Pricing and Promotions: The Analytics Opportunity
Pricing and Promotions: The Analytics Opportunity
Pricing and Promotions: The Analytics Opportunity
© Getty Images
June 2021
The scenario is a familiar one to e-commerce level. (For more on dynamic pricing, see “How
retailers: a supplier increases prices on an item, so retailers can drive profitable growth through
a category manager increases the item’s selling dynamic pricing,” on McKinsey.com.)
price. But this effort to make sales of the item
more profitable is promptly undermined by a While most companies consider price sensitivity
well-intentioned marketing manager, who lowers when they make pricing decisions, the scores often
the price of the item by 20 percent as part of a don’t incorporate enough factors and thus aren’t as
promotion. accurate as they could be. The best price-sensitivity
scores are calculated with advanced analytics, using
Such uncoordinated and counterproductive input factors that take customer, competitor, and
decisions happen much more often than most company considerations into account. For instance,
retailers realize, and they are expensive. Many price elasticity is based on different models for
promotions don’t turn a profit at all, or at least they each product category, because customer behavior,
don’t add nearly as much profitable revenue as including purchase frequency and reaction to price
retailers expect. Addressing this conflict can quickly changes, differs for each product. By aggregating
turn into a game of cat and mouse, in which retailers individual input factors for price sensitivity and
find themselves constantly chasing the next issue in promotion affinity, individual scores for each
a highly reactive way. Sometimes they simply avoid product category can be developed. With price
the problem by keeping prices low or making small sensitivity identified for all products, items are then
adjustments across the board, in effect creating a grouped into three buckets based on their scores:
permanent discount on their entire assortment.
— Key value items: top sensitivity. These are
With better analytics, though, e-commerce retailers everyday products, such as common grocery
can create value by intelligently linking pricing and items, whose prices most customers know,
promotions based on optimal price setting and making it easy for them to comparison shop. As
promotion design. We have observed, for example, a result, customers are especially price sensitive
several innovative e-commerce retailers increase when it comes to these products, and their
revenue and profits by three to five percentage prices must be competitive. These products
points using a highly differentiated analytics typically account for 10 to 20 percent of a
process and often achieve improved customer retailer’s sales. Investing in keeping their prices
satisfaction and loyalty as well. low pays off in significantly better customer
price perception, which leads to more frequent
Companies that effectively and profitably link visits and larger baskets.
pricing and promotions through advanced analytics
engage in the following three-step process, which — Foreground items: midlevel sensitivity. These
first determines customer price sensitivity, then are items that have attributes that are more
gauges the likely effectiveness of promotions for important to customers than price and hence do
every product, and, finally, links the two: not require as great a degree of competitiveness
in pricing as key value items. However, their
prices should be competitive enough to avoid
1. Use a wider range of factors to any negative impact on customers’ price
determine price sensitivity perception. Prices for these items can be set
A price-sensitivity score considers the extent to within a range that falls between, for example,
which customers perceive a product’s price and, the highest and lowest competitor prices, or the
as a result, react to price changes. If a product has minimum margin and recommended retail price.
a higher price-sensitivity score, it means that a
customer is less likely to accept a price increase. In — Background articles: low sensitivity. The items in
this case, the price should be kept at a competitive this bucket are either products for which price is
Exhibit 1
AAvariety
variety of
of input factors
factorscan
canbebeused
usedto to
determine eacheach
determine product’s precise
product’s price
precise
sensitivity
price and promotion
sensitivity affinity.affinity.
and promotion
Competitors Customers/
Price elasticity market Customer value
Number of product views (traffic) Transaction elasticity
Direct-search-results share Basket elasticity
Conversion rate (number of product Order elasticity
views vs orders) Difference from list price
Exhibit 2
A price-promotion matrix
A price-promotion matrixidentifies
identifiesthe
theoptimal
optimalbalance
balanceofof
pricing and
pricing promotion
and
for each product
promotion being
for each sold. being sold.
product
# of product categories Everyday value pricing (EDVP):
Share of gross profit Everyday low pricing (EDLP): very limited low-low strategy
promotions Match competitors’ prices and
Low prices, few promotions; match competitors’ promotions on products with high
base price and long-term promotions price-sensitivity score
High
100 EDLP/very limited EDVP+ (low-low
Higher 90 promotions strategy)
score 80 39 18% 14 15%
indicates 70 Cutoff indicative—
.
higher 60 not fixed
Price–
customer
sensitivity 50
sensitivity
score 40
to price
levels for 30
a certain 20
124 55% 51 11%
product 10
EDVP/optimize for value High-low strategy
0
Low 0 10 20 30 40 50 60 70 80 90 100
Low High
Claus Heintzeler is an associate partner in McKinsey’s Berlin office; Mathias Kullmann is a senior partner in the Düsseldorf
office, where Maximilian Totzauer is a specialist; and Karin Lauer is a consultant in the Frankfurt office.
The authors wish to thank Katharina Buhtz, Marcus Keutel, Anna Kloss, Felix Lösch, Krzysztof Szyszkiewicz, and Stephanie
Vehrenkemper for their contributions to this article.