Product Clustering
The Product Clustering model is an unsupervised learning model that groups
customers based on the type of products they buy or do not buy. In other words,
this model groups customers based on their buying behavior of specific products
or categories.
The model uses the k-means algorithm and iteratively assign customers to the
nearest cluster based on their Euclidean distance from the centroid of the
clusters.
Use cases
The following use cases describe how Acquia product-based clusters help you
produce relevant and personalized touches, thereby increasing customer
engagement and reducing marketing costs:
More targeted product or brand campaigns
- When creating a product-focused campaign, quickly identify and target
the customers or clusters that have previously purchased these products.
Such campaigns help to increase affinity toward the products that customers
already like.
- When launching a new product, identify a current similar product,
locate the customer cluster aligned with the existing product, and target
customers in those clusters in the launch campaign. This ensures a
high response rate because you target customers with similar product
affinities.
Broaden customers’ purchased product mix
- Improve retention, loyalty, and lifetime spend by encouraging customers to buy
different products.
- Create tests providing higher discounts for product X (or similar
products) to customers in clusters that are either negatively indexed towards
product X or do not feature product X at all. When you convert such
customers to purchase product X, you improve retention and gain incremental
revenue without cannibalizing margin. The benefits are large because these
customers, in all likelihood, would not have purchased these products prior
to your tests.
Use product clusters to identify single-product buyers
Use Metrics to identify the most common one-time purchased product. Using this
information, you can create campaigns to encourage single-product buyers
toward buying other different products, thereby improving the retention and
lifetime spend.
Target audience
The Product Clustering model predicts the behavior of buyers.
- A buyer is a customer with a transaction in the past three years.
- The model is trained on the customers with two+ transactions in the time
window.
- The customers with two+ transactions in the time window are scored by being
assigned to their top three clusters and probabilities.
- The customers with one transaction are labeled as
One-time buyer
.
- The customers with no transactions are labeled as
Non buyer
.
How does the model work?
For each customer, the model:
- Considers the number of transactions (not total revenue or number of
transaction items) in a timeframe per product category at a given level.
- Applies balancing to product categories based on:
- Importance of the product category. For example, customers might frequently
buy “Carry Bag”, but it may not be important to the client.
- Popularity of the product category. For example, customers might frequently
buy chewing gums at checkout. However, it can be a known fact to the client
and may not be deemed necessary or useful to surface in product cluster.
The output for each scored customer consists of the product cluster id, such as
Menswear, Womenswear, Accessories, that the customer is most closely associated
with.
Using the Product Clustering model
CDP displays the output from the Product Clustering model in:
- 360 > 360 Profiles
- Actions > Campaigns and Actions > Campaigns+
- Analytics > Metrics
To learn more about how the Product Clustering model can enhance
your marketing workflow, contact us.