Unlocking Value of Data in a Digital Age
1 like1,078 views
The document discusses the evolution of digital organizations driven by big data and the necessity for businesses to innovate in the face of market changes. It emphasizes the role of intelligence hubs in supporting data-driven decision-making while addressing the challenges companies face in utilizing data effectively. Additionally, it highlights the growth of the online retail market and the importance of adapting business models to maintain competitive advantages amid increasing consumer expectations and security concerns.
Unlocking Value of Data in a Digital Age
- 1. Intelligence hubs as accelerator of the digital organisation smartphone went on to purchased on a smartphone Forecast of European online retail sales in 2017 by country in billion Euro's Challenges UNLOCKING VALUE FROM DATA MEANS THINKING BIG – WHILE ACTING SMALL. Understanding the dynamics. Implement through a stepwise approach: • Acting small to make ‘fact-based decisions’ essential • Thinking big: Intelligence hubs as accelerator the digital organization • Acting big to make it stick OVER THE LAST DECADE, DATA HAS BEEN INCREASING IN VOLUME AND DIVERSITY – and the speed of production continues to grow. 62 Central Europe Middle East and Africa MAKE DECISIONS BASED ON FACTS: The end of assumption-based decision making. THE DIGITISATION OF DAY-TO-DAY LIFE IS DRIVING MARKET CHANGE. Classical business models across industries are under pressure. BUSINESSES HAVE TO INNOVATE AND SECURE THEIR BUSINESS MODEL TO RELEASE THE VALUE OF DATA. Some are more successful than others to overcome the key challenges. Non-mobile devices Mobile devices Interactions through mobile devices will grow. Today there are 7 billion connected devices in the world, in 2020 there will be 32 billion devices. Data transactions through these devices will grow at 78% a year. In 2016, there will be 12 exabyte data transactions a day - in other words 66 billion pictures sent a day! The digital customer is on the rise. The internet has redefined the consumer decision-making process. The average number of information sources used by shoppers has doubled in the last years. Customer loyalty is reducing as switching becomes easier (80% of customers will abandon a mobile site following a bad user experience). Online sales are growing at 11% per year in Europe. The European internet economy is already 3.5% of total GDP - this will double by 2016, and triple by 2020. There has been an increase of 18% in the number of retailers going bankrupt over the last three years as they have not adapted their business model in time. In businesses, only 5% of data is used, while There are several challenges that businesses face in this process of change: Global challenges. Security: of people have no idea who holds their personal data. The number of privacy injunction applications more than doubled in the last year. People have little trust in the way companies handle their personal data and what they use it for. Online crime: 2.7 billion people use the internet today and this is expected to grow to 4 billion by 2017. As a result, online crime is growing. Organisational challenges. Value: “I want to do big data, but don’t know how to apply any of the outcomes” Capability: professional level” Governance: “Data management and governance processes aren’t defined and are fragmented throughout the organisation” Technology: “We have a fragmented landscape of applications and warehouses” Sweden $ 6.5 Germany The digital organisation is on the rise. Mass customisation has become the norm, boosting flexible production and changing the traditional organisation. Next to that, the value chain is increasingly digitised to support flexible production, personalised service and products. The big data market has exploded. A consequence of the digitisation and the growing amount of data, is the growth of the big data industry. The big data technology and services market will grow 27% per year to $32 billion, and that the internet of things will generate 30 billion autonomously-connected endpoints in 2017. In a connected world, emerging economies will drive further data creation. of data today is generated by the West, other parts of the world are emerging. So is Asia expected to be the biggest data producer by 2019. The internet of things will generate 30 billion autonomously-connected endpoints in 2017. Total data volume will increase exponentially. The total amount of data will grow from 5 zettabytes - to nearly 45 zettabytes, that is the equivalent of 62 billion iPhones. Talent matters as much as technology routinely run experiments to test the impact of changes to things like marketing strategies and recommendation systems. Amazon is able to monitor the impact of tiny changes, such as a of making the change,” says Mr Wiengend of the Social Data Lab. MANAGE THE PENDULUM EFFECT BETWEEN STRICT SECURITY AND MASSIVE OPPORTUNITIES Accompanying the digital revolution are multiple security and privacy concerns. Cyber criminals in search of financial gain (representing 60% of cyber crime) and intellectual property spies (about 25%) give cause for concern. Companies have to be aware of the security, moral and legal choices they make regarding data protection. Digital shoppers are on the rise and represent a massive opportunity Despite the risks accompanying the increase in online data, there are also opportunities: • Reputation can be positively influenced by a security strategy – when a large bank openly informed their customers about phishing emails, positive sentiment increased. • Customer satisfaction increases when companies react quickly to public opinion. 