Marketing
3. Empirical generalizations
Marketing
"Very often I get the question:
'What will change in the next 10
years?' But I almost never get
the question, 'What's not going
to change in the next 10 years?'
And I would argue that the
latter question is actually more
important because you can
build a business strategy on
things that are stable over
time."
Jeff Bezos
Content
• What are empirical generalisations?
• Negative binomial distributions and how brands grow
• Double jeopardy
• Duplication of purchase
• Natural monopoly
Blood letting
• Blood letting is the removal of a small amount of blood from a
superficial vein on the body.
• This practice was common in medicine from ancient times
until the mid-19th century.
• Blood letting is based on the ancient medical idea that a
patient's health is related to the balance of bodily fluids.
• According to this theory, blood is one of the body's four basic
juices, and if there is too much blood in the body, it can cause
illness.
• Therefore, blood letting removes excess blood and restores the
balance of the body juices.
Blood letting
• In the 19th century, several studies were conducted to
investigate the effectiveness of blood letting.
• These studies found no evidence that this technique was an
effective treatment for any disease.
• Today, we use veiblood letting only rarely (for example, for
excessive iron in the blood) because it is not only in medicine
that the scientific method has become widespread.
Scientific method
• Evidence-based
• Observation - I observe my environment and the phenomena
that interest me.
• Hypothesis - I infer what might be causing the phenomena.
• Data - I collect data to prove a cause.
• Result - I decide whether my hypothesis is valid.
• So in addition to theories, there is also empirics
• Theory is based on logical reasoning often without evidence,
whereas empirics is based on observation and measurement.
Blood letting example
• I observe the practice of blood
letting.
• I hypothesize that it is a method
that does not benefit the patient.
• I am collecting data on cases
where blood letting has and has
not been used, which I will
compare. The data show that in
the group where blood letting
was used, deaths were
significantly more frequent.
• The result is evidence of the
ineffectiveness of this method.
Chocolate bar example:
• We observe a chocolate brand that
has a significantly higher market
share than its competitors.
• We hypothesize that the brand
succeeds in attracting customers to
repeat purchases.
• The data show that the brand
differs significantly from others in
the number of customers, not in
the amount of repeat purchases.
• As a result, the success of the brand
is driven by customers who have
purchased only once not by those
who repeat purchase.
Why do marketing textbooks remain
theories instead of evidence?
• The Hungarian physician Ignaz
Semmelweis found that the death rate from
puerperal fever is significantly higher in
maternity hospitals staffed by doctors than
in those staffed by midwives.
• After excluding other variables, he
concluded that doctors, who often move
directly from autopsies to deliveries, carry
infectious material on their hands.
• Semmelweis instituted a hand-washing
protocol using a chlorinated lime solution,
which significantly reduced the incidence
of puerperal fever.
Why do marketing textbooks remain
theories instead of evidence?
• Despite the impressive results,
Semmelweis's conclusions were met
with skepticism and resistance from
the medical community.
• His ideas contradicted established
medical views and were largely
rejected.
• Empirical generalizations in
marketing go against established
theories, which is why most authors
ignore them.
"It is not difficult for
people to accept new
ideas but forget old
ones."
J.M.Keynes
Empirical generalisations in general
Empirické – evidence based
Generalizace – applies generally in
most situations
Empirical generalisations in general
• Generalization of gravity:
• Gravity is more or less the same
on planet earth.
• On the moon, however, it is
significantly different, due to
specific conditions.
• So we can generalize the validity
of gravity to most conditions on
earth.
Empirical generalisations in general
• An empirical generalization is like a rule or formula that applies in most situations.
This rule helps us understand how different things or situations are related.
• But beware, these rules may not always and everywhere apply. There are situations
where the rule does not apply, and that is good to know.
• For example, water usually boils at 100 degrees Celsius, but in the mountains, where
the pressure is lower, this rule does not apply.
• So even if managers and marketers are creative, they need to know the rules of the
market they are working in.
• It's like architects who design beautiful buildings. They can be creative, but they have to
respect the laws of physics to make their buildings safe and stable. They can't ignore
the law of gravity or their buildings will fall down.
NBD model of repeat purchases
Customer acquisition and retention
• To understand the importance of
customer base modeling, we must first
understand how a brand's sales grow.
• Essentially, there are two ways to do
this: the first is a new customer and the
second is a repeat purchase from an
existing customer.
Customer acquisition and retention
• Acquisitions
• Acquiring new customers
• Indicator: market penetration
• Retention
• Retention of existing customers
• Indicator: repeat purchase/loyalty
• Under conditions of limited resources, the
marketing manager must choose.
Market penetration
• Market penetration is the percentage of potential
customers who have already purchased a company's
product or service.
• Market penetration is measured as the proportion of
customers who have purchased a product or service
out of all potential customers.
• High market penetration means that a company serves
a large part of the market.
• If the potential market is 100 people and 50 of them
have already purchased the company's product or
service, then the market penetration is 50%. This
means that the company serves 50% of the market.
Repeat pudchase / Loyalty
• If a customer buys a product from a
company once, it is a one-time
purchase.
• If the customer buys the product again,
it is a repeat purchase.
