Intelligence 360

Business Landscape

In today’s world, businesses are no longer competing on price, product, and promotion. Enterprise Customer Experience is today’s competitive advantage, and this strategy requires a consolidated and comprehensive understanding of the customer relationship.

Organizations collect and activate customer data in silos that are owned by various lines of business. Fragmented, incomplete, and conflicting customer data results in disconnected and inconsistent customer experiences. To create a comprehensive single customer view, an organization must untangle the knots of first, second, and third-party data, as well as integrate structured data and unstructured interactions, and deliver unified customer perspectives to support the differing needs of the multiple lines of business. Advanced AI is needed to mine context, relationships, and customer intent as well as to automate the vast amounts of data available.

Intelligence solutions leveraging Informatica Customer 360 Insights (C360i) will generate deep customer insight from a broad range of data sources and continuously learn and refine intelligence from customer interactions. C360i is built using next generation technologies such as Apache Hadoop, Graph data stores, Elasticsearch, and machine learning. C360i creates an actionable customer 360 to address both operational and analytical use cases across the enterprise.

Critical Implementation Success Factors 

There are four critical success factors for an Intelligence Solution implementation: 

• Enterprise Thinking -- The team needs to be willing to break down silos and work across business units.
• Business Vision & Business Case -- Defining the business case for the initiative is especially important to drive business value. Be clear in the vision. Define a measurable set of business goals, objectives, and KPIs related to revenue, cost, and engagement. 
• Use Case Roadmap -- Customer experience, technology modernization, and digital transformation initiatives are examples of corporate strategies that are often journeys themselves. It is essential to support the long-term vision by breaking these initiatives down into smaller, achievable steps that provide quick wins. 
• Communication Planning -- Enterprise adoption will only happen if people know about the advancements and are informed on an ongoing basis. 

Getting Started with C360i

The implementation strategy for C360i is to “Accelerate First Value” by launching use cases that will drive business benefits quickly. For expediency, Informatica recommends leveraging many out-of-the-box capabilities and building upon these services while traveling through the Use Case Roadmap. This approach is iterative and agile, delivering incremental value with each step along the journey.

1. Ingestion

As use cases are prioritized, data source priorities will be identified. The initial sources should drive the most value and satisfy requirements for multiple use cases, where possible. C360i ingests structured and unstructured data from a wide variety of sources including customer details, account information, customer transactions, product meta data, and customer interactions.

The ingestion services leverage big data technologies to consume batch and real time streaming data. The services are designed for an Extract, Load, and Transform (ELT) approach to lift and shift data from the source into the extensible C360i data model. The ingestion services adapt as model changes are introduced.

Best Practices:

• Rank Use Cases by impact and complexity.
• Prioritize data sources that support multiple use cases.

2. Enrichment

Data enrichments enhance the quality of the data and identify key elements that will inform the synthesis (matching) processes. Common out-of-the-box enrichments include demographic enrichment, sentiment enrichment, and standardizations for names, address, email, phone, gender, and dates. The enrichments consist of rule-based enrichments and machine learning enrichments. Enrichment functions can be easily added to address the specific requirements of an implementation.

Best Practices:

• Profile data sources for sparsity, valid values, and cardinality.
• Enrichments may be processed post-ingestion to affect a single record or after entity management to consider data across records within an entity.

3. Synthesize

The Synthesis process performs contextual matching and linking for party-to-party and party-to-object matching. By default, the C360i match engine links records that are classified as Confirmed Links and Probable Links. The linked records are automatically merged by applying pre-configured survivorship rules that determine which data attributes and business objects are maintained and prioritized in the merged records.

The match engine leverages an advanced genetics algorithm that is trained during the implementation process. During training the Data Stewards review match groups and provide input into the validity of the match. Data Stewards select any of the records classified under the various buckets and choose to link or unlink them. Further they can provide a reasoning for their action of linking or unlinking. These actions can be performed on single match pairs or in bulk mode.

The matching engine learns from Data Stewards decisions and tunes the matching processes. In general, 2 to 4 iterations are executed before finalizing the match algorithms. All match relationships are stored in a graph with confidence intervals ranging from zero to one indicating the strength of the match.

Linking is used for data sources that provide a foreign key relationship, and the provided identifiers are used to make and maintain the connections. An example of linking would be tying order line items to the correct order.

Many times, an organization has a “trusted” data source (e.g., MDM implementation). C360i will honor the linkages from the trusted source and augment connections with intelligence from interactions, third party data or other sources.

Best Practices:

• In match training, assign match pairs within a match group to multiple Data Stewards.
• Leverage OOTB match types provided by the core product -- Individual, Organization, Address, Email and Webchat – to speed implementation.

4. Entity Management

C360i provides different perspectives (views) of the same underlying customer data based on configured match thresholds, business objects, and attributes. C360i provides two OOTB perspectives -- Marketing Customer Perspective and Operational Customer Perspective.

Entity ID’s (EID) are assigned within a perspective organizing data relationships for the specific customer view. Synchronizing real time and batch sources is simplified within C360i. For example, when transaction data is ingested before the associated party information, C360i will store the transaction and incorporate it into the Perspective once the party entity has been ingested and an EID assigned

Best Practices:

• Minimize the time lag across related subject area sources, e.g. transaction and party.
• Schedule near real time processes to enrich and synthesize data frequently to maintain the freshness of the Perspectives.

5. Reasoning

Use Visual Analytics to drive business decisions with the clean, standardized, enriched, de-duped, and insight inferred data. Visual Analytics include preconfigured dashboards to surface the business intelligence in the data. Custom dashboards can be created for any perspective, configured in a chosen presentation style.

Embedded models are available to analyze perspectives and source data for life event, customer’s sentiment, and next best action/interaction. Custom models and analytics are easily embedded to take advantage of the best data.

Best Practices:

• Organize dashboards by team to streamline information sharing.
• Customize dashboards by role drive user engagement.