Customer Driven Big-Data Analytics for
the Companies’ Servitization
Eugen Molnár, Natalia Kryvinska, Michal Greguš
Principal interactions in a PSS delivering
an advanced Service
Data – an important dimension of
servitization
Data – an important dimension of
servitization
Successful delivery of advanced services is enabled by information and
communication technologies that are focused on informing and advancing actions of maintenance, repair and use.
Source: BAINES, J.-LIGHTFOOT, H. Made to Serve: How Manufacturers Can Compete Through Servitization and Product Service Systems. John Wiley & Sons, 2013.
Tomorrow, information deriving from servitization and exploited on an ecosystem level, could represent the third source of revenue streams for the manufacturer. It could even become the manufacturer’s main revenue stream. To depict the impact of the introduced concepts while positioning it in relevant literature, a third layer of added value was added to Thoben’s representation of servitization levels, the
information layer.
Source: David Opresnik, Manuel Hirsch, Christian Zanetti, and Marco Taisch: Information – The Hidden Value of Servitization
An excellent example joining of information with business demonstrated Fred Smith, founder of FedEx, who said that ‘‘the information about the package is as important as
the package itself’’ and applied this insight to develop the real-time tracking tools that
gave his company a huge advantage in the marketplace.
Wheel of Social Media
Holistic customer view
Speech Analytics
3 Pillars of Customer Experience
Source: Soudagar, R., V. Iyer, and V.G. Hilderbrand. 2012. The Customer Experience Edge: Technology and Techniques for Delivering an Enduring, Profitable, and Positive Experience to Your
Big Data & Unstructured Data Analytics
• Customer analytics is an emerging approach to truly create meaningful, lasting, and profitable customer interactions through the systematic examination of a company’s customer information (internal, syndicated or social; structured or unstructured) to attract, retain, and grow the most profitable customers.
The main research directions in Business intelligence and analytics:
• Big Data Analytics – “technology oriented” research questions related to Hadoop
and MapReduce
• Text Analytics - was moved from search engines to enterprise search systems and
from information exraction to Question Answering systems.
• Network Analytics - Link mining, community detection, social recommendation
are the main areas in this kind of research.
Source: LIM, E.-P.-CHEN, H.-CHEN, G. 2012. Business intelligence and analytics: Research directions. In: ACM Trans.Manage. Inf. Syst. Vol.3, No.4, Article 17 (January 2013).
Innovation in Monitor, Analyze and
Respond
Source: BAINES, J.-LIGHTFOOT, H. Made to Serve: How Manufacturers Can Compete Through Servitization and Product Service Systems. John Wiley & Sons, 2013.
Monitor – Social Media, corporate web pages, chats, discussions, blogs, Call center interaction
Analyze – Text Analytics
Respond – dynamic dashboard, multichannel interaction with customer, pushing relevant data for example e-commerce, FAQ
Text Analytics
What can one mine from unstructured (text)
data?
Information processing
Text Corpus Preprocessing Representation Knowledge DiscoveryInformation processing
Text Corpus Preprocessing Representation Knowledge Discovery Categorization
Process of deriving precise categories
through conceptual understanding.
Rule-based training also supported.
Automatic Categorization
Why are Taxonomies useful?
Makes unstructured information more
accessible through directed navigation
Clustering Visualizations
Visualizations
Clustering of data
Clustering of users
Time analysis
Cascade clustering
Node Decomposition Analysis
(NDA)
Cluster-on-demand
Drag and click defined regions of
interest
Architecture of an Enterprise Intelligent
System
Rules Engine/BPM
Intelligent Engine
Connectors
Application Framework
Analytics
Rules Engine/BPM
Intelligent Engine Connectors
Application Framework Analytics
Core of this architecture is Intelligent Engine that
covers such capabilities as: Conceptual search
Advanced search methods Categorization Clustering Sentiment analysis Hyperlinking Entity extraction Personalization
A Rules Engine/BPM increases traditional Enterprise Search Engine and allows
implementing certain kind of business logic that can use Intelligent Engine to deliver more sophisticated and more relevant answers.
On the top of this architecture is an Application
Framework - APIs those provide IT capabilities and allow to compose flexibly business specific application. Such applications cover different business requirements.
Analytics tool provides capabilities for data and information processing to discover
enhanced insight and decision making.
Connectors represent an integration layer and the boundary between data sources and Enterprise Intelligent System. Most of
systems consist already of a set of connectors to standard system or social media.
Architecture of an Enterprise Intelligent
System
Sample of Rule
Source: http://decisions.com/platform/rules/ and http://docs.jboss.org/drools/release/5.3.0.Final/drools-expert-docs/html/ch01.html
Real Knowledge Management
• Company
• Knowledge discovery from different sources
• Knowledge from KB and relevant information are used during an
interaction with customer
• Customer – Relevant information are provided
Benefits
• Company
• provides higher quality of service
• reduces costs
Screens and story
Corporate Web: FAQ Engine
User fill contact form
User gets list of
related FAQs
before request is sent
Screens and story