Trial Model:
t = qwv ………..... (2)
q = ultimate probability of trial given awareness and
availability
w = long run awareness of the new brand
v = long run availability of the new brand (volume
weighted percentage of stores carrying the brand)
The market share of those who have used
the brand is modeled as the equilibrium of a
two- state Markov process
5. Steady- state Share Process:
S = P21/ (1+ P21 – P11) ……….. (3)
for those who have tried the new brand
Pij = Probability of purchase of brand j at next purchase
opportunity when brand i was purchased last, 0 ≤Pij≤ 1.0,
i or j = 1 refers to new brand,
i or j = 2 refers to all other brands.
This Markov formulation assumes that the
frequency of purchase of the new brand is the
same as existing brands.
6. Perception and Preference Mapping
Xby = ∑Aα=1fyarba …………………… (4)
xby, = coordinates of brand b on dimension y for the
perceptual map of those who have not tried our brand,
but who are aware of concept (b = 1, 2, . . . B, where B =
new brand; y = 1,2,. . .Y)
fya = factor score coefficient for dimension y and attribute
scale a (a = 1,2,. . . A)
rba = standardized average rating of brand b on scale a.
Iy = coordinates of average ideal point of
dimension y for map of those who have not
tried our brand but are aware of the concept.
This Iy would be used subsequently
7. Probability of Purchase (Trial)
q= α0 + α1dB2 ……………… (5)
q = as defined in (2)
dB = squared distance from ideal point to the new
2
brand on map for those who are aware but have not
yet used the new brand
α0 , α1 = coefficients to be determined empirically
8. In case of a perceptual space having z
dimensions where the new brand occupies
the position Xz and ideal position is Iz, then:
dB2 = ∑zz=1hz(Xz-Iz)2 …………….. (6)
Hz = importance weights for dimension z
9. Probability of Repeat
P11 = β 0 + β1d2B …………… (7)
pll = probability of repeat purchase if new brand was
purchased last. (0 ≤p11 ≤ 1.0)
d2B = distance squared from ideal point to new brand
after use perceptual map for those who have used the
brand)
β0 , β 1 = coefficients to be determined empirically
The probability of repeat purchase if the new
product is not chosen (p21)is empirically
determined.
10. Sources of New Brand Share
After substituting the trial (5) and repeat (7)
probabilities in (2) and (3) an estimate of the share
of the new brand is obtained by (1).
It is assumed that share will come from the brands
that are perceptually most similar to the new brand
(i.e. nearest to it on the map) and from the brands
that are most widely evoked.
A brand is defined as evoked by a consumer if it has
been used, if it would be considered at the next
purchase, or if it is rejected as a purchase
alternative.
11. Sources of New Brand Share (contd.)
Kb = m [(eb/D2bB) / ∑B=1B-1(eb/D2bB)] …………..
(8)
kb = loss in market share of existing brand b
m = market share of new brand
eb = fraction of people who have brand b in their
evoked set
D2bB = distance squared from brand b to new brand B
in users map
This completes the model description.