Digital Servicing Using Artificial Intelligence
4 likes270 views
Celcom Axiata leverages artificial intelligence (AI) to enhance customer service and improve business results through data unification and interaction analytics. The company emphasizes the importance of creating rich customer experience data to enable AI applications, resulting in better service delivery and cost efficiency. Current AI deployments include predictive analytics for network performance and customer interaction, as well as the introduction of smart virtual agents to enhance digital customer engagement.
![We Have Achieved a Lot With This Approach.
By bringing together the total ecosystem inclusive of AI, Celcom has achieved in parallel a
substantial positive impact on customer experience and cost.
+30
3Q173Q16 1Q183Q15 4Q17
+32
3Q15 3Q173Q16 4Q17 1Q18
RETAIL SHOPS CONTACT CENTRE
44
1247
77
TOUCH POINT NPS DIGITAL INTERACTION GROWTH
SOCIAL MEDIA GROWTH
60%
Celcom
4x
Peers*
15%
SOCIAL MEDIA REACH
[YoY Growth – ‘17 vs. ´16 in %]
-22%
CelcomPeers*
28%
SOCIAL MEDIA ENGAGEMENT
[YoY Growth – `17 vs. `16 in %]
61
55
28
14
2018TD201720162015
% of All Interactions
Source: SocialBaker](https://image.slidesharecdn.com/2018-06-07connectechconferencecelcomrenewernerdigitalservicingusingartificialintelligence-180701081849/85/Digital-Servicing-Using-Artificial-Intelligence-7-320.jpg)
Digital Servicing Using Artificial Intelligence
- 1. Digital Servicing Using Artificial Intelligence René Werner, Celcom Axiata Berhad 7/1/2018
- 2. Artificial Intelligence & The Broader View. At Celcom we focus on leveraging AI to support our purpose statement. AI is not a self serving purpose. CELCOM´S PURPOSE STATEMENT
- 3. 34 21 Growth in Available Bandwidth (Nielsen‘s Law of Internet Bandwidth: CAGR 50%) Growth in Compute Power (Moore‘s Law: CAGR 60%) Growth in Storage Density (Kryder‘s Law: CAGR 40%) Growth in Compute Energy Efficiency (Koomey‘s Law: CAGR 50%) 0.000 0.000 0.001 0.010 0.100 1.000 10.000 100.000 1,000.000 10,000.000 100,000.000 3/10/1954 11/6/1968 18/2/1982 28/10/1995 6/7/2009 USD/MByte Artificial Intelligence Made Possible. The progression of technologies along four key „laws“ has made AI a reality now with completely changed economics to penetrate the day-to-day live of businesses & consumers. Example: Historical Cost of Computer Memory & Storage Big DataCompute AI Source: hblock.net/storage
- 4. Artificial Intelligence Today is NARROW. Further technology progress needed to move todays NARROW AI towards a GENERAL AI that has a broad use spectrum vs. purpose built deployments. Source: http://www.visualcapitalist.com/visualizing-trillion-fold-increase-computing-power/ TODAY´s AI vs. HUMAN BRAIN AI Assisted Movie Recommendations GENERAL vs. NARROW AI General AINarrow AI AI Assisted Natural Language Processing AI Assisted Driving AI Based „GO Expert System“ (Google Deepmind) 0%of the use cases today Computers thinking at a level that meets or surpasses people Abstract and independent reasoning & thinking without limitations to expertise domains Time of Arrival: 2035+ (Estimates) 100%of the use cases today Computers detecting useful patterns to solve problems by crunching vast amounts of data Pattern recognition & automation usually limited to one expertise domain Time of Arrival: today General Purpose AI
- 5. Artificial Intelligence & Customer Service. Large amounts of unstructured (interaction) data and need to understand patterns to improve service & experience are ideal conditions to leverage narrow AI. Messy CX Data multiple interaction touchpoints divergent customer identifiers varying interaction data types (voice, text, clickstreams, ...) heterogeneous CX „Health“ data (network, service, ...) multichannel triggers for interaction journeys heterogeneous customer sentiment data Sub-Optimal Service & Experience Delivery Generate Interaction Data (Enrich Interaction with Data) Insights Creation (Pattern Recognition) Leverage Insights for Better Service & Experience Delivery (Automation) More Customers/ Lower Cost AI Enhanced Service & Experience Delivery Better Business Results Create CX Data (Clean & Rich Data Sets via Data Unification) 1 2 3 4
- 6. LEVERAGE INSIGHTS (Automation) INSIGHTS CREATION (Pattern Recognition) CREATE CX DATA (Data Unification) Get Going on Artificial Intelligence. To be able to leverage on artificial intelligence, rich CX data sets from interaction data have to be available. Building those up and connecting them is key for reaping AI benefits. GENERATE INTERACTION DATA Retail Shops/ Branches Contact Centres Traditional Channels: Enrich Interaction Data IVR ... Digital Channels: Drive Migration to Digital Company Website Social Media Company App Others: Get deeper Usage Insights Network Experience Product Usage Experience ... 1 2 3 4 Customer Identification Journey Stitching Experience Issue Clustering Core AI Application Experience Opportunity Identification Channel Preferences & Effectiveness, Interaction Bottlenecks Customer Sentiment (State & Forecast) Channel Volumes (State & Forecasts) Automation: Use AI to Drive Action Intent Recognition (State & Forecast) Customer Service Bots Customer Service Agent Scheduling Network Field Force Dispatch ... Support: Use Insights to Drive Better Human Action Better Up- & Cross-Selling Better Network Planning Ambient Retail Influencer Marketing Six Sigma Initiatives AI Enabling AI Harvest