95% of popular brands have a webcare strategy, including service delivery. Successful companies have response times varying between zero and two hours, while the average response time in the Netherlands is 15 hours. Mobile phones Profit increase 27.5% Fraud detection 80% Security Marketing VS. Fraud reduction 30% Business unit operations Identify a visionary to sponsor your first data insights project that can be executed in eight weeks (small) and will delver insights that exceed the investment at least five times (essential). Data insight projects follow a five step proven approach: A. Formulate hypotheses as your point of departure: • Engage resources from business, data science and consultancy, and bring them together into a team to brainstorm on hypotheses • Ensure the insights are tangible input for a realistic business case B-D. Prepare, analyse, validate your data: • Use your existing infrastructure and analytics tooling, where necessary, complemented by low-investment, open-source software, for the first data insight projects • Continuously manage and iterate scope and outcomes with business and project teams E. Manage benefits realisation with a disciplined approach: • Formulate clear next steps to capitalise your insights • Monitor the insights regularly to justify the realised benefits • Share success throughout the organisation to trigger demand We believe that a digital organisation requires an enterprise-functional approach to maximise the potential of available data. By thinking big and initiating an enterprise-wide intelligence hub, companies can speed up their journey to becoming digital organisations. Stakeholders – Who should we take into account? • Your data playing field should give you insights into who your stakeholders are (customers, regulators, shareholders or employees) • This rich set of stakeholders, and limited capacity of an intelligence hub, requires a careful prioritisation to deliver high value to all stakeholders Customer and Value - Why do you need an intelligence hub? What is the customer value the intelligence hub creates? What services do the intelligence hub oer to stakeholders? This goes beyond the value of the analysis itself . The type of service dimensions are: the speed/flexibility of the analytics and type of data that is requested by the business. Capability – What are the resources and how do we organise them? • Building the capability for intelligence hubs goes beyond recruiting data scientists and procuring software tools. • A digitally capable organisation has the following ingredients in place: data driven marketeers and managers, processes to deliver insight projects, benefit reporting platforms, data governance, agile infrastructure and a view on roles and responsibilities Financial - How do we finance and keep costs in control? Most organisations use financial triggers as a driver for change. Initiating data insight projects requires a short-term return on investment (ROI) and shoud be used to drive such experiments. But, a long-term investment is required in the near future to accelerate the transformation into a data-driven organisation. Hence, one should define a budget and cost and performance mechanism that fits the requirements to manage the pendulum between short and long-term investment requirements. Hacking Malware Social Physical Misuse Negative Neutral Positive BUILD A DATA CAPABILITY BEYOND RECRUITING DATA SCIENTISTS AND BUYING BIG DATA PLATFORMS. Building a data capability requires an integrated approach over The skills and knowledge of data scientists are precious. The number of job opportunities for data scientists are increasing: the US is expected to create around 400,000 new data science jobs, but is likely to produce only about 140,000 qualified graduates to fill them in 2015. Analytical methods have not evolved as quickly as technology. Most methods used today stem from the 1950s to 1980s. Due to innovation in technology, application of several analytical methods is now possible compared with a couple of years ago. 0,02 0,01 Risk of mis-using oversimplified analytical software. Analytical software tools make data analytics mainstream, but there is a risk of over-simplification. Programming skills are less important than before. This increases the risk of miss-use of the applications New platforms challenge the way corporations look at IT. There are an increasing number of open-source, service subscription and cloud solutions that companies can benefit from. For