Segmentation Assignment Version Two
Supermarket Question
The following data analysis has been completed on a supermarket, with an aim to provide a greater understanding of their customers in order to reorganise marketing efforts with an overall goal of increasing sales. This analysis will focus on segmentation and clustering customers. Ending with marketing suggestions based on the clusters discovered.
Question 2(a)
- The data in this scenario did need to be scaled as the variables were not using a set scale. Annual income is measured in €,000’s and spending score is a calculated number given to consumers based on the amount they’ve spent with the supermarket, with the higher number relating to a higher spend.
Question 3
Question 4
- When looking at distance matrix using a heat-map, we’re looking for light yellow colored blocks on the diagonal line. These light yellow blocks represent groups of customers who are similar to each other. On the heat-map shown we can see several light yellow blocks on the diagonal, telling us that there is evidence of similar customers within this data-set.
Question 5
Silhouette of 200 units in 5 clusters from silhouette.default(x = clusters1, dist = d1) :
Cluster sizes and average silhouette widths:
23 21 79 39 38
0.5065849 0.6199558 0.6140651 0.5094809 0.4623933
Individual silhouette widths:
Min. 1st Qu. Median Mean 3rd Qu. Max.
-0.1401 0.4993 0.5965 0.5531 0.6610 0.7645 When accessing the quality of the clusters we’ve created we look at both the individual silhouette score and the overall average silhouette score.
Cluster 1 Score
- 0.51 - A reasonable cluster structure
Cluster 2 Score
- 0.62 - A reasonable cluster structure
Cluster 3 Score
- 0.61 - A reasonable cluster structure
Cluster 4 Score
- 0.51 - A reasonable cluster structure
Cluster 5 Score
- 0.46 - A weak structure which could be artificial
Overall Average Cluster Score
- 0.55 - A reasonable cluster structure
Segmentation of Supermarket Customers based on Annual Income and Spending Score
Cluster 1
- Have a low annual income and a low spending score.
Cluster 2
- Have a low annual income and a high spending score
Cluster 3
- Have a moderate annual income and a moderate spending score
Cluster 4
- Have a high annual income and a high spending score
Cluster 5
- Have a high annual income a low spending score
Summary of Supermarket Customer Clusters
| cluster | num_customers | avg_annual_income | avg_spending_score | avg_age |
|---|---|---|---|---|
| 1 | 23 | 26.30435 | 20.91304 | 45.21739 |
| 2 | 21 | 25.09524 | 80.04762 | 25.33333 |
| 3 | 79 | 54.41772 | 50.21519 | 42.88608 |
| 4 | 39 | 86.53846 | 82.12821 | 32.69231 |
| 5 | 38 | 87.00000 | 18.63158 | 40.39474 |
Cluster 1
This cluster has 23 customers with an average annual income of €26,304.35. They have a relatively low spending score of 20.9 and tend to be older, with an average age of 45 years old.
Leads us to believe that these customers prioritize saving over spending, being conservative with the budget they have to spend.
Cluster 2
This cluster has 21 customers with a lower average annual income €25,095.24. However, they have an incredibly high spending score of 80.05. With an average age of 25 years old.
This tells us that these customers are a younger audience who prefer to spend their money focusing on the current moment in time, liking to treat themselves.
Cluster 3
This cluster has 79 customers with a moderate average annual income €54,417.72 and spending score 50.22. The average age of customers in this cluster is 43 years old.
This cluster has the largest population of our data inside, it shows the moderate consumer. Customer with the ability and want to balance both spending and saving on a moderate income.
Cluster 4
This cluster has 39 customers with a high average annual income €86,538.46 and a very high spending score 82.13. The average age of customers in this cluster is 33 years old.
These customers represent the consumers with a large amount of disposable income, customers who potentially enjoy the luxurious of life and don’t mind indulging in them.
