The Individualized Customer Offer Recommendation Engine helps you to make individualized product offers to your customers. 

In this era of « mass information », in which the customer's attention is harder to get, targeting your customer with the most suitable offer is a key success factor for any business (big or small) and differentiates you from competitors in the customer’s eye.

If you can identify this "next best offer" to make to the customer before your competitors, you will gain a competitive edge and increase your sales volume.

It is all about knowing your customer and acting on this knowledge. « » understood it years ago and improved their customer’s online experience by showing « you might also like this… » kind of specific suggestions resulting in increased sales. « Colruyt » in Belgium are also « targeting » or adapting their promotional offers (by email or mail) to each customer’s purchasing habits.  

You need to understand your customers’ purchasing habits & product preferences in order to make the right offer at the right time and increase the return of your marketing campaigns

ICORE's Visualization of Product Associations

ICORE's visualization of the associations between products such as analyzed from the Customers transaction database. 
This shows the relationship between products in a "network" like fashion where products bought together are close to one another. The size of each circle gives an idea of the penetration rate of the product within the customer population analyzed (% of customers who bought the product)
On hovering above the circle; the penetration rate appears (see example of 71% for the product "i-s"). The figure on the arrow is another parameter to illustrate the strength of the relationship (lift) between two products.

ICORE has been designed to specifically suit medium and smaller sized companies' needs. The tool is easy and intuitive to use via three interactive visualization interfaces (ICORE DASHBOARD, ICORE SEGMENTATION & ICORE SELECTIONS; see description hereunder) and does not require "data science" knowledge. ICORE has been developped in R and the visualization interfaces  have been developped in "Shiny R". The installation of these visualization tools is also easy and straightforward even on a stand alone PC at your office. Version française: Les 3 interfaces (ICORE DASHBOARD, ICORE SEGMENTATION & ICORE SELECTIONS) sont disponibles en langue française.

How can ICORE help you to make the right product offer at the right time to each of your customers ?

The answer lies within YOUR customer’s purchases database. You have the « raw material », you now need to exploit it and "let your data speak" . That is why ICORE has been developped: To analyze and extract useful and practical customer knowledge from your customer purchases database.

ICORE uses your customer transactions information (Date of transaction, Customer ID, Product ID, quantity and value of the transaction)  & gives you a full overview of each customer’s individual buying habits along with individual product offer recommendation(s). 

ICORE will deliver:

· A segmentation of your customer base into distinct groups (segments) based on products bought. This is a "product preferences" based segmentation where every single customer is assigned to a group of other customers buying similar products. An exhaustive comparison of each of the segment with the others is available to help you to understand its distinct characteristics.

· An individual customer report showing  :

    • A description of the segment the customer belongs to.
    • Individualized recommendations on the best  product offer(s) to be made to each individual customer.
    • Purchasing habits (recency of latest purchase, frequency of purchases, monetary value and number of products).

· An analysis of distinct product associations patterns (baskets) within your customer base. This helps you understand how products relate to one another in customer's mind. You probably already know some obvious associations between products but the tool will screen all potential product associations (product baskets) and select the most significant ones. You can then analyze them and act on these newly discovered product associations in new marketing initiatives ("bundled product offers", or rearrange the assortment in your catalog or in your shop).

The benefits of using ICORE:  

1) Individualize your message to your customer: The clear customer view given by the ICORE DASHBOARD will guide you, before or during customer contact, to offer the right product(s), whether it be a new product for the customer (Cross-sell) or a product he has already bought before (Up-sell). Take up rates of up to 7% have been consistently observed on Cross-sell offers on inbound customer calls, this goes up to 30% observed take up rates for Up-sell offers.

2) An efficient way to initiate a segmented approach to your mass marketing campaign (email, mail, catalog) based on “customer’s product preferences "  via the ICORE SEGMENTATION.

3)  Increase take up rate to your marketing campaigns by providing better selection of customer targets for (mail, email, phone campaigns or catalog cover individualization) in order to promote a specific product via ICORE SELECTIONS. 

Description of the 3 Visualization interfaces :  

1) The ICORE DASHBOARD is an interactive HTML interface that can be used before or during a contact with a customer  (visit to or from a customer, inbound or outbound phone call), to have a clear and complete view of the customer's purchases and receive a recommendation on which offer(s) to make. You type in a customer id (in the upper left panel of the page) and see details about the customer's profile (his segment number with a brief description). 

The section « GENERAL CUSTOMER OVERVIEW » gives key measures about the chosen customer’s purchasing behaviour (over the timespan of the data provided to ICORE): recency(number of days since latest purchase), total sales, frequency of purchases, number of distinct products bought. If product categories were provided (in this example two categories had been created "banque" & "assurance") ICORE would also show the total number of distinct product categories in which the customer bought a product.

