Enabling Cross-Selling Across Group Companies By Centralizing Data

How about making an extra $500 million in two years time with what you already have? Central data hubs and data marts have generated significant benefits for group companies that have utilized them. More and more group companies are taking heed and are launching similar initiatives to reap the benefits.

You can download PDF version of this whitepaper here.


Customer data hubs and data marts, which bring together customer data from all key operational systems and structure them in a more manageable and informative manner, have been popular among most leading companies for years, serving as a sound base for analytically driven customer strategies and tactics. Central data hubs and data marts are the next step for conglomerates wanting to take the concept to the next level, bringing together customer data from across all group companies, to facilitate group-level understanding of customers and opportunities.

Central data hubs and data marts create the data foundations for facilitating synergies, enabling customer data sharing in a controlled and well-governed way, while increasing the data quality and quantity for all. Although it is heavily supported by data and technology, companies shall avoid considering this as yet another technology solution, and make it a business-driven initiative, as the business implications and impacts are vast.


Various leading companies in North America and Europe, such as RBC Canada, Hilton Hotels Corp, and Harrah’s, have been using central data models to facilitate sharing and making use of customer information compiled from all individual business lines and group companies for years, and have realized handsome returns for their investments. The concept has caught on in the Middle East more recently; one regional conglomerate attributed over $500 million of its revenues earned over the past two years to the use of its central data mart, leveraged through its group loyalty program.

Here are a couple of reasons why such a relatively low cost initiative can generate such impact on the bottom-line:

1. The first and most obvious benefit of central data models is the opportunities created to cross-sell to group customers across companies. A central data hub and data mart brings together detailed information about customers, collected across various industries, building a very solid lead list, with details on customer value and potential, needs, and behavior. Such a central data model presents all the untapped potential within the customer and non-customer base for all group companies. This benefit extends to the new business development areas for conglomerates, building a strong lead list for a quick and successful launch in new ventures.

2. Similar to understanding the opportunities in the group customer base, a central data model can also facilitate better understanding of risks – in both financial and relational terms. For example, a customer who has defaulted or delayed payments in one of the business lines can be flagged as high risk in other group companies. Or, similarly, another customer churning from one of the group companies may require special attention in other lines, as the churn might indicate disloyalty to the group brand as a whole.

3. Another important benefit from a central data hub is the availability of most up-to-date and accurate customer information for all group companies. More often than not, when group companies bring together their customer data, they realize that some of the other group companies have more recent contact information or more complete demographics data on their customers. Centralizing customer data is one of the cheapest and most effective means for enriching the data for all. It is even possible to take this one step further and direct group companies to gather information for not just their own benefits, but for the other companies in the group. A simple example would be a car rental company, which can easily collect information on customers who rent automobiles for a relatively long-term and identify those who do so because of a recent relocation. Although this information has limited benefits for the car rental company, it could become a valuable piece of information for one of the other group companies; say, a home appliances retailer.

4. It is a known and fairly common case; one of the highest value customers of a company visits a store of its sister company. Expecting that he would receive the same special treatments he has enjoyed in other company, the customer becomes frustrated when no one in the sister company even recognizes him. For the first company, this means dissatisfying a high value customer, whereas for the sister company, this means a customer with relatively high potential gone unnoticed. It is simply a lose-lose situation for all parties involved. Without merging the customer data from all group companies, such cases will continue to exist in any given conglomerate.

5. Another common case in conglomerates without central data and customer relations management is the overwhelming amount of communications their customers get. Without knowing how much each group company sends to a specific customer, these groups end up sending tens of messages to the same customer, and sometimes with conflicting propositions or even offering competing products. Of course, resolution of such issue goes beyond just centralizing the data and calls for more organized approach to communications with customers as well.

Nowadays, everyone talks about synergy within group companies, and managing the total customer relationship and experience across a group, but very few take solid steps towards building them. Central data hubs and data marts are one of the few quick win steps and major enablers for such objectives, which shall become one of the building blocks for any conglomerate.


There are six main steps that need to be taken towards building and leveraging central data hub and data marts in group companies:

1. Design of the Data Model: The first step in design is understanding of the business requirements across the group, to come up with a comprehensive listing of the critical data elements and customer information expected from the central data model. The key challenge at this stage is to support business teams in thinking outside of their own boxes (own business lines) and envisioning what they could get from other lines of businesses. Benchmarks and best practices definitely help at this stage, but for best results, cross-industry experts and group level brainstorming sessions should be part of such a process. Based on the business requirements, the ideal data model design should be prepared, which requires flexibility to cater for not only existing business lines’ data needs, but also potential future ventures of the group as well. The output from this step is usually a relatively large-scale data model, which shall be prioritized for implementation as well. One of the leading conglomerates in the Middle East has started with over 7,000 data elements for its central data mart design, which included only the first priority fields.

