Unleashing data to make better customer and business
Unleashing data to make better customer and business decisions within Sky Avengers: Endgame
Agenda Discussion today: • Overview of Sky & NOW TV • Our Data Journey & Use Cases • Top Tips. . Key things to leave the room with: 1 A sense of how data is used in the working world 2 An understanding of what to think about in any future career 2
My Journey so far… Insight Director Credit Analyst Data Intelligence Director Head of Insight – Viewing Capability Career Goals!! Lead Analyst Maths Degree 3
Overview 4
Our Purpose How NOWTV articulates this To create bright new ways to wow and entertain our customers so they never miss out
Our proposition 1 Pick & choose your pass We offer à la carte (no contract) subscription passes and transactional sports passes 2 Watch on 60+ devices Broad distribution allows our customers to watch wherever they choose 3 Mass scale Our OTT footprint spans 5 territories building material scale 5 Core Passes Partner Pass Transactional Sports Passes Territories
What are we trying to achieve? 7
Data Intelligence Purpose is to Unleash the Power of Data to Make More Intelligent Business & Customer Decisions Understand Predict, Optimise & Act Govern & Improve Insight Data Journeys & Personalisation Data Governance & Business Enablement Empowering OTT Businesses to make Bigger, Bolder & more Intelligent Business Decision by putting Insight at the Heart and Information at your Fingertips To use Consumer Data to provide a tailored experience across all touchpoints with relevant messaging and content at the right time through the right channels Enabling OTT Businesses to use data securely and seamlessly, connecting the dots through alignment of data expertise and strategic partnerships to compliantly deliver business value Operate Excellently Strategic Data Platform Create a robust, connected & secure data environment with fast feedback loops seamlessly supporting OTT use both flexibly in market & efficiently for core products 8
Research Execs Audience Planners Communications Execs Insight Analysts Data Business Analysts Service Manager Optimisation Analysts “Uber” Technical Commercial We have different roles to enable us to deliver this Data Scientists Data Governance & Quality Data Engineers 9
• Cancellation Journeys • Purchase Experience • Offer Management & Xsell Marketing • Navigation Experience • Content recs • Ordering of rails • Channel Management • Contact Strategy • Promos Service Experience Product Experience • Content attribution • Brand affinity • Headroom & Pricing Customer Experience Proposition @NOW TV we talk about initiatives that impact every area of the business to help create more meaningful relationships with customers • Concierge: Contact Handling • Value through Service • Warning signs Greater engagement Better value for money perception Lower churn More efficient and effective allocation of resources Higher EBIDA
Storage Central. Activation Brain Channels To put that in perspective. . We have a multitude of channels to talk to our customers through Customer Journeys Communication channels Email Paid Social Push SMS Pay Per Click Canx Journey Shop Help My Acct Adobe Marketing Cloud (Campaign & Audience Manager) Customer Decisioning Library Content Library Trading Engagement Marketing In Product Trailers Homepage Banners In App Segmentation Engine Insight Marts (Measurement) Service Attitudes
There are multiple ways we enhance the customer journey 12
As we do this we assess different ways to get to the best solution Traditional approach “Out of box thinking” approach Ents: Churn Movies: Churn Sport : Churn Ents: X-sell Movies: X-sell Ents: X-sell Sports Ents: X-sell BB Ents: Cancellation Ents: Price elasticity Movies: X-sell Sports Movies: X-sell BB Movies: Cancellation Movies: Price elasticity Sport: X-sell Ents Sport: X-sell Movies Sport: X-sell BB Sport: Cancellation Sport: Price elasticity Ents: Winback Movies: Winback Sport: Winback Ents: Lapse Movies: Lapse Sport: Lapse 1. Deep learning image based Sports NBA optimisation Sport: Acquisition Movies NBA optimisation Movies: Acquisition Ents NBA optimisation Ents: Acquisition 2. Sequential modelling 3. Reinforcement Learning Channel model Timing model Message type model Next-Best-Action 4. Personalised segmentation with our own brains Typical problems: • Time consuming to build • Propensity not uplifts models • Cumbersome to maintain • More channels and proposition makes it even more complex 13
What are the typical projects you can work on? Foundations Exploratory Analysis Data Transformation Forecasting / Headroom Model Maintenance Build & Optimise Act Bespoke offers in cancellation journey Strong Measurement Prototypes Next Best Action Business Simulation Content Recommendati ons in comms Sales through service with explainable ML
Top Tips 15
What Makes a Great “Data Intelligencer” Get Up and Go Technical Excellence Great Leadership Commercial Savviness Strategic Thinking Source: Group Finance, June 2018
What Makes a Great “Data Intelligencer” Adaptable to any situation • • • Breadth of experience Objective thinking Trusted Advisor with business Understand apply the art of the possible • • In depth / expert Understand & apply range of techniques Deliver high quality, accurately & reliably knowledge of one Innovative approach to problem solving specialism Land business impact • • • Know your onions (business levers) • • Explain Simply & Clearly • Tell a Story Motivated & Productive team • • • Develop others Clear Planning & area ownership Engaging Purpose What Next? • • • Challenge current beliefs • Anticipate Change Cross –Functional view • Quick Decisions 2 year+ planning • • • Time Management / multitasking Line management skills Drive change Fit for purpose output Know your audience Priority management Positivity Influencing other teams
Top Tips To help you transition to business world Perfection doesn’t exist! • Both in the data and in the use cases - Use the 80/20 rule to drive value Occams Razor – The simplest way is most likely correct • Don’t over complicate your approach – the best model is still useless if not understood & deployed properly Mindset is as important as any qualification • Think of the art of the possible and how you can achieve it – seek to understand the business objective Be Kind to yourself. . • You can’t achieve everything all at once 18
- Slides: 18