Discounts
Discounts and promotions are pivotal in steering customers towards stores and have a substantial impact on sales and margins. We have all seen those massive Black Friday (and like in various different parts of the world) sales, where stores are teeming with people. Intuitively, one is convinced that a price reduction has a psychological impact on customers, and helps retailers either get rid of the products on their way to obsolescence or earn brute margins by virtue of increased sales figures. In the following analysis, we try to delve deeper into the impact of discounts not only on sales but also on people belonging to different categories.
Impact of discounts on sales
To study the impact of sales, we analyze the weekend sales of products belonging to different categories over a period of one year. We highlight that no specific brands were analysed. Rather, different brands were bundled into the following four categories.
Bag Snacks
Cold Cereal
Frozen Pizza
Oral Hygiene
Furthermore, the discounts provided by the stores are of three different types:
- Feature discount: product was is in-store circular,
- Display discount: product was a part of an in-store circular, and
- TPR discount: temporary price reduction only (i.e., shelf tag only. Product’s price was reduced but was not on display or in an advertisement)
Total sale, sale in discount period, sale in normal period per category
The figures below show the cumulative sale of the products across multiple categories.
Regardless of how counterintuitive and incongruous it may appear, more units are sold on normal days than on promotion days. In addition, feature and display discounts taken separately sell far more units as compared to TPR discount, which is in line with the anticipated behaviour. Tellingly, the units publicised heavily in in-store circulars and finding their way onto store display boards are likely to sell more than the ones whose price reduction appears in some random shelf of an aisle.
Normalised sales
The above figures can be misleading - for one, they take into account only the absolute sales numbers. The results may be skewed as the number of promotion and non-promotion days is discarded. Moreover, the number of stores a type of promotion was featured in is also cast aside in the cumulative sales figures. In conclusion, it is important to normalise the sales to do way with the aforementioned limitations.
How discount percentages impact sales
How people use discount tickets
In this section we analyze how discounts impact people’s shopping habits, more precisely how people spend them. We look at these habits from 2 different perspective: the status of the person (owner or renter) and the composition of the person’s family. These two categories are defining for this part of the analysis since they offer a way to split our people into groups that might have different habits. Intuitively we expect to see that those people who are renters and those who have families use more discount tickets.
How owners and renters spend their discount tickets
As we see below, the number of renters is much smaller than the number of owners. In order to have a fair comparison we look at the normalised values.
Age Number of owners Number of renters
19-24 13 8
25-34 75 11
35-44 130 15
45-54 194 15
55-64 44 2
65+ 59 2
Looking at the normalised plots from above we see that owners use more discount coupons than renters when they pay for products. The only age category where renters use more is for people between 19-24 years old, which seems fair if you are young, you have a rent and want to save money. This makes us think that most of the owners with age between 25-64 years have families and might have a more regular buying habit than renters do.
What kind of people redeemed the coupons and on what kind of products
We saw in the previous plots that renters are much less likely to use discount coupons than the owners and that owners redeem many more discount tickets, which is not what we expected to see in the first instance. Since owners between 25-64 years redeem many more tickets than the others, in this section we look at what kinds of families these people come from.
Owners families
1 Adult with Kids 25
2 Adults with Kids 146
2 Adults without Kids 206
Single Female 63
Single Male 38
Unknown 37
Above we see some numbers that start to explain the intuition before, that most of the people have families regardless of whether they have kids or not. We see that most of the redeemed tickets come from families, but because their number is dominant we again look at both the casual and the normalized plot. In order to have a better understanding about how people use their tickets, we observe below the number of transactions for every category of products.
There are many categories of products, many of them quite similar, so we grouped the similar ones into higher categories. In the following analysis we chose five basic categories of products that every household has. In the left we have the broader category groupings considered during the analysis and on the right we have the departments of products as they appear in the dataset:
- Groceries - GROCERY, PASTRY, GRO BAKERY, PASTRY, FROZEN GROCERY
- Meat - MEAT-PCKGD, MEAT, MEAT-WHSE, PORK
- Seafood - SEAFOOD-PCKGD, SEAFOOD
- Medicines - DRUG GM, PHARMACY SUPPLY
- Nutrition - NUTRITION
Most popular product categories people use discount tickets on
Select category:
We see that people use most of their discount tickets on groceries. Every person, regardless of family type, has the peak on the transactions that involve groceries.
People that come from families with 1 parent and kids use considerably more discount tickets than people from other types of families, which makes sense - if you are the only parent and you have kids it would be in your interest to use discounts to buy products cheaper.
People from families with two adults with/without kids do not redeem as many tickets as expected. There might be two reasons for this behavior: they either have a more balanced behavior when it comes to shopping, or they shop less, but when they do they redeem tickets or they do not seem interested in using tickets. If we look at the distribution of coupons redeemed we can draw the conclusion the high number of redeemed tickets comes indeed from the high number of people with families and that they do not seem as interested in using tickets as the others.
Women take care of what they eat much more than men do. Women tend to use a comparable amount of discount on groceries, more on seafood and meat, and much less on nutrition products that contain unhealthy products like chips, juice, and sugar based products. These kind of products are avoided by women.