example: Hadoop, an open-source technology, has become mainstream. Stage gate | Deliverable Stage gate | Deliverable Stage gate | Deliverable It is our experience that digital organisations develop incrementally instead of a traditional belief of top-down design. Acting big requires finding sponsorship to initiate an ocial intelligence hub. We use four incremental steps to make intelligence hubs true accelerators of a digital organisation: Make it essential - acting small to create urgency and commitment: • Build success by promoting the success of your first client-insight projects • Identify stakeholders and support throughout the organisation for new insight projects • Build a community around the successful projects and show benefits delivered Make it ready - design the business model of the intelligence hub: • Create leadership and mobilise first movers into a virtual team • Start defining the business model of an intelligence hub • Agree on a corporate-wide data governance and strategy, balancing speed and flexibility of analysis with data quality management Make it happen - secure advanced data analytics in your organisation through the intelligence hub: • Create a roadmap that allows short-term experimenting and long-term capability • Incorporate data activities in existing processes and adopt new ways of working • Make the intelligence hub an ocial team Netezza Terradata Google Hadoop BQ Storm Spark Exploration | A successful implementation of information services will be built over time through stage gates. In every stage gate high impact business requirement will be fullfilled. Operation | Through the delivery of the stage gate deliverables, an accumulated experience is built that will be embedded in the organisation (playing field customer, capabilities and finance). Make it stick - continuously innovate while producing regular insights: • Incorporate new ways of working in your organisation’s performance management system • Continuously allow and invest in innovation and discover the unknown (allow for a lab mentality) • Create a ecosystem of clients, competitors and suppliers - understanding that ideas and innovation are born in partnership and new coalitions 1. Make decisions based on facts: the end of assumption-based decision making 2. Break through organisational silos by focusing on the client 5. Manage the between strict security and massive opportunities 3. Establish a leadership that facilitates a data-driven way of working 4. Buil a data capability beyond recruiting data scientists and buying big data platforms BREAK THROUGH ORGANISATIONAL SILOS BY FOCUSING ON THE CLIENT. ESTABLISH LEADERSHIP THAT FACILITATES A DATA-DRIVEN WAY OF WORKING. ACTING SMALL TO MAKE ‘FACT-BASED DECISIONS’ ESSENTIAL. THINKING BIG: INTELLIGENCE HUBS AS ACCELETATOR OF THE DIGITAL ORGANIZATION ACTING BIG TO MAKE THE ‘DATA CHANGE’ HAPPEN. Marketing is building brands through personalised campaigns. They are using customer data to understand their customer better and to make personal on your location or previous buying behaviour. Security is worried about the impact of personalised campaigns that may breach privacy and regulations. VS. IT Business unit operations develop their own data-related projects based on their are often not reused and can’t be used by other units. IT tends to introduce large-scale data projects that try to provide a one-size fits-all solution. They are looking for eciency gains through standardisation and predictable quality. Data provider Data user VS. The data provider is focused on local benefits of data and compliant to regulations. He experiences no incentive to realise corporate overall benefits. The data user is focused on the upside potential of data and reuses data of Integrate data governance responsibility Fewer than one in five companies have “a well-defined data management strategy that focuses resources on collecting and analysing the most valuable data. We need to balance our initiatives across our silos in the bank and meanwhile maintain the drive to use data in an agile way.” CIO of a Dutch retail bank. Balance big data investments with real business need “Many executives believe that the right technology can produce ‘magical results’. But companies should start by prioritising the challenges they want to tackle, and then build an appropriate data strategy around those objectives. You need to know what problem you want to solve.” Mr Dumbill, chair of the O’Reilly Strata Conference, a leading big data event. Job trends from Indeed.com % of data science job postings Jan ’06 0 Jan ’07 Jan ’08 Jan ’09 Jan ’10 Jan ’11 Linear regression 1950s Neural networks 1960s Decision treed 1970-1980s Temporal dierence learning 1990s Deep learning 2000s Generic use case Proven Leading edge Specific use case Mongo DB Neo4J Commercial open source Number of breaches per threat category over time 600 400 200 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Error 70% of organisations indicate that information security policies are owned at the highest organisation level 3500 3000 2500 2000 1500 1000 500 0 1 apr 