BT – Data Analytics
- 2. ● A British multinational telecommunications company with head offices in London, United Kingdom and was founded in 1969. ● CEO: Gavin Patterson (since 2013) ● It operates in over 180 countries and is the largest provider of fixed-line, mobile and broadband services in the UK, and also provides subscription television and IT services. ● Purpose: to use the power of communications to make a better world ● Revenue: £24.062B ● 106,400 employees BT Overview
- 4. How is it relevant to us? Source: Accenture
- 5. Process of examining data sets in order to draw conclusions about the information they contain, with the aid of specialized systems and software, in order to improve products and services, decision- making and business operations. What is Data Analytics?
- 6. “We can now process the billions of messages that are being alerted from millions of pieces of our network electronics. Five years ago that data would not have been processable in near real time, but today it is.” – Clive Selley, BT CIO
- 7. Business Model Canvas Provide world-class: telecommunications, information products & services, and to develop & exploit our networks, at home and overseas - 106,400 employees - Offices over 180 countries - Telecom infrastructure: - £3B worth + superfast optic fibre network - 4G network - Phone line + TV network (broadband) ￡520m in R&D 102 patent fields £14.1bn spent with suppliers Online service (company website & My BT app) Call Centres Account managers - Retail Telecom Services - Retail IT Services - Business Services Solutions - Provide Telecom Infrastructure - Network maintenance - Marketing & Sales Revenue: £24.062 B (-5.26% drop from 2014/2017) Communication Providers (CPs) Media Companies and Broadcasters Businesses (SMEs, MNCs) Governments Public Institutions General Public (households) Contractual based Call Centres BT Chat Shift towards self- service for BT’s Consumer business division - Newness - Improved Performance - Accessibility (e.g: 4G coverage over 80% of UK ) - Convenience/ usability
- 8. Multinational companies Small and medium-sized enterprises Large enterprises and public sector - Voice and data - Mobility - IT services Products and services in: - Fixed voice - Broadband - Mobile - TV and broadcasting Services in: - Fixed voice - Cloud - Mobile - Networking - IT Households Network and infrastructure servicesOpenreach – nationwide Infrastructure: Copper and Fibre connections Communications providers and media companies Network products and services - Data transmission services
- 10. Our Competitors National (direct competitors) Worldwide (indirect comparable competitors) Mobile carriers Vodafone Group plc O2 GiffGaff Three China Telecom Deutsche Telekom Telefónica Verizon Worldwide IT solutions Kcom Vodafone Group plc Deutsche Telekom AG Cisco Softbank Dell Cloud Intel Internet services Kcom Eurocoms O2 Three UK Vodafone Group plc Sky UK Virgin Media China Telecom Deutsche Telekom AG Verizon SoftBank Group Corp Cisco Sprint Broadband Sky UK Virgin Media Eurocoms TalkTalk Group O2 Three UK Vodafone Group plc China Telecom Telefonica Europe AT&T Inc Verizon SoftBank Group Corp America Movil Movistar Sprint Division Companies
- 11. Understanding the maturity of the Telecom Industry “Telecommunications Industry at cliff’s edge” – McKinsey (2016)
- 12. Current challenges for BT ● Growth in number of contracts is slowing down (Smartphone Saturation) ● Decline in number of TV & broadband & fixed line subscriptions ● Consumers demand the latest services/technologies at a lower price, while data consumption is predicted to grow by 53% by 2020 (EY) ● Highly competitive → Price war between operators in tariffs ● High expenditure in infrastructure ○ BT is still expanding its 4G network (15% in the UK yet to be covered) ○ Estimated cost to implement 5G in the EU → €57B by 2020 WILL DATA & ANALYTICS HAVE ANY IMPACT ON THESE? WHAT KEY APPLICATIONS DO DATA & ANALYTICS HAVE?