Cluster 5
This cluster has 38 customers with a high average annual income €87,000.00. However, they have a noticeably lower spending score 18.63. With an average age of 40 years old.
These customers are savers, planning for the future on a high income.
Percentage of Males and Females in Each Cluster
Cluster 1
Female: 14 (60.87%)
Male: 9 (39.13%)
Cluster 2
Female: 12 (57.14%)
Male: 9 (42.86%)
Cluster 3
Female: 47 (59.49%)
Male: 32 (40.51%)
Cluster 4
Female: 21 (53.85%)
Male: 18 (46.15%)
Cluster 5
Female: 18 (47.37%)
Male: 20 (52.63%)
Summary of Clusters (Named Clusters)
Budget Veterans (Cluster 1)
This cluster has 23 customers with an average annual income of €26,304.35 and a relatively low spending score of 20.91. The average age of customers in this cluster is 45 years old. With females making up the majority of this cluster, accounting for 60.87%.
Marketing Actions
- Showcase the budget options of products as well as pushing the supermarket brand alternatives available. Pushing TV and social campaigns showing what can be achieved with the budget products in store, focusing on the savings made at the end of the advertisment.
Trendsetting Spenders (Cluster 2)
This cluster comprises 21 customers with an average annual income of €25,095.24 and a high spending score of 80.05. The average age of customers in this cluster is 25.33 years. With females having the majority, accounting for 57.14% of the cluster.
Marketing Actions
- Ensure that all trending products are held in store initially, then use the online trends to make social media content showcasing the trend using the supermarkets version of the product. Leverage this clusters youth and want to spend to stay on trend and keep the supermarket top of mind for this cluster.
Value Seekers (Cluster 3)
This cluster has 79 customers, with a moderate average annual income of €54,417.72 and a spending score of 50.22. The average age of customers in this cluster is 42 years old. With females having the majority, accounting for 59.49% of the cluster.
Marketing Actions
- Show this customer in store campaigns showcasing the value of mix and match deals. Push the idea of spending a little more than usual to increase the quantity of their purchase leveraging the clusters want for value. In this case value being the experience this cluster still craves, for example dinner mix and match meal deals. Good products individually, but a better “dining experience” mixed together.
Luxurious Elites (Cluster 4)
This cluster has 39 customers with a relatively high average annual income of €86,538.46 and a correspondingly high spending score of 82.13. The average age of customers in this cluster is 33 years old. With females slightly representing the majority with 53.85% of the cluster.
Marketing Actions
- Release a VIP newsletter for high spending customers in this cluster, offering them the most luxurious products in store pre-release. Mix this with private tasting opportunities. Leverage these clusters need to spend items and feel elite.
Future Planners (Cluster 5)
This cluster has 38 customers with a high average annual income of €87,000.00, but a relatively low spending score of 18.63. The average age of customers in this cluster is 40 years old. With men having a slightly larger proportion, representing 52.63% of the cluster.
Marketing Actions
- Focus on pushing bulk purchase savings to these customers. These customers are committed to long term savings, to leverage this push out lines of “Wholesale” type product quantities at lower prices.
Recycling Question
This analysis has been completed under the idea that Carlow County Council are looking to further understand the recycling perceptions of people living in Carlow. In order to do achieve this, this analysis will break the database down into specific clusters in order to further understand what each cluster feels when it comes to recycling. With an overall goal of making marketing suggestions for each cluster, in order to encourage them to recycle more.
Task 3
- The data does not need to be scaled after being put into the sub-data set, as all of the variables have been measured on the same scale. We can see that each answer has been given on a scale of 1-4.
Task 5
Silhouette of 366 units in 3 clusters from silhouette.default(x = kmeans1$cluster, dist = d1) :
Cluster sizes and average silhouette widths:
169 162 35
0.2682631 0.2054575 0.3271905
Individual silhouette widths:
Min. 1st Qu. Median Mean 3rd Qu. Max.
