Building, measuring and profing from customer loyalty
Watson, Beck, Henderson & Palmater
Conceptual model and hypotheses
Loyalty antecedents
Commitment, trust, satssacton and loyalty incenties haie all been positiely linked to customer
loyalty, but we expect them to haie diferental efects on attudinal and behaiioural loyalty.
Attudinal loyalty results srom positie eialuatons os a seller based on preiious exchangeve
experience. Driiers os loyalty that primarily enhance a customer’s eialuaton os the exchangeve
should haie a strongever efect on attudinal than on behaiioural loyalty.
Behaiioural loyalty results srom situatonal trigevgevers and habit, which may not iniolie a strongev
attudinal component.
Commitment, trust and satssacton all contribute to a customer’s positie experience. Theresore,
they should haie strongever efects on attudinal loyalty than on behaiioural loyalty. Alternatiely,
loyalty incenties are extrinsic, so they migevht operate as repurchase reminders that reduce eforrul
purchase consideratons and encourageve habitual purchasingev or as rewards sor the positie behaiiour
os repurchasingev.
Hypothesis 1: (a) commitment, (b) trust and (c) satssacton haie strongever positie efects on
attudinal (is. behaiioural) loyalty
Hypothesis 2: loyalty incenties haie strongever positie efects on behaiioural (is. attudinal) loyalty.
Hypothesis 3: attudinal loyalty positiely afects behaiioural loyalty.
Loyalty outcomes
Because attudinal loyalty is associated with positie eialuatons which are easy to communicate,
the efect os attudinal loyalty should be strongevest sor WOM. Behaiioural loyalty instead migevht not
include a strongev, accessible attudinal component, which proiides the basis sor WOM. Theresore, the
efect os behaiioural loyalty on WOM should be weaker than that os attudinal loyalty.
Attudinal loyalty tends to be based on consormity and may exist despite situatonal constraints, so
we expect its efect on persormance to be weaker. Instead, behaiioural loyalty, which is based on
readiness to act and is ted directly to purchase, should haie a strongever efect on persormance.
Hypothesis 4: The efect os attudinal (is. behaiioural) loyalty on WOM is gevreater.
Hypothesis 5: The efect os behaiioural (is. attudinal) loyalty on persormance is gevreater.
Role of measurements compositon and study-based characteristcs
Neither higevh attudinal nor behaiioural loyalty are sufcient to capture customer loyalty sully.
Customers with higevh attudinal loyalty may lack the ability or opportunity to support the seller.
Behaiioural loyalty directly increases seller reienues, but these benefts may be short is customers
lack the motiaton to contnue their purchase behaiiours when their eniironment changeves. Thus,
customer loyalty should capture both behaiioural and attudinal aspects.
Measures os loyalty that include WOM items migevht weaken the predictie efect os loyalty on
persormance outcomes. Theresore, measures os loyalty that include WOM items may capture
unrelated efects, so they will be less efectie at predictngev oierall persormance.
Watson, Beck, Henderson & Palmater
Conceptual model and hypotheses
Loyalty antecedents
Commitment, trust, satssacton and loyalty incenties haie all been positiely linked to customer
loyalty, but we expect them to haie diferental efects on attudinal and behaiioural loyalty.
Attudinal loyalty results srom positie eialuatons os a seller based on preiious exchangeve
experience. Driiers os loyalty that primarily enhance a customer’s eialuaton os the exchangeve
should haie a strongever efect on attudinal than on behaiioural loyalty.
Behaiioural loyalty results srom situatonal trigevgevers and habit, which may not iniolie a strongev
attudinal component.
Commitment, trust and satssacton all contribute to a customer’s positie experience. Theresore,
they should haie strongever efects on attudinal loyalty than on behaiioural loyalty. Alternatiely,
loyalty incenties are extrinsic, so they migevht operate as repurchase reminders that reduce eforrul
purchase consideratons and encourageve habitual purchasingev or as rewards sor the positie behaiiour
os repurchasingev.
Hypothesis 1: (a) commitment, (b) trust and (c) satssacton haie strongever positie efects on
attudinal (is. behaiioural) loyalty
Hypothesis 2: loyalty incenties haie strongever positie efects on behaiioural (is. attudinal) loyalty.
Hypothesis 3: attudinal loyalty positiely afects behaiioural loyalty.
Loyalty outcomes
Because attudinal loyalty is associated with positie eialuatons which are easy to communicate,
the efect os attudinal loyalty should be strongevest sor WOM. Behaiioural loyalty instead migevht not
include a strongev, accessible attudinal component, which proiides the basis sor WOM. Theresore, the
efect os behaiioural loyalty on WOM should be weaker than that os attudinal loyalty.
Attudinal loyalty tends to be based on consormity and may exist despite situatonal constraints, so
we expect its efect on persormance to be weaker. Instead, behaiioural loyalty, which is based on
readiness to act and is ted directly to purchase, should haie a strongever efect on persormance.
Hypothesis 4: The efect os attudinal (is. behaiioural) loyalty on WOM is gevreater.
Hypothesis 5: The efect os behaiioural (is. attudinal) loyalty on persormance is gevreater.
Role of measurements compositon and study-based characteristcs
Neither higevh attudinal nor behaiioural loyalty are sufcient to capture customer loyalty sully.
Customers with higevh attudinal loyalty may lack the ability or opportunity to support the seller.
Behaiioural loyalty directly increases seller reienues, but these benefts may be short is customers
lack the motiaton to contnue their purchase behaiiours when their eniironment changeves. Thus,
customer loyalty should capture both behaiioural and attudinal aspects.
Measures os loyalty that include WOM items migevht weaken the predictie efect os loyalty on
persormance outcomes. Theresore, measures os loyalty that include WOM items may capture
unrelated efects, so they will be less efectie at predictngev oierall persormance.
