What is household identification?
Let‘s meet Miguel, his wife Sandra, and son Carlos. They use 8 different telecommunication services, and share some of these services (e.g., TV, Netflix, Spotify) as a family.
Both Miguel and Sandra have a postpaid mobile rate plan subscriptions with iPhone leasing contracts, while Carlos has a prepaid SIM card.
As leasing contracts for Miguel and Sandra will expire by the end of the month, it would surely serve as an excellent opportunity for communication services provider (CSP) to contact Miguel or Sandra and recommend them a special family bundle with 2 new iPhones, a family postpaid rate plan and a Kids TV channel package for additional $10 per month.
Highly relevant and customer context-focused recommendations like these can significantly improve overall customer experience and help CSPs deliver more value for their customers. Unfortunately, not many telecoms can deliver such personalized services even in this age of digital technology.
Extracting household context tends to be an incredibly complex task for telecommunication services providers due to siloed legacy business systems that contain different customer information.
For instance, Miguel‘s and Sandra‘s family throughout all IT systems would be identified as 6 different customers.
The siloed customer understanding serves as the main driver of inappropriate and costly customer engagements, like call center agents upselling customers with services they already possess. Lack of proper intelligence and contextual understanding often dilutes the returns of such commercial initiatives.
Key challenges when building household identification
While working with telecoms globally, we‘ve learned that there are 4 key challenges telecommunication services providers face when building a household identification process:
1. Collection of data
The customer information is based on multiple internal & external systems associated with their unique data structure and logic.
|Internal traditional data||Internal non-traditional data||External data|
Data collection not only requires infrastructure with the capability of managing these large data sets but also needs the implementation of complex data pipelines with strict governance processes to ensure data quality.
2. Data normalization and unification
Data collection and centralization might not be enough. Data also needs to be cleaned, normalized and unified for identification of unique customers and insights extraction.
In this stage, data engineering teams mainly face data inconsistency and quality challenges which can be solved with predefined business logic and algorithms or by rigorous customer data validation processes.
3. Building data hierarchy
The next big challenge is to define how and in what format normalized data should be stored to meet different analytical needs in the company.
If data will be stored in the original format (e.g., detail event logs), analysts working in the company will always need to carry out complex data processing actions in order to extract simple insights. If data will be pre-aggregated, part of important information could be lost in the process. Furthermore, it is essential to clarify what system IDs should be used in aggregation to align with complex customer identification hierarchy. In other words, information needs to be processed on different customer views (e.g. subscription, client, household) to ensure the efficiency of analytics processes.
Typically data engineers cope with this challenge by defining data model architecture, which sets clear rules on how the data need to be stored. Over time, such data models become too complex to work with and require significant efforts to be maintained.
4. Activate collected and processed data
However, the biggest challenge is at `the last mile‘. Processed data need to be exposed to the rest of the organization for the implementation of personalization initiatives. With every new customer-focused initiative, the engineering team needs to build specialized data load processes or integrations for data exposure to other IT systems.
Over time, it becomes very complex to govern all of these processes, especially as they have a direct impact on customer experience and business processes.
But, only through the democratization of the customer data, telecoms can transform their current processes, enable proactive customer engagement and personalize the customer experience.
Streamline personalization initiatives with Exacaster Customer Data Platform (CDP)
To help our Telco clients to streamline personalization initiatives and cope with these challenges, the Exacaster team has developed a customer data platform – Exacaster Customer 360.
Our platform works as a centralized repository for the automatic collection, processing, and exposure of all internal and external customer data. The platform is equipped with out-of-the-box solutions tailored for telecom use cases, enabling lower time to market and higher ROI of personalization initiatives.