10 apr 20 apr 30 apr Strongly negative Strongly positive Managing security breaches leads to positvie sentiment Security breach Proven approach to drive insights and benefits Benefits realisation Data preparation B Validate Analyse Hypotheses business case A D C E Value for the business Continuously improving information factory Playing field Customer Capabilities Finance Make it essential Make it ready Make it happen Make it stick Exabyte per Month 2011 2012 2013 2014 2015 2016 North America Latin America Asia Pacific 2019 2019 2019 2013 2013 2019 2013 2019 2013 2013 Western Europe Mobile PCs, tablets and mobile routers Billion iPhones 46%researched on a smartphone went on to purchase in store 41% researched on a 19% researched on a smartphone, visited the store and then bought by computer UK $ 170 France $ 60 Spain $ 30 Netherlands $ 12.5 Italy $ 20 $ 110 Global internet device installed base forecast 0 2004 2005 Devices 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 20,000,000 15,000,000 10,000,000 5,000,000 Wearables Smart TVs Internet Of things Tablets Smartphones Personal Computers (Desktop and Notebooks) Play 78% increase in mobile data transactions 66bn pictures sent per day Real-life example Improving conversion ratio through improved understanding of the omni-channel client journey Due to changing regulations, the mortgage services of a retail bank needed to be digitised to become more ecient. However, it was unclear whether customers could find their PA analysed the online click and transaction data in a eight week’s project, to understand the bottlenecks especially where online visitors did not convert into potential customers. With the results, the client improved the conversion ratio and decreased throughput time in mortgage process dramatically. Real-life example Securing governance and structural data insights at a large Dutch insurer We were asked to help this insurer accelerate the creation of business value f internal and external data. Through iterative projects, and within a short timeframe, the business value from available data was demonstrated through an ‘intelligence’ cockpit. Existing data sources revealed business value through clever permutations and combinations which were previously unthinkable. This is the foundation of a true digital journey for this client. Business model Playing field Stakeholders Value Who should we take Why do you need an intelligenge hub? Operating model How do we finance and keep costs in control? into account How do we create value for our customers? What are the resources and how do we organise them? Real-life example Customer Capability Financial Accelerating omni-channel A financial company in the process of transforming into a omni-channel service provider closed local branches and ramped up new online services. To accelerate this change process, additional customer insights from digital channels were essential. Hence the CCO and CIO agreed to imple ment a joint customer insights competence centre to build up the required capabilities that the existing organisation either lacks or has scattered across divisions. PA supported the creation of this intellgence hub which now successfully completes client insights projects. 1. Business creation and innovation Alternate and realtime pricing scenarios Product and service reinvention based on data and sensoring Customer relationship dynamic pricing Churn prevention Breakthrough product innovation in healthcare Store location, supply chain optimisation and operations management optimisation Stock replenishment, asset management and forecast of performance and equipment failures End-to-end omni-channel customer journey optimisation Forecasting bandwidth in response to customer behaviour Scenario analysis of the impact of tax policy and budget decisions 3. Risk control and fraud detection Fraud detection prevention Individual risk profiling Determination of level of credit exposure to particular customers Fraud prevention Identification of tax and benefit fraud Retail Manufacturing Banking Telecoms utility Government Internet users in bn 2014 4 2 0 2017 5% used data Customer intelligence: There is an average profit increase of 27.5% at selected US banks through analysing client motivation and behaviour patterns, or predicting customer churn. Data analytics: Analytical platforms for fraud detection highlight suspicious behaviour with 80% accuracy - leading to 30% reduction in fraud cost. ‘Watch our ’Digital Democracy’ video’ www.paconsulting.com/digital-insights/ ‘Watch our ’Future Role of Insurance’ video’ www.paconsulting.com/digital-insights/ European internet economy % of total GDP 2014 10.5 7 3.5 0 2016 2020 “Machine learning finally matured and now outperforms humans in most data analysis problems. I have seen the number of machine-learning students more than quadruple in the last three years because of this renewed interest in the field.” Prof. Dr. Patrick van der Smagt, Associate Professor of Biomimetic robotics and machine learning. DATA IN THE DIGITAL AGE Setting the scene Suggestions “It’s dicult to find and