- 7. We Have Achieved a Lot With This Approach. By bringing together the total ecosystem inclusive of AI, Celcom has achieved in parallel a substantial positive impact on customer experience and cost. +30 3Q173Q16 1Q183Q15 4Q17 +32 3Q15 3Q173Q16 4Q17 1Q18 RETAIL SHOPS CONTACT CENTRE 44 1247 77 TOUCH POINT NPS DIGITAL INTERACTION GROWTH SOCIAL MEDIA GROWTH 60% Celcom 4x Peers* 15% SOCIAL MEDIA REACH [YoY Growth – ‘17 vs. ´16 in %] -22% CelcomPeers* 28% SOCIAL MEDIA ENGAGEMENT [YoY Growth – `17 vs. `16 in %] 61 55 28 14 2018TD201720162015 % of All Interactions Source: SocialBaker
- 8. LEVERAGE INSIGHTS (Automation) INSIGHTS CREATION (Pattern Recognition) CREATE CX DATA (Data Unification) Example of Deployments in Customer Service. From numerous deployments we had made in the past. GENERATE INTERACTION DATA 1 2 3 4 Retail Shops/ Branches Contact Centers Traditional Channels: Enrich Interaction Data IVR ... Digital Channels: Drive Migration to Digital Company Website Social Media Company App Other: Get deeper Usage Insights Network Experience Product Usage Experience ... Customer Identification Journey Stitching Experience Issue Clustering Core AI Application Experience Opportunity Identification Channel Preferences & Effectiveness, Interaction Bottlenecks Customer Sentiment (State & Forecast) Channel Volumes (State & Forecasts) Automation: Use AI to Drive Action Intent Recognition (State & Forecast) Customer Service Bots Customer Service Agent Scheduling Network Field Force Dispatch ... Support: Use Insights to Drive Better Human Action Better Up- & Cross-Selling Better Network Planning Ambient Retail Influencer Marketing Six Sigma Initiatives AI Enabling AI Harvest AI - Great Network Experience AI - Superior Social Media Growth AI - Contact Centre AI – Birth of Virtual Assistant AI - Voice of Customer Feedback AI – The Next Level A B C D F E
- 9. AI - Voice of Customer Feedback. Realtime customer feedback with analytics to address problem areas. Automated Generation of Heatmap from VOC “Pounding the Rock”Realtime VOC Feedback Crowdsourcing of “Voice of Customer” on various channels & customer journey Per channels, we create heat maps by automatically tagging to the channel & journey Daily scrums leveraging on heatmaps to address problem areas A >5mn Yearly Survey
- 10. AI - Superior Social Media Growth. Holistic approach to Social Media underpin by AI foundation. Customer Engagement Analytics and Measurement Generate Data (Quality & Quantity) • Connect and crawl the social web • Derive actionable insights • Intelligent listening • Engage key influencers • Drive media value • Posting the right content • Develop actionable insights. Eg, complain resolution, social selling & conversion etc • Define and track social metrics • Identify key influencers • Identify key customer segments Social Media Servicing Celcom Community Social Media Analytics CELCOM SOCIAL MEDIA EXCHANGE (SMEX) Setting up of social media hub to service & engage with customer, managing brand and contextual offerings B
- 11. AI - Great Network Experience. Predictive analytics tool for automatic root cause analysis and resolution. Generation of Insights Action AreasGathering of Data Internal Network Performance Data From probes on RAN and Core Network Crowd-sourced (Anonymous) Usage Data From ~ 2.5m App users Journey NPS data From >5mn surveys 1 2 3 Example: Measuring Network Performance vs tNPS • Self-optimized network • Better network planning • Early communication of network issues to customers Example: Network Heatmap – Hot zones of the network C
- 12. AI - Contact Centre. Predictive analytics solution to analyze customers‘ calls to improve communication and future interaction. Predictive analytics solution Predictive analytics solution to gauge customers’ emotion based on voice tone and manner Benefits • Improve interactions quality and call experience • Automatic documentation of calls • Better up-sell /cross-sell opportunities • Reduce customer attrition * Interaction with ~300k calls per month D
- 13. AI – The Next Level. Bringing everything together. Customer analytics solution to understand and visualize individual journeys for customer across all touchpoints; seamlessly and in real-time. Improve Net Promoter Score (NPS) o Identify areas to improve pain points o Identify journeys and customers who log frequent complaints across channels Up-sell / Cross-sell opportunities o Predict next customer interaction step/channel in a journey o Linking customers to journeys and analyzing customers’ demographics, segments and sentiments Reduction in cost-to-serve o Automatic prioritization of journeys / pain points driving experience, costs, churn with built-in algorithms Customer Journey Analytics Solution Predictive analytics solution to understand and visualise customer journey across touchpoints; seamlessly and in real-time Benefits E
- 14. AI – Birth to Virtual Agents. Launched smart virtual agents with distinct character for a more personalized experience. Clive & Emma State-of-art intelligent Virtual Agent with 2 personas powered by Microsoft technology. Currently, it can perform the following functions: • Prepaid & Postpaid – check bill/balances, reload/pay bills, check data usage, add VAS & Mobile Internet • Humanized tone • Small Talk & Personas • Ability to handover to Live Agents Next phase is to grow the Virtual Agent to perform more activities • Handling of natural language capabilities • Interface from “voice to text” and from “text to voice” • Handling of more transactions • Expansion to other servicing channels like FB Messenger F
- 15. Artificial Intelligence & The Broader View. At Celcom we focus on leveraging AI to support our purpose statement. AI is not a self serving purpose. CELCOM´S PURPOSE STATEMENT
- 16. THANK YOU !!! If you have questions contact me as below: rene.werner@celcom.com.my Linkedin.com/in/rene-werner