Products & Solutions / Intelligent Clustering

SAS® Intelligent Clustering for Retail

Enable customer-centric merchandising with tailored local market assortments

The SAS Intelligent Clustering for Retail solution identifies clusters of stores with a history of similar customer preferences. The results can then be used with other planning systems so intelligence can be integrated into the planning process and store clusters can be planned together to drive better sales, margins and customer satisfaction.

Benefits

  • Increase sales, margin, inventory productivity.
  • Increase customer satisfaction in each local market.
  • Find hidden business insights to drive greater results.
  • Simulate new assortment clusters to avoid costly mistakes.

Read more

Features

  • Statistically based cluster generation
  • Attribute profiling
  • Return on investment
  • Intelligent service delivery

Read more

" The retail-thought process has irrevocably changed. Retailers are injecting science into a process that used to be heavily driven by art. For local market assortment decisions, science can be critical in providing local customer insights that might otherwise be missed."

— Paula Rosenblum

Managing Partner, RSR Research


Screenshot

Group stores based on historical sales patterns at the product level.



How SAS® Is Different

Only SAS provides:

  • SAS Solutions OnDemand, which require no capital investment in hardware or software because SAS Intelligent Clustering for Retail is independent of a planning system.
  • Statistical measures of all combinations of stores and products to find an optimal number of clusters.
  • An output that includes profiles for each cluster to inform the retailer who the consumers are and what they're buying.

Benefits

  • Increase sales, margin, inventory productivity. When customers find the products they want through more localized assortments, retailers sell more items at the full price and reduce lost sales due to out-of-stock items. Inventory capital is not tied up in unwanted merchandise.
  • Increase customer satisfaction in each local market. Customers find the products that they want to buy, where and when they want to buy them, when assortments take local sales patterns into account.
  • Find hidden business insights to drive greater results. SAS provides the analytical strength using a scalable approach to analyze sales patterns and determine local customer insights. You learn who your customers are by evaluating store attributes of the consumers within the cluster. You learn what they buy from analyzing merchandise attributes of products sold within each cluster.
  • Simulate new assortment clusters to avoid costly mistakes. SAS Intelligent Clustering for Retail allows you to simulate planning and review the percentage of unit improvement and accuracy. You can see the ideal cluster for each assortment prior to execution.

Features

Statistically based cluster generation
  • Groups stores with similar sales patterns at the product level.
  • Provides store clusters that align as closely as possible with the local consumer demand to grow sales and reduce waste.
  • Measures percent of unit improvement and percent of accuracy (based on mixed percent variance cluster to store) as clusters are added.
  • Shows financial opportunity per cluster group to estimate the potential value in assorting to the recommended clusters.
  • Statistically measures all combinations of stores and products to find an optimal number of clusters.
  • Analyzes at the lowest merchandise level of the assortment plan (style/color or SKU) to determine store sales pattern, not the traditional volume-based cluster approach based on sales of class or category.
  • Runs unconstrained, multiple cluster solutions and selects the optimal number of clusters based on calculated unit shift and mix accuracy.
  • Measures quantitative value to show cluster improvement.
  • Tunes statistical model with specific business criteria:
    • Filter on sell-through percent to exclude sales where the product was sold in clearance.
    • Input minimum and maximum number of clusters to define the number of scenarios.
    • Input minimum number of stores per cluster.
Attribute profiling
  • Can incorporate third-party information, such as syndicated demographic data, to add richness to profiles.
  • Creates store profiles using location and geodemographic data, which may include a combination of:
    • Location attributes (mall, high-volume, rural).
    • Census data (median income, education, sex, marital status, ethnicity, age).
    • Nielsen data (brand management, category management).
  • Finds common traits in the cluster region using attributes.
  • Helps buyers understand who buys the assortment based on store cluster profiles.
  • Determines what merchandise attributes of product are sold within the cluster using product profiles:
    • Surfaces key attributes that drive sales (Cluster A sells more wool, V-neck, $50-$75 range).
    • Provides sales-mix percent and units by attribute (fabric – 20 percent; cotton – 30 percent; wool – 5 percent).
Return on investment
  • Measures percent of unit improvement and percent of accuracy.
  • Identifies financial opportunity per cluster group.
  • Estimates the potential value in assorting to the recommended clusters.
Intelligent service delivery
  • Allows retailers to overcome the difficult challenge of limited computing and analytics resources.
  • Requires no capital investment in hardware or software.
  • Is independent of a planning system.
  • Delivers clustering results in Excel format.

Screenshots

Screenshot
Group stores based on historical sales patterns at the product level.

SAS® Intelligent Clustering for Retail is a managed service that provides immediate value with no long-term hardware investment. Clusters can be integrated into third-party merchandise planning tools and processes. The solution groups stores based on historical sales patterns at the product level, revealing differences among local markets and the appropriate product volume and mix. It simulates planning to converge on the ideal cluster for every assortment and profiles consumer and product attribute insight to reveal local product preferences.

View Screenshot

System Requirements

SAS Intelligent Clustering for Retail is offered to retailers as a managed service to speed results. SAS' knowledgeable retail business and technical consultants are able to provide optimal clusters effectively and efficiently for retail organizations, which are then put into action with planning and store systems. This allows retailers to rapidly improve results with current staffing, and eliminates the need for a long-term hardware investment.

Ready to learn more?

Call us at 1-800-727-0025 (US and Canada) or request more information.