• If the customer buys the company's
products regularly, it is a sign of
loyalty.
How do academics view loyalty?
• Customer loyalty has become a strategic objective to
increase brand value and profitability (Baloglu et al. 2017).
• Increasing loyalty should be critical to becoming one of
the growing brands (Baldinger and Rubinson, 1997).
• A critical issue for a company's continued success is its
ability to retain existing customers and make them loyal
(Dekimpe et al. 1997).
• If research shows a high level of loyal customers, it means
that we do not need to find new ones to replace those
leaving (Sethna and Blythe, 2016, p.16).
Build loylaty!
Sounds like generalization!
Build loylaty!
How about some evidence?
NBD Model
• Andrew Ehrenberg applied NBD to customer behaviour when he was
working on market research and Cadbury's Drinking Chocolate
brand seemed to have an unusually high level of overall sales.
• One idea was that the brand had a large number of intensive
shoppers. That is, those who bought the brand repeatedly during the
period under review.
• However, the empirical data revealed that there was no significant
surplus of intensive shoppers for Cadbury's Drinking Chocolate and
the majority were those who had only purchased the brand once.
• Gradually, other research has also shown this pattern across most
customer bases of brands both large and small.
• Thus, the NBD model is that curve in the other slides in an inverted J
shape that maps the structure of the customer base.
Repeat purchases in the customer base
Source: http://marketing-
bulletin.massey.ac.nz/V19/MB_V19_T2_212_Habel.pdf
Source: https://papers.ssrn.com/sol3/papers.cfm?abstract
_id=3042332
Repeat purchases in the customer base
Source: https://www.marketingscience.info/the-
unbearable-lightness-of-buying-as-told-by-an-old-jar-
of-pesto/
Source: https://www.emerald.com/insight/content/d
oi/10.1108/JBIM-05-2014-
0105/full/html?skipTracking=true
Repeat purchases in the customer base
Source: https://www.slu.cz/file/cul/eed27791-
ea38-466f-8522-8433cc8490bb
Source: https://www.sciencedirect.com/science/article/abs/pi
i/S1441358214000044
Repeat purchases in the customer base
Source: https://www.emerald.com/insight/content/doi/10.1108/JBIM-
05-2014-0105/full/html
How to grow?
How to grow?
We don't see this curve shape in the research. So it is not possible for a brand to grow
solely by increasing repeat purchases at the expense of new customer acquisition.
How to grow?
How to grow?
Source: Anschuetz (2002)
• When comparing larger
and smaller brands, it is
evident that larger brands
have more customers who
have only purchased once.
• But they also have slightly
more repeat customers
(double jeopardy)
How to grow?
Zdroj: Anschuetz (2002)
• We see that the first brand
in the market with 18%
share is more successful
especially in the amount of
light shoppers.
• In essence, this can be
thought of as pulling the
curve upwards beyond the
extreme point at the
beginning of the curve.
Double Jeopardy
Double Jeopardy
• Small brands are hit twice, hence the phenomenon is called
Double Jeopardy.
• Not only do they have fewer customers (the first threat), but
they also attract less loyalty among their buyers (the second
threat) than big brands.
• The average purchase frequency of a brand decreases only
slightly with decreasing market share. In other words, the
number of customers has a greater impact on market share.
Double Jeopardy
Source: Dowling a Uncles (1997)
Double Jeopardy
• Super-loyalty brands: are brands with high market
share and above-average loyalty (repeat purchase
frequency).
• Niche brands: are smaller brands with a small
number of very loyal customers.
• Change-of-pace brands: are brands with a larger than
expected number of customers, but they show less
loyalty.
Double Jeopardy
• Most of the brands are on the DJ line, and the cases
cited of superloyal brands, niche brands or speedchanging
brands are rather the exception that proves
the rule.
• It is true that small brands with low turnover can be
significantly profitable (turnover ≠ profit) as they
choose profit instead of growth (investment in
communication or distribution). But then they cannot
expect to grow market share.
Duplication of purchase
Duplication of purchase
• People buy from a repertoire of brands.
• Purchase duplication describes the overlap of a
brand's customer base with competing brands in line
with its penetration.
• Over the time period under consideration, a brand will
share more of its customer base with large brands and
less with small brands.
• Large brands will inherently be at an advantage.
Duplication of purchase
Source: Snijders (2020, p. 22)
Duplication of purchase
Natural Monopoly
Natural Monopoly
• Not to be confused with economics!
• The effect of Natural Monopoly is that popular products or
brands tend to attract lighter or less knowledgeable buyers of
the product category compared to less popular products and
brands.
• Impacts: advertising for the whole product category?
• Small brands should avoid marketing communications that aim to
promote an entire product category, as they are very likely to attract
customers to other, stronger brands.
Natural Monopoly
Zdroj: Dawes (2020)
Natural Monopoly
Source: https://www.slu.cz/file/cul/eed27791-ea38-466f-8522-8433cc8490bb
The amount of less experienced customers in the customer base. The firms are ranked by market share
and it is therefore apparent that, with exceptions, the natural monopoly rule applies.
Marketing
Thank you for your
attention