- 13. Competitors and their use of Data Analytics News/Blog database: 1619 stories
- 14. News/Blogs Database: News article network with 1109 stories Quid BT & Big Data Network
- 15. Timeline of Predominance of AI potential and Cloud in media in regards to BT News/Blogs Database: News article bar chart with 962 stories BT implemented the Cloud of Clouds Announcement on future launch of CoC
- 16. Consultant sources we borrowed information from: Consultant Sources ArticlesReports KPMG (2) Deloitte (1) Gartner (1) McKinsey (7) PwC (2) Bain & Company (2) BCG (2) Forbes Insights (2) MIT Sloan (8) IBM (1) KPMG (2) Deloitte (1) Gartner (1) McKinsey (1) PwC (3) Bain & Company (2) Accenture (2) EY (2) Forbes Insights (2) MIT Sloan (1) IBM (3)
- 17. Application Clusters Network Operations Service Customer Management & Relations Business Assurance Optimisation of network utilization (21) Network monitoring and protection (20) Network capacity planning (9) Anomaly and detection troubleshooting (1) Total : 51 Product development & tariff optimization (19) Service personalization (5) New product rollout visibility (2) Subscriber profiling and segmentation (4) Total : 30 Campaign management & precision marketing (42) Customer loyalty management (6) New Customer Experience (11) Existing customer information/ data (2) Churn Prevention (15) Total : 76 Fraud detection/ management (21) Suspicious traffic management (14) Cloud Utilization (13) Real time reports (6) Forecasting and planning (2) Credit scoring system (1) Cost reduction (4) Enhance decision-making process (9) Total : 70
- 18. How are Data & Analytics going to change our... 1. Key Products & Services 1. Key Decisions 1. Key Processes
- 19. 1. Key Products & Services: Multinational companies Small and medium-sized enterprises Large enterprises and public sector - Voice and data - Mobility - IT services Products and services in: - Fixed voice - Broadband - Mobile - TV and broadcasting Services in: - Fixed voice - Cloud - Mobile - Networking - IT Households Network and infrastructure services Openreach – nationwide Infrastructure: Copper and Fibre connections Communications providers and media companies Network products and services - Data transmission services
- 20. a. Households Products and services in: - Fixed voice - Broadband - Mobile - TV and broadcasting Households
- 23. a. Households Enriching Customer Experience: 2014 MTS Telecom – By using customer analytics they implemented a system of user gratification based on their loyalty and plan usage (‘customer delight’) → Doubled their conversion rate, from 3% to 6% ● Prescriptive Analytics (Internal) Virgin Mobile – Used predictive analytics to improve retention by targeting at-risk/new subscribers → Increased their data usage by 19% and SMS utilisation by 41% ● Descriptive Analytics (Internal + External) BT –Reducing faults and predicting churn at EE → Churn reduced from 2.1% (2014) to 1.1% (2015) ● Predictive Analytics (Internal Data) 2012 T-Mobile – Data-based customer care in order to fight churn → Reduced net customer losses from 99,000 to 50,000 the quarter after implementing ● Descriptive Analytics (Internal Data) TalkTalk – By using an analytics and reporting platform for improved customer experience, they reduced number of calls to the customer service centre by 40%, (2011–2012) ● Predictive Analytics (Internal Data) BT – By using cross-channel interaction with their customers and by tracking them in those channels (360º view) → Reduced complaints by 50% and dissatisfaction contracts reduced from 28% to 12% from 2011 to 2014 ● Descriptive Analytics (Internal+External) 2022 BT – Prescriptive Analytics (External + Internal Data) + AI + Outside Insight to enrich customer experience, improve customer relationship and deliver increased value as a service wrapped around their products ● £30M Investment in Belfast R&D AI, Analytics and IoT Centre Verizon – Customer service with Cognitive Intelligence and personalised marketing → Customer satisfaction went up from 81% to 87% (2016 to 2017) ● Predictive Analytics (Internal + External) + AI T-Mobile – By leveraging social media data together with historical and transaction data from the Customer Relationship Management and billing system, they are able to better predict customer defections and react on time → In a single quarter (Q4 2016) they cut down the churn rate by 50% ● Predictive Analytics (External + Internal) + AI BT – Real-time data analytics and customer behavior analysis in BT Mobile together with fault reductions → Churn rate reduction 2.8% (2016) to 1.9% (2017) ● Predictive Analytics (Internal Data) 2017
- 24. Key Characteristics of Applications (.a): 2012 2014 2017 2022 Technology Data Analytics Data Analytics Data Analytics + AI Data Analytics + AI + Automation Type(s) of Analytics Descriptive Introduction of Predictive Predictive Prescriptive Data Source(s) Internal Internal Internal + External Internal + External + OI Cluster Churn prevention Precision marketing & Churn prevention Precision marketing & Churn prevention Precision marketing & Churn prevention
