This model has helped Tesco generate annual aggregate gross
Transcription
This model has helped Tesco generate annual aggregate gross
Data Science Consulting Monétisation des données Big Data Paris 2015 Julien Cabot, Directeur La Monétisation des données? Générer un revenu, direct et/ou indirect, par l’exploitation des données, comme actifs de l’organisation © Quantmetry 2015 | Salon Big Data Paris 2015 2 Quatre Business Models clefs de la Monétisation des données Amélioration des processus internes… Activité dédiée… Performance Marketing et Commerciale Performance Opérationnelle et Financière Commercialisation de Données brutes, enrichies ou indicateurs Commercialisation de Produits & Services Intelligents SFR, AXA, SG, Solocal, Amazon, Tesco, … Bouygues Telecom, Orange, EDF, Toyota, Amazon, UPS, … Météo France, GERS, BIEN, Médiamétrie, Bloomberg, Twitter, Tesco, … 1000Mercis, AXA Drive, Google, SFR Geomarketing, Equifax, Intuit, … Revenu d’amélioration © Quantmetry 2015 | Salon Big Data Paris 2015 Revenu de croissance 3 De combien parle-t-on? “Tesco uses a data ecosystem to monetize data from its loyalty program, the Tesco ClubCard. This model has helped Tesco generate annual aggregate gross billings of US$500 million globally.” Source : Forbes, Janvier 2014 “For financial services firms, the opportunity to monetize customer and transaction is new and compelling – with revenues of US$175-300 billion per year.” Source : Strategy& PwC, formely Booz&Company, Juin 2013 © Quantmetry 2015 | Salon Big Data Paris 2015 4 Quel raffinage des données? Niveaux de raffinage des données Typologies de données Usages 5 Service Données « intelligentes » Processus adaptatif, règles de gestion adaptative, … 4 Prédiction Données prédites, probabilités Anticipation, recommandation, alerte préventive, … 3 Indicateur Données agrégées, calculées KPI, scores, alerte, dashboard, infographie, … 2 Caractéristique Données validées, complétées, nettoyées Analyse détaillée, gestion, reporting, … 1 Brute Données brutes de logs, capteurs, formulaires, … Audit, détection d’incidents, … © Quantmetry 2015 | Salon Big Data Paris 2015 5 Les constituants de la Valeur des données Utilisabilité Temps Précision Exhaustivité Rareté Les données sont-elles utilisables, compréhensibles, utiles pour mes besoins? La fraicheur des données permettentelles de réagir rapidement et/ou efficacement? La granularité et la significativité des données est-elle adaptée à mes besoins? L’exhaustivité et la complétude des données correspondent-elles à mes besoins? La donnée est-elle facilement accessible et/ou largement diffusée? Structuration de données, Infographies … Remarketing en temps réel, attrition long terme, anticipation d’incident, … Ciblage comportemental, Lutte contre la fraude, personnalisation de l’expérience client, … Benchmark, indice de référence, … Avantage concurrentiel, exclusivité, … © Quantmetry 2015 | Salon Big Data Paris 2015 6 Quels prix pour les données? Tesco Bank, +3% Market Share, +39% Car insurance Premiums, +44% Pet Insurance Premiums Valeur complète des données de Tesco ClubCard pour Tesco Bank Source : Strategy&, PwC U.K.-based supermarket giant Tesco is a prime example of a nonfinancial company that’s using data to compete effectively with traditional financial players. Until 2008, the company ran Tesco Bank as a 50/50 joint venture with the Royal Bank of Scotland. That year Tesco bought out RBS and began developing a completely new infrastructure for the business, built a new team, and brought in new expertise. The transition was not always smooth — for instance, online customers were locked out of accounts for several days in 2011 when Tesco moved data from the RBS systems to its own — but it’s now complete. To fully exploit this treasure trove of data, the company took a significant stake in Dunnhumby, a U.K. data mining firm that will help Tesco monetize the consumer data from both the retail and banking operations. At its core, Tesco Bank is underpinned by the Clubcard. The insights the bank gains from the Clubcard customer data allow the company to understand customer needs and make the most relevant offers in the store and in the bank. The Clubcard credit card rewards customers with points whenever they use their card — one Clubcard point for every £4 (US$6.12) spent. Clubcard customers can also receive preferential deals when buying Tesco Bank products — including discounts on car, home, pet, and travel insurance — and can use points to buy Tesco Bank insurance. This year, Tesco Bank gave customers around £70 million (US$107 million) worth of points to spend in the store or on Clubcard rewards. In terms of systems and IT, Tesco’s new platforms significantly improve customer service. Instant decisions are now possible on loan applications, and customers can open and fund savings accounts in just 10 minutes rather than the two weeks required in the past. The conversion is still in its early days, but Tesco’s efforts are paying dividends in the form of increased market share across a range of products. In 2009, Tesco Bank credit cards made up 9 percent of all MasterCard and Visa credit card transactions in the U.K., and by 2012 that figure had grown to 12 percent. Meanwhile, from 2008 to 2012, the company’s car insurance gross written premiums increased by 39 percent and pet insurance gross written premiums rose 44 percent. © Quantmetry 2015 | Salon Big Data Paris 2015 Valeur Comptable Coût de production des données pour Tesco Clubcard Valeur d’Usage Revenu incrémental estimé (« UpLift ») lié à l’utilisation des données par Tesco Bank Valeur de Marché Prix des données Prix auquel Tesco ClubCard vend ses données à des tiers 7 Quels moyens pour monétiser les données? Types de solutions Exemples 5 Service Web Service Predictive Business Process Data Management Platform, Data Exchange, RTB, … 4 Prédiction Modèle prédictif Data Science Studio, OpenScoring, RTD engines, Model as Code, … 3 Indicateur Data Mart, Data Visualization SGBD, InMemory, BI, DataViz Tool, … 2 Caractéristique Data Lab Hadoop, Datawarehouse MPP, ETL, BI, … 1 Brute Data Lake Hadoop, NoSQL, Stream Processing, … © Quantmetry 2015 | Salon Big Data Paris 2015 8 Conclusion 5 Etapes essentielles à la Monétisation des Données Par Chris Twogood, Teradata 5 4 3 2 1 Start with Questions Look for Data Patterns © Quantmetry 2015 | Salon Big Data Paris 2015 Search for External Data Sharpen Your Analytics Skills Understand Your Data Monetization Identity 9 Merci!!! Julien Cabot [email protected] 06 64 45 53 73 10