Speaker presentation from TCI conference:
Big Data & Analytics in Call Centres Conference 2016
26 - 27 October 2016 | Indaba Hotel | Johannesburg
Cognitive technology and analytics to drive value in call centers
• extended single view of customer to support & guide call centre client conversations
• enhanced customer insight through matching social media data to
enterprise customer data
• using real time analytics to propose the next best action
• using cognitive technology to drive better outcomes for both vendor & customer
Wolfgang Knupp, Research: Business Analytics, IBM
Beginners Guide to TikTok for Search - Rachel Pearson - We are Tilt __ Bright...
Cognitive technology and analytics to drive value in call centres v1.0
1. COGNITIVE TECHNOLOGY
AND ANALYTICS TO DRIVE
VALUE IN CALL CENTERS
Wolfgang Knupp
Senior Analytics Architect
IBM South Africa
wknupp@za.ibm.com
2. PREDICTION: IN THE NEXT 30 MINUTES
YOU WILL HEAR ABOUT…
enhanced customer insight through
matching social media data to enterprise customer data
using real time analytics to propose the next best action
using cognitive technology to drive better outcomes for both vendor & customer
extended single view of customer to support & guide call centre client conversations
3. EXTENDED SINGLE VIEW OF CUSTOMER TO
SUPPORT & GUIDE & IMPROVE CALL
CENTRE CLIENT CONVERSATIONS
My name is Eric – I work in a call center for Travelers Insurance.
I get measured on successful call handling time, number of calls
and customer satisfaction. My company wants to keep
average call handle time below 2 minutes..
I am about to complete authenticating my current caller as
customer Jane Robertson
7. ENHANCED CUSTOMER INSIGHT THROUGH
MATCHING SOCIAL MEDIA DATA TO
ENTERPRISE CUSTOMER DATA
So you think you know your customer based on the
internal system information you have available?
What If : you could tell that the person you are
speaking to has publicly complained about your
service on Hello Peter / Twitter under a pseudonym ?
What If : you could tell that the person you are
speaking to is intending to purchase a competitive
product ?
8. SOCIAL MDM MATCHING –
OBJECTIVES AND CHALLENGES
External Social Data
Lead Generation Brand Management …
Enhanced Customer 360O View
(internal + external)
CRM DataCRM Data
Contact Center DataContact Center Data
Web Analytics DataWeb Analytics Data
Enterprise Internal Data
Fraud AnalyticsCustomer Retention 8
9. SOCIAL MASTER DATA MATCHING -
GAINING INSIGHTS FROM SOCIAL MEDIA
DATA
Location announcements
I'm at Woolies Sandton City
http://4sq.com/fAeMRc
Personal Attributes
• Identifiers: name, address,
gender, occupation…
• Life Cycle Status: marital,
parental
Relationships
• Personal relationships: family,
friends,…
• Business relationships: occupation,
employer
Products Interests
• Sentiment on products, services,
campaigns
• Personal preferences of products
• Product Purchase historySocial Media
extended
Customer
Profile
Life Events
• Life-changing events: relocation,
having a baby, getting married,
getting divorced, buying a house…
All names and identifiers are fictitious
Intent to buy a car
The Toyota has finally seen the last of its days,
crying at the fact I have to look for a new car
9
10. ENTITY RESOLUTION - RESOLVING IDENTITIES
ACROSS SOCIAL MEDIA AND CUSTOMER DATA
Example social media user
profile, with sample activity:
Screen name: @kathyg
Name: Kathy Greene
Location: Somewhere in CA
I'm at HP Arena http://4sq.com/ZibpbUI'm at HP Arena http://4sq.com/ZibpbU
@banksupport What happened to the credit
card payment I made last week?
@banksupport What happened to the credit
card payment I made last week?
I'm at Walmart (San Jose, CA) w/ 2 others
http://4sq.com/1TL20yP
I'm at Walmart (San Jose, CA) w/ 2 others
http://4sq.com/1TL20yP
Great promo from @airline. Even more
excited that they're flying PE to BLOEM soon!
Great promo from @airline. Even more
excited that they're flying PE to BLOEM soon!
Customer database
Is this social media user
one of my customers?
If yes, which one?
10
11. USING REAL TIME ANALYTICS TO
PROPOSE THE NEXT BEST ACTION
Derive and act on insights using real time analytics based on
real time context in combination with more traditional attributes
Should this call be automatically routed to a more senior agent?
Is the time right to apologize and provide some complimentary
offer ?
Is there a short lived opportunity to upsell ?
Is there an opportunity to warn the caller of a coming hail storm ?
weather
14. USING COGNITIVE TECHNOLOGY TO DRIVE
BETTER OUTCOMES FOR BOTH VENDOR &
CUSTOMER
Enter Chatbots & Intelligent Virtual Agents :
Better customer experience:
• reduce annoying menu navigation
• less misunderstandings due to accents
• less affected by mood of agent
• better communication (less noise)
• better cultural / demographic style
match between agent and customer
• pace can be adjusted to suit each
customer
• less risk of handing over PII information
to potentially untrustworthy people
Better vendor outcome:
• improved efficiency
• protects brand reputation
• improved customer service keeps customers
returning
• scales easily
• provides consistent service levels,
• promotes and enhances self – service
• Improves by learning from interactions
• easy to train with new products, processes
• elevates personnel job satisfaction to deal
with more exceptional / complex issues
requiring human interactions
• reduces need for large / low skilled workforce
16. USING COGNITIVE TECHNOLOGY TO DRIVE
BETTER OUTCOMES FOR BOTH VENDOR &
CUSTOMER
Chatbots / Intelligent Virtual Agent Platform examples:
Echo - primarily designed for voice interaction in the home
virtual voice based personal assistance
Engagement Advisor / Watson Developercloud / Watson Virtual Agent
(Sept 2016)
Microsoft: LUIS / Cortana (Virtual Agent) and Bot framework
17. USING COGNITIVE TECHNOLOGY TO DRIVE
BETTER OUTCOMES FOR BOTH VENDOR &
CUSTOMER
Building your Intelligent Agent example 1:
IBM Bluemix Watson Deverlopercloud …
API … Services:
• AlchemyLanguage
• Speech to Text
• Text to Speech
• Conversation
• Language Translator
• Natural Language Classifier
• Document Conversion
• Personality Insights
• Retrieve and Rank
• Tradeoff Analytics
• Tone Analyzer
Building your Intelligent Agent example 2:
Facebook Messenger Platform…
• Setup
Configure your Facebook App and
Page for Messenger
• Conversation
Send and Receive messages using
an API and webhooks
• Entry Points
Learn about how people can enter a
conversation with your bot
• Payments (BETA)
Enable your bot to take payments
• Analytics
Use metrics to see how your bot is
performing
18. USING COGNITIVE TECHNOLOGY TO DRIVE BETTER
OUTCOMES FOR BOTH VENDOR & CUSTOMER
The question is not if you should consider cognitive technology in your call center,
but how best to do so to remain relevant.
Where to start on the journey to becoming a Cognitive Call Center…
Build
Ensure that your existing technology
platform is open to integrate new
capabilities:
• in-house development capabilities
or
• a technology partner
• choose appropriate cognitive
technology platform - SaaS
• simple chatbots can be built in
minutes…
Buy
• Work with your existing Call Center
Technology Platform provider to
see if and how they are planning to
deliver new capabilities
• All the major Call Center Platform
Vendors have some roadmaps that
include some cognitive capabilities
– ensure that yours is not leaving
you behind the competition.