- 25. b. Businesses & Public Institutions Multinational companies Small and medium-sized enterprises Large enterprises and public sector - Voice and data - Mobility - IT services Services in: - Fixed voice - Cloud - Mobile - Networking - IT Network and infrastructure services
- 27. Cloud service improvement: b. Businesses & Public Institutions 2017 Hewlett-Packard - Design, deploy and implement data analytics consulting solutions → Leading to an application improvement in response time by 33% for RI-solution ● Prescriptive analytics (internal & external) Telefonica - It has a simple web services interface (API) that can be used to store and retrieve any amount of data, at any time, from anywhere on the web → Resulted in a high return in investment ● Predictive analytics (internal & external) BT - Becomes the first global network services provider to offer direct access to the Oracle cloud → Cutting the typical dedicated connection deployment time from months to days ● Cloud computing (internal & external) 2014 Vodafone x Argyle Data - combat fraud → $4.5m additional funding to help companies detect and combat fraud through real-time analytics which can be accessed as soon as data enter the system ● Predictive analytics (internal) Cloudera - Provide an open source big data analytics framework (Apache Spark) to support in-memory processing → Perform significantly 10 - 100x faster than MapReduce ● Prescriptive analytics (internal) BT - Announces the “Cloud of Clouds”, the first ever integration between different cloud providers, third party services, analytics solutions, customer data centres and customer users together with the BT Network. 2012 Dell Cloud - Customization of cloud deployment (build, manage, use, consult) → They cut deployment time by 80% compared to Cisco; cut operating cost; respond faster to business demand ● Predictive analytics (internal) Hewlett-Packard - Speeds the deployment of application-based services → They decreased spam by 70% and increased administrator efficiency up to 33% ● Descriptive analytics (internal & external) + Cloud computing BT - Expansion of data centre capacity in the Benelux with a new facility in Rotterdam which supported Microsoft Lync and Cisco’s Hosted Collaboration Solution (HCS) → Increased IT services capacity in the Benelux region by 50% ● Predictive analytics (internal) 2022 BT is currently one of the leaders in Cloud services (“Cloud of Clouds”). Nevertheless, they don’t have a Data Lake that could allow them to systematically analyse their customers’ data. Together with external data, and Outside Insights, they could potentially offer a better service. In the future, they could also include IoT data analysis. - The data lakes market size is expected to grow from USD 2.53 Billion in 2016 to USD 8.81 Billion by 2021, at a Compound Annual Growth Rate (CAGR) of 28.3% during the forecast period.
- 28. Key Characteristics of Applications (.b): 2012 2014 2017 2022 Technology Data Analytics + Cloud Computing Data Analytics + Cloud Computing Data Analytics + Cloud Computing Data Analytics + Cloud Computing + AI Type(s) of Analytics Predictive Predictive & Prescriptive Predictive & Prescriptive Prescriptive Data Source(s) Internal Internal Internal + External Internal + External + OI Cluster Cloud Utilization Cloud Utilization Cloud Utilization Cloud Utilization + Real time analytics
- 29. 2012 2014 2017 2022 Technology Data Analytics + Cloud Computing Data Analytics + Cloud Computing Data Analytics + Cloud Computing Data Analytics + Cloud Computing + AI Type(s) of Analytics Predictive Predictive & Prescriptive Predictive & Prescriptive Prescriptive Data Source(s) Internal Internal Internal + External Internal + External Data Variety Structured Structured Structured + Unstructured Structured + Unstructured Data Velocity Near Real-time Real-time Real-time Real-time Cluster Cloud Utilization Cloud Utilization Cloud Utilization Cloud Utilization + Real time analytics
- 30. c. Telecommunication Companies Openreach – nationwide Infrastructure: Copper and Fibre connections Communications providers and media companies Network products and services - Data transmission services
- 31. Sprint – Real-time data analytics to control the network driving → Increased 90% of its capacity ● Predictive analytics (Internal) + AI BT – Real time threat monitoring ● Descriptive analytics (Internal+external) + AI c. Telecommunication Companies Infrastructure Improvement: ! 