Customer-Segmentation

Use Case

Use Case Summary
Objective Statement:
- Get business insight about how many product sold every month.
- Get business insight about how much customer spend their money every month.
- To reduce risk in deciding where, when, how, and to whom a product, service, or brand will be marketed.
- To increase marketing efficiency by directing effort specifically toward the designated segment in a manner consistent with that segment’s characteristics.
Challenges:
- Large size of data, can not maintain by excel spreadsheet.
- Need several coordination from each department.
- Demography data have a lot missing values and typo.
Methodology / Analytic Technique:
- Descriptive analysis
- Graph analysis
- Segment Analysis
Business Benefit:
- Helping Business Development Team to create product differentiation based on the characteristic for each customer.
- Know how to treat customer with specific criteria.
Expected Outcome:
- Know how many product sold every month.
- Know how much customer spend their money every month.
- Customer segmentation analysis.
- Recommendation based on customer segmentation.

Retail is the process of selling consumer goods or services to customers through multiple channels of distribution to earn a profit.- This case has some business question using the data:
How many product sold every month?
How much customer spend their money every month?
How about Customer segmentation analysis?
How about recommendation based on customer segmentation?

Data of Retail Transaction from 01 December 2010 to 09 December 2011
Source Data: Online retail dataset by UCI Machine Learning Library. https://archive.ics.uci.edu/ml/datasets/Online+Retail
The dataset has 8 columns and 541909 rows.
Data Dictionary:
InvoiceNo: Invoice number uniquely assigned to each transaction.
StockCode: Product (item) code.
Description: Product (item) name.
Quantity: The quantities of each product (item) per transaction.
InvoiceDate: The day and time when each transaction was generated.
UnitPrice: Product price per unit in sterling.
CustomerID: Customer number uniquely assigned to each customer.
Country: The name of the country where each customer resides.

There are about 25% of Null CustomerID in the data. We need to remove them as there is no way we can get the number of CustomerID.
There are few records with UnitPrice<0 and Quantity<0. We need to remove them from the analysis. This could represent cancelled or returned orders.
There is more than 90% of 'United Kingdom' customers, therefore we will restrict the data to only United Kingdom customers.

How many product sold every month? Product sold in November has highest quantity that has around 13,41% product sold from all transaction along 1 year. Therefore the business team can increase sales in this month such as promoting new products to customers in this month.
How much customer spend their money every month? Revenue in November has highest amount that has 13,41% revenue from total revenue along 1 year. Therefore the business team can replicate the success of sales strategies in November to be implemented in other months

Recency Frequency Monetary (RFM)
RFM analysis allows you to segment customers by the frequency and value of purchases and identify those customers who spend the most money.
Recency — how long it’s been since a customer bought something from us.
Frequency — how often a customer buys from us.
Monetary value — the total value of purchases a customer has made.

Now we split the metrics into segments using quantiles.
We will assign a score from 1 to 4 to each Recency, Frequency and Monetary respectively.
1 is the highest value, and 4 is the lowest value.
A final RFM score (Overall Value) is calculated simply by combining individual RFM score numbers.

K-Means clustering algorithm is an unsupervised machine learning algorithm that uses multiple iterations to segment the unlabeled data points into K different clusters in a way such that each data point belongs to only a single group that has similar properties.
K-means gives the best result under the following conditions:
Data’s distribution is not skewed.
Data is standardised.
The data is highly skewed, therefore I will perform log transformations to reduce the skewness of each variable and I standardised the data.
Finding the optimal number of clusters

Davies Bouldin Score is a metric for evaluating clustering algorithms.
The smaller Davies Bouldin Score is The more optimal the cluster. K-Means 4 clusters has lowest Davies Bouldin Score than other cluster. Therefore the optimum cluster is 4.

"Cluster 0" has 29% customers. It belongs to the “Loyal Customer" segment as they Haven’t purchased for some time, but used to purchase frequently (F=2) and spent a lot.
"Cluster 1" has 20% customers. It can be interpreted as “Almost Lost". They purchase recently (R=2). However they do not purchase frequently and do not spent a lot.
"Cluster 2“ has 30% customers. It can be interpreted as "Lost Cheap Customers". Their last purchase is long ago (R=4), purchased very few (F=4) and spent little (M=4).
"Cluster 3“has 21% customers. It belongs to the "Best Customers" segment which we saw earlier as they purchase recently (R=1), frequent buyers (F=1), and spent the most (M=1).

Recommendation for “Best Customers" segment: Focus on increasing customer purchases therefore it is necessary to form a cross/Up Selling Strategy.
Recommendation for “Loyal Customers" segment: The business team must optimize the budget campaign and the time campaign for this customer segment in order to maintain their loyalty and increase their value.
Recommendation for “Almost Lost" segment: This customer segment is very at risk for churn, so focus on activating customers and making repurchases by forming a Reactivation Strategy, Retention Strategy.
Recommendation for “Lost Cheap Customers" segment: This customer segment has churned, so the focus of the campaign is to reactivate the customer by forming a Reactivation strategy.