Integration, Information and Insights: A Contextual Perspective on Customer Interactions
1. Integration, Information and
Insights: A Contextual Perspective
on Customer Interactions
Igor Nesmyanovich
PhD, CIISP, Partner and CIO, dthree inc.
Steve Vermeiren
Director – CRM and Retention, SIRIUS Satellite Radio
Data integration challenges and the risks of deriving data insights out of context Learn about contextual data integration and how it can help you in your marketing efforts Gain insights into enabling multi-channel, customer focused marketing strategies: Sirius Satellite Radio Canada Case study
Statistical Data Integration Map known characteristic of one group of individuals onto another group of individuals based on a set of common attributes: Demographic Profile Matching Geo-demographic Clustering Audience-topology clustering Calibration Fusion Challenges: Complex process of validation and auditing Difficult apply to non demographic data such as behavioral. CRM Data Integration Data integration on individual Consumer/Customer level Extract data from multiple systems Build data warehouses, data marts, Online Analytical Processing (OLAP) for multi-dimensional queries Data mining for pattern extraction and predictive analytics Challenges: Data formats and metrics vary from system to system Semantic integration problems Integration of structured and unstructured data Quality of data Knowledge extraction from massive integrated data Advertising and branding: Market Share, Brand Development Index, Awareness CRM: Churn, Retention rate, RFM, Customer Life Time ValueDirect marketing: Response Rate, Cost Per Lead, Lead-to Sales Conversion RateAdvertising media: Impressions, Gross Rating Points (GRP), Effective ReachWeb and online metrics: Unique Visitors, Unique Visits, Pageviews, e-mail Opens, Clickthrough Rate
Data integration process that is focused on the particular business or marketing strategy and marketing business decision makers audience Business users of the data and information are able to make decisions and act on the information Decisions and actions can be automated
Define contextual integration domain, business stakeholders and target user information audience Acquiring new customers context Target information user audience: marketing managers, and other business stakeholders involvement of customer acquisition process Generating more business from existing customers context Reducing cost of acquiring context Identifying and serving the customers that generate the most profit Improving customer retention rate context
Determine target actions based on the business context Track target action events Contextual data tagging Use a first party-cookie to maximize life cycle management Online media to get customer centric vs. media centric view Build profiles of anonymous users Data integration with overlay behavioral, demographical and transactional information Use offline to online contextual tracking codes, landing pages to understand channel to channel conversion Enable data quality monitoring Measure data “trust factor” Enrich data with contextual artifacts Use contextual dimensions and measurements
Use of anonymous persona model to integrate data Create anonymous profile with all customer attributes but Personal Identifiable Information (PII). Group similar profiles into clusters of digital persona groups. Apply data mining to find clusters and build automated classification, scoring and predictive models. Apply run time classification and scoring to recognize digital persona. Deliver the most relevant marketing message to the persona group. Privacy Sand box
Personal identifiable information (PII) is never stored in clear text files and can not be accessed directly using general purpose data access or query tools. Access to personal identifiable information (PII) is restricted and enforced based on privacy policy. The “Need to know” security principle is followed. All access to the PII is audited. Each event of explicit or implicit personal data collection has visible contextual explanation what personal data is collected, why and what are benefits for consumer.
Automated rules based campaigns: direct, phone, e-mail, web, online media: Use of static and dynamic rule engines Event triggered communication Online media personalization Multi step or multiple waves and multi channel campaigns use previous step or wave to tune communication in different channel.
- Contest – Where did the name SIRIUS come from? And how is relevant to the company logo? - Started in December 2005 One of the fastest growing subscription based services Achieved 500K subscribers in under 2 years time. Canadian satellite radio leader with over 80% share of the market
In order to stimulate demand and facilitate engagement in our product the online listening trial was developed to give potential customers a sample of the content they could receive.
Message to existing customer to notify them of their existing subscription and method for retrieving Login ID / Password.
SIRIUS has established partnerships with a significant number of automotive manufacturers to enable the reception of SIRIUS Satellite services from integrated/built in stereo receivers. Each automotive partner is unique in the way they enable and offer the SIRIUS radio and service to their customers Ford – Standard feature on most vehicles with 6 month complimentary trial subscription included Chrysler – Factory option on select vehicles and trim levels with 12 month complimentary trial subscription. Mercedes – Factory option on select vehicles / trim levels with 12 month complimentary trial subscription.
Each automotive partner data process set up is unique and challenging. Many data processes go from partner to SIRIUS U.S. to SIRIUS Canada. Timing is critical to ensure confirmation of sold data to SIRIUS Canada triggers trial subscription period and conversion communication streams. Working closely with partner marketing departments we need to ensure that customers are aware that SIRIUS is an option for their vehicle or comes with their vehicle. Training dealers and providing them with marketing materials is another key focus to ensure that customers have a good knowledge of how to access and use SIRIUS as soon as they drive off the lot.
Why do customers leave us? Identify actionable reasons – no need or do not listen anymore not a legitimate reason that we can action. - product not meeting expectations? Received as a gift? Temporary / seasonal? Financial hardship – low on the hierarchy of needs Radio lost or stolen Changed jobs Payment failure Voluntary vs. Involuntary reasons Targeting Reason Code(s) – source of issue Plan type – monthly, annual, two-year, longer? Account value – how many subscriptions on account? Renewal date – when does the subscription rollover to a new term? Radio source type (i.e. Automotive, Retail, Direct, Special Markets)
Value of offer determined by customer profile attributes such as: Number of radio subscriptions Subscription term Customer Tenure Product Acquisition channel
Customer Retention – when are customers leaving? Why? Is this a symptom of a bigger issue? Automation - to alleviate resource constraints, concentrate on continuous refinement and optimization.