The Sub-Segment number shown is the number of the smaller more "precise" Sub-Segment of the customer. Each Sub-Segment being fully part of one specific Segment. Analysis and description of each Sub-Segment is available in ICORE SEGMENTATION.

The next section of the dashboard shows the customer’s « FREQUENTLY BOUGHT ITEMS (up-sell) », these are potential Up-sell opportunities to be explored by your sales team. 

They can also propose product(s) which are new to the customer using the « RECOMMENDED OFFERS (x-sell) » section. These are « individualized » offers recommended by ICORE for the selected customer. As you can see from the "Offer Type" column ICORE provides 2 types of recommended offers for Cross-selling:

1.« INDIVIDUAL GAP OFFER » : Based on product association rules (baskets) 
2.« POPULAR IN THE SEGMENT»: Based on the top 5 most popular products in the segment.


2) The ICORE SEGMENTATION interface lets you explore each segment individually, Once you understand the characteristics of the segment you can initiate Marketing campaigns specifically targeted towards this segment (also called "cluster" in technical jargon). You type in a segment number;  a brief « Segment description (label) » appears along with a complete overview of the segment characteristics. 

«COMPARISON between the Population & (Sub-) Segment » helps you to understand the specificities of this group on key measures (% of the total population accounted for by the segment, % of total sales accounted for by the segment,total count, total sales, average number of distinct products per customer, recency and frequency. You also see a comparison of the « TOP 5 PRODUCTS » (most frequently bought products) for the Population Versus the top 5 for the segment along with the respective percentage of customers having the product.

Once you understand the segment’s distinct characteristics, You can, for example, set up a new marketing action promoting  products  in the « TOP 5 RECOMMENDED OFFERS » or combine them with « TOP 5  PRODUCTS » in « bundle offers ». the "%" column in « TOP 5 RECOMMENDED OFFERS » shows the percentage of offers accounted for by each "top 5" product in the total number of recommended offers for the group. 

As you can see, the analysis is also possible at "Sub-Segment" level, a Sub-segment is a "smaller" more "refine" group of customers within the bigger segment. Each Sub-segment is included into one unique segment. It is sometimes worth having a look at them especially when looking for a very specific "niche" of product's preferences within your customer base.

Note that you can save the selected (sub-)segment's customer list by ticking “Yes” to the question "Save this segment/Sub-segment selection to an external flat file?" (see bottom of the left panel)

The interface provides 2 graphs (not illustrated here) comparing the distribution of the number of products per customer for the population versus the segment selected and the distribution of the number of days since latest purchase (recency).

  The second tab in the interface « Product Association rules applied to this Segment (or Sub-Segment) » gives an overview of the association between products within the selected Segment. ICORE presents a "Product Network Graph" for the segment showing how the products relate to each other in the customer's buying behaviour. Products close to each other are often bought together, the arrow links these associated products, the figure shown on the arrow is a measure of the strength of the relationship between the products. Hovering over a product circle will let the penetration rate of the product within the segment appear (% of customers who bought the product). This helps understand the strongest product associations ( market baskets) in the Segment.

 Right under the graph is a list of all the product association rules applied in order to define the  Recommended offers for each customer in the segment.

 Each association (basket) is described by the trigger product associated and the selected  recommended offer (Gap offer product) associated with.

3) The ICORE SELECTIONS interface allows you to initiate Marketing Cross-Sell Campaigns based on recommended offers:  you type in up to 5 products and the interface will  select customers having one or more of these specific product(s) as recommended offer(s). You can also "tick in" a specific type of recommended offer (individual or popular in segments or both) the tool will then further select customers with the specific offer type only. This choice depends on how "pure" you want the customer selection to be.

The sections:  «GENERAL OVERVIEW » ,  « TOP 5 PRODUCTS » & « TOP 5 RECOMMENDED OFFERS » give a comparison of the selection with the Population using the same measures as in ICORE SEGMENTATION.

On ticking "Yes" to "Save this Selection ?" You can then save this list of selected customers in a flat file for later use in the setup of a Marketing campaign. 

The interface produces 2 graphs (only one shown here) comparing the distribution of the number of customers for the population and the selection across the clusters. This graph helps to better understand the "profile" of customers selected via specific recommended offers. Deviation from the population distribution indicates how different the selection is from the all population. it is very common that a specific selection for a recommended offer will "focus" itself on one or two clusters in particular. It is all natural as the clusters themselves were built around "product preferences". 

A second tab 
« Product Association rules with the selected product as Gap offers » is also available in ICORE SELECTIONS (not illustrated here) & lists all the product association rules (baskets) that define the specific products as recommended (gap) offers.
This helps understand the product associated with the selected recommended offer and also analyze the potential differences in association rules across clusters (different «trigger products» in different clusters) .
Each association (basket) is described by the trigger product associated with the selected recommended offer (Gap offer product).

Other specific statistical measures which indicate the " statistical strength" of the rule are given as well as the product category