2014 2012 2020 BT - Will use IoT data analytics and security technologies to do 5G testing for future implementation and will develop the network with Nokia and Huawei→ It is expected from BT to provide improved ultrafast speeds and latency in the range of one millisecond. ● Prescriptive analytics (External + Internal data) Cablelabs-proactive network maintenance (PNM) ● Descriptive analytics (Internal+external) Amazon – Integrates machine learning to transactional and operational data centres → To extract meaning and forecasts out of their data now no longer need third-party machine learning platforms. ● Predictive analytics (Internal) + Machine Learning BT- Implemented Cloudera-powered enterprise data hub to accelerate data velocity → Process five times more customer data and achieve a velocity increase of 15%. ● Descriptive (Internal) 2017 Vodafone – Uses AI to increase network optimisation speed by over 45,000%. Tested in C-SON, the Initial results confirmed an average 6% improvement in the mobile download speed and lower interference at the cell sites. ● Prescriptive analytics (Internal) + AI + Machine learning Cisco – By using Global Mobile Data Traffic Forecast, an ongoing initiative to track global networks → Mobile data traffic will grow at a compound annual growth rate (CAGR) of 47 percent, reaching 49.0 exabytes per month by 2021, and with the implementation they will be able to cope with it. ● Predictive analytics (External + Internal data) BT Openreach – By using data transmission technology and FTTP, BT developed the world’s first ever live demonstration of 100Gbps ‘hyperfast’ broadband with Huawei → Boost the broadband signal with enough capacity to stream 4,000 ultra HD quality movies simultaneously ● Predictive analytics (External +Internal data)
- 32. Key Characteristics of Applications (.c): 2012 2014 2017 2020 Technology Data Analytics + AI Data Analytics + Machine Learning Data Analytics + Machine Learning + AI Data Analytics + Machine Learning + AI Type(s) of Analytics Descriptive & Predictive Descriptive Predictive Prescriptive Data Source(s) Internal Mainly internal Internal + External Internal + External Data Velocity Real-time Real-time & Near Real-time Real-time Real-time Data Variety Structured Structured + Unstructured Structured + Unstructured Structured + Unstructured Cluster Suspicious traffic management Optimisation of network utilization Optimisation of network utilization Optimisation of network utilization + Network monitoring and protection
- 33. 2. Key Decisions: Basis
- 34. Company - wide ● 2: Shrink an existing business ● 3 : Grow an existing business ● 4: Corporate restructuring Develop products ● 5: Enter new industry/ start a new business Market and sell ● 6: Negotiating a major contract Deliver & Support ● 1: Collaborate with competitors ● 7: Major business investment ● 8: Corporate financing Decision architecture One-off decisions Ongoing decisions 6 5 7 32 41 8
- 35. BT Key Decisions Company-wide: Reconstructing/adapting business model/culture (ongoing) Develop Products: - New launches for product/services or reductions for existing ones (ongoing) - Infrastructure upgrades or maintenance (ongoing) Market and Sell: - Advertising (ongoing) - Maintaining and gaining new customer relationships and segments (ongoing) Deliver: - Mergers and acquisitions (once off) - Development of new or existing investments (ongoing) - Collaboration - ventures (ongoing) Support: - Security/ fraud management (once off) - Obtaining financing (once off)
- 36. 2. Key Decisions 2017 2014 2012 Analytics: Predictive and Prescriptive Database: Internal and External Technology: AI, machine algorithms, data analytics. In addition, Outside Insight may be used to augment decision-making: AI Vodafone - Simulate the expected effects of possible new pricing packages or services, and use the results to decide “when and where is best to roll them out.” ● Predictive Analytics (Internal data) Celcom Axiata Bhd. - Used customer experience analytics and performance indicators data to make decisions surrounding campaign launches and reduced campaign launch time by 80% ● Diagnostic Analytics (Internal data) BT – Continues using the cloud-based business intelligence system → Allowing agile decision making ● Descriptive Analytics (Internal Data) Telefonica O2 – Used analytics for decision-support. In specific, to support corporate planning and management decisions. This led to an 18% increase in precision of forecasts and an increase of 200% in forecast modelling speed. ● Predictive Analytics (Internal Data) Aruba – Used analytics and cloud to generate automated reports based on internal data, ultimately reducing decision-making time ● Descriptive Analytics (Internal Data) + Automation BT – Used a cloud-based business intelligence system to glean information and unify reporting from 5000 Salesforce users across 3 BT divisions to accelerate (saved time = 93%) and improve the accuracy of their decision-making process. ● Descriptive Analytics (Internal Data) Verizon – Combining artificial intelligence with human judgement for key decision-making → “Using AI to illuminate choices but not to abdicate authority.” ● Prescriptive Analytics (Internal Data) + AI OTE Cosmote - Uses analytics to analyse growing volumes of network and customer data in seconds instead of minutes, this accelerates the decision-making process. ● Prescriptive Analytics (Internal Data) + Big data BT - They are using Big Data to reveal key business insights, highlight key trends and needs that will guide the direction and speed of new product development → Allowing them to make smarter and faster decisions ● Advanced prescriptive analytics (internal data) → + Big data - Structured & unstructured data 2020