-0.08256 0.14303 0.23963 0.24610 0.35704 0.52470 Cluster 1 has a silhouette width of 0.27, which falls in the range of 0.25-0.5. This suggests that the structure of this cluster is weak and could potentially be artificial.
Cluster 2 has a silhouette width of 0.21, which falls closer to the lower end of the range 0.25-0.5. While it’s not substantially below 0.25, it’s closer to that threshold. This suggests that the clustering structure for Cluster 2 is weak, but it’s not as weak as a cluster that might fall into the range below 0.25.
Cluster 3 has a silhouette width of 0.33, which is within the range of 0.25-0.5 suggesting a weak clustering structure as well.
Based on the overall mean cluster silhouette width of .25, we can tell that analysis has uncovered a weak structure that could potentially be artificial.
Task 6
| Count | Cluster | pos_impact | environ | money | bins | local | avoid_waste |
|---|---|---|---|---|---|---|---|
| 169 | 1 | 3.846154 | 3.674556 | 3.928994 | 3.118343 | 2.976331 | 1.337278 |
| 162 | 2 | 3.586420 | 3.481482 | 3.827161 | 2.617284 | 2.969136 | 3.104938 |
| 35 | 3 | 1.800000 | 1.228571 | 1.657143 | 1.114286 | 2.285714 | 2.657143 |
Task 7
Cluster 1
Primarily middle-aged individuals between 45-54 and 35-44.
Generally positive perception towards recycling, seeing it as beneficial for the environment and cost-effective.
Moderate confusion about different bins, but relatively good awareness of local recycling facilities. Low difficulty in avoiding waste from packaging.
Cluster 2
Mainly individuals in their late 30s to early 50s.
Positive attitude towards recycling, recognizing its environmental and financial benefits.
Some confusion about different bins, but good awareness of local recycling facilities. Moderate difficulty in avoiding waste from packaging.
Cluster 3
Mainly young adults in the 18-24 range and individuals in their mid-30s.
Least positive attitude towards recycling, with lower recognition of its benefits.
Minimal confusion about different bins and moderate awareness of local recycling facilities. Moderate difficulty in avoiding waste from packaging.
Task 8
Cluster 1
Send a flyer to the houses of the people in this cluster.
On the flyer provide clarity on which items go into which bins, as this cluster seems to have confusion on the different bin types.
With this flyer also address the issue regarding the lack of local recycling centre knowledge. On the other side of this flyer provide map of Carlow, pointing out where the main recycling centres are and what each of them do.
Cluster 2
With cluster 2 the main issue seems to be based around people struggling to avoid waste coming into the home.
With this cluster, use a local social and flyer campaign educating the cluster on the use of the newly introduced Re-turn system introduced across the country.
As this system is already being highly advertised across the country, use a more localised strategy. Tell the cluster where specifically these machines are in the major towns across the county, giving tips on the best way to use these machines.
- Example being the best way to store the leftover bottles and cans
In a situation where the county council had additional budget to spend, the main suggestion would be to send a re-usable cotton fruit and vegetable bag to each person in the cluster along with a flyer. On the flyer push the sustainable, economic and environmental benefits of bringing a re-usable bag on their shopping trips to pick loose fruit and vegetables instead of purchasing pre-packaged goods.
Cluster 3
This cluster is the youngest of the three, with this cluster having the least positive recycling outlook. The main goal is increasing the overall perception of recycling in this cluster.
Firstly connect with SETU, as much of this cluster resides within the demographic of students that go to SETU. Run talks and sponsored giveaways and social nights in the college and Carlow town in order to increase awareness when it comes to the recycling pain points this cluster has.
Run a social media campaign based around a competition. Make a landing page with various questions around recycling practices that this cluster may use most. Set a minimum pass % of the quiz, where upon pass the persons social media account is entered into a competition to win a prize.
- Prizes can be recycling based, based around seasonal events or based around college events (i.e. rag week etc)