2. Data and Customer Governance: Although the governance model does not get into action before implementation of the data model, being one of the most challenging and make-or-break aspects of the central data models, companies need to start discussing the governance model as early as possible in the process. The governance model not only affects how much customer satisfaction or dissatisfaction the central data model can create, but also assigns ultimate power and responsibilities, making it a highly political and long-duration discussion topic. Companies need to take into account various policies within this step, including:

  • Customer Data Feed: Whether or not to include all sensitive customer information (e.g. including private data such as medical details, as well as sensitive customers such as VIPs of companies.
  • Customer Data Access: To what extend the group companies and external parties can have access to combined customer data (e.g. summarized data only, details for only their own customers, details for all customers and leads)
  • Customer Communications: How frequently, with what priorities and through what means the customers can be reached out to (e.g. via a central CRM team, with 1 message per month from each group company.
  • Customer Permissions: Whether opt-in or opt-out option will be used for sharing customer data across group companies and taking actions
  • Customer and Data Ownership: Who are the owners and guardians for the shared customers and customer data (e.g. central CRM and IT teams, shared ownership between companies, one company acting as the guardian)

Although there exist various rules of thumb and proven methods for each of these aspects, various conglomerates opt for different decisions in them, based on their market and business conditions, calling for a well-thought out solution instead of a one-size-fits-all solution. For example, one European conglomerate uses a centralized newsletter to integrate communications with customers, which limits the number of communications a customer could receive, while giving opportunity to each group company for reaching out to him. On the other hand, a North American conglomerate prefers centralizing the campaign management for the whole group, selecting the best offer for each customer from across the group and communicating them, following a truly customer centric approach. Companies need to analyze their own requirements and customer expectations and come up with a solution of their own, while studying such best practices.

3. Preparation of the Data: Of course, the actual work of bringing the data together, merging separate data sources from across the group companies, is a major step of its own. This step involves mapping of sources’ system fields with the central data model, extraction of vast amounts of data from them, cleansing and standardizing, transforming and loading into the central data hub and data marts. One of the key activities within this step is unification of customers – and possibly households – which have records with varying levels of accuracy and completeness across the group. Here, companies need to set up customer matching rules, which define what constitutes a unique customer (e.g. records with same name, similar address and same phone number). Additionally, it is necessary to define the rules for specifying the golden records when performing the merge (e.g. if a customer has two different age information in two different business lines, which record should be used).

4. Data Improvements across the Group: Central data hub and data mart development activities always demonstrate substantial gaps across the group companies, as it gives the opportunity to compare quality of data between them, and identifies additional requirements to support each other. As a natural consequence, central data model development process requires data improvements across most business lines. Although the journey for ideal customer data is a continuous one for most companies, conglomerates need to take this into account when planning for implementation of their central data hubs and data marts, assigning ownership and responsibilities for data quality improvements.

5. Group-Level Customer Analytics: Once the central data model is in place, the next step is analysis of compiled customer data, running various customer analytics models at the group level. Similar to applications at individual company levels, customer analytics can reveal critical insights for the group as a whole. For example, companies can identify their high value and high potential group customers, and can use bits of information across the group to identify customer segments. As the variety of channels, products and services offered to the customers increase, so too do the opportunities to understand them better (such as understanding their lifestyle or socio-economical status). For example, a conglomerate, with a customer who has recently purchase a Ferrari from its automotive business line can easily cross-sell from its luxury retail lines, identifying such customer as trendy and high net-worth.

6. Business Use of Data: Last, but not the least, companies need to put the customer insights and leads into use, incorporating them into their day to day operations (e.g. prioritizing the high value group customers in their call center, even if they are not the high value customers for their own business line), and come up with numerous campaigns to tap into the available potential for benefits. These activities shall bring the group companies closer, with task-forces working to come up with cross-company campaigns and opportunities, making the synergy myth a reality for all.

Each of these six steps requires detailed planning and involvement from various stakeholders, requiring a well-structured program management office.

What Next?

Building a central data hub and data mart, and establishing a group level customer and data governance model around this involves a lot of convincing and company politics, and requires top level buy-in and sponsorship. As a result, we recommend starting with some proof of concept, such as bringing data on sample lists of customers across group companies and testing cross-company pilots to demonstrate the benefits up-front. This has worked extremely well in various cases, such as one of the leading groups in the North America, where contributions to the central data model was initially voluntary for group companies, but once the results started to come in, no business line wanted to stay out.

To learn more about making most out of your group customer data, please contact info@forteconsultancy.com.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s