- 37. Characteristics of Applications in Decision-making: 2012 2014 2017 2020 Technology Data Analytics Data Analytics + AI Big Data + AI Machine algorithms + AI + Big Data Type(s) of Analytics Descriptive Predictive + Descriptive + Diagnostic Mainly Prescriptive Predictive + Prescriptive Data Source(s) Internal Mainly Internal Internal Internal + External Actual decision made by Human Human Machine and Human Machine and Human Main Cluster - Forecasting and planning - Real-time reports - Cloud Utilization - Campaign management & Precision marketing, - Existing customer information/ data - Cloud Utilization - Existing customer information/ data - Cost reduction - enhance decision- making process -
- 39. Source: PwC, 2016 Future Decision Making “The combination of data analytics and human intuition adds up to judgment that is more capable and effective.” - PWC, 2016
- 40. Talk Talk – Natural Language Call Steering (routes calls to the correct division in Talk Talk). They were able to save £3 million and reduce call times by 23% ● AI + Automation + Descriptive Analytics (Internal Data) Vodafone – Uses fraud analytics to detect fraud. Saved 99% time (from 24 hours to 3.5 minutes) spent. ● Prescriptive Analytics (Internal Data) BT – Debatescape - social monitoring tool - uses AI to search out customers who might need help but don't know who to approach in BT to get it. Saved ￡2 million. ● Big data (social media) & Predictive Analytics (Internal Data) + AI Vodafone – Implemented an IBM Business Intelligence (BI) platform used by top management for querying and KPI reporting → Saved £105k the first year after implementation. ● Descriptive Analytics (Internal Data) O2 (Telefonica) – Automation Blue Prism software applied to reduce back office operation costs and remove reliance on offshore activities to improve workload management during peak hours → reduced offshore costs which for equate to ¼ of onshore costs. ● Descriptive Analytics (Internal Data) + Automation BT – Used Business Intelligence (BI) software (Oracle BI), able to cut down from 8,000 systems to 4,000, consolidating and homogenising the company’s technology. Allowed a clearer view of processes leading to an increase of Key Performance Indicators (KPI). ● Prescriptive Analytics (Internal Data) Rostelecom - SAS credit scoring software used to automate credit management process. It reduced international operators’ bad debt reserves by US$3.8 million, achieved by reducing the amount of funds the company should set aside to cover risky debts. ● Descriptive, Predictive & Prescriptive Analytics (External + Internal Data) + Automation CenturyLink - Implemented Conversica AI powered assistant (Angie), used in sales automation to help sales reps and save costs of hiring. Around 90,000 sales leads sent to company every quarter. Angie sends ≈ 30,000 emails a month, analyses responses, identifies and sends most promising leads to the right reps. Earnt $20 in new contracts for every dollar invested in the software. ● Descriptive & Predictive Analytics (Internal Data) + AI BT - Truecall technology blacklists nuisance numbers (unwanted calls). Real time analysis of data from radio access networks (RAN) used to identify rouge numbers (large numbers dialled in quick succession). Aim to block 25 million calls /week. ● Descriptive, Predictive & Prescriptive Analytics (Internal data) BT – Analysing constant process improvements by using internal and external data and AI algorithms. More real-time analytics and virtual assistants are expected to help in areas such as fraud detection, cost reduction and real-time reports. ● Prescriptive Analytics (External + Internal Data, OI) + AI & Algorithms 3. Key Processes
- 41. Characteristics of Applications in Key Processes: 2012 2014 2017 2022 Technology Data Analytics + BI Data Analytics + AI + Automation Data Analytics + Automation Data Analytics + AI Type(s) of Analytics Mainly Descriptive Descriptive Predictive Prescriptive Mainly Descriptive & Predictive Prescriptive Data Source(s) Internal Internal Internal Internal & External Main Cluster Optimisation of network utilization Optimisation of network utilization, Fraud & Anomaly detection Product development & optimisation of tariffs -
- 42. Example of a Key Process (Network Capacity Planning):
- 43. Yesterday (until 5pm), BT had 11 cases of Broadband Outage across the UK WHY? WHY? WHY? WHY? WHY? We extensively rely on our internal database, which is why we are unable to anticipate through Big Data. We currently don’t make any use of external data (IoT analytics nor Outside Insight) in our predictive model. Our predictive model is not as sophisticated as it should to be. Our estimations are not accurate enough. Our chase demand does not match the actual demand.
- 44. * Full details on value creation will be provided in the report (along with the impacts on processes, operations and value propositions for customers) * *
- 45. BT could eventually regain its competitive edge whilst: ● Enriching customer service ● Better network utilisation and service improvement ● Making smart decisions faster ● Potentially reducing operation costs An additional Recommendation Internal + External data (OI)
- 46. Thank you