Ventiv Technology

HARNESSING THE WEB The Aon Data Centre had the task of consolidating information from 400 separate databases and applications onto one platform. Aon moved from the 'tins and wires' of traditional databases to online cloud and amazon based servers. Data became more accessible to customers, employees and key stakeholders as all information was held on the same infrastructure, ensuring common and consistent messaging across all parties. The web-based platform allowed use of online computation models and reduced costs as storage capacity can be turned off and on to meet the peaks of data collection across the business. DIFFICULTY ACCESSING DATA The top reported reason for risk managers not using analytics is inability to access the required data. Firstly, they may struggle against business units limiting the information they share with the risk function or sanitising it beyond use. Even where data is accessible many businesses fall foul of collecting data for single-purpose or single-function use. Data becomes locked within silos where each set has its own classification and taxology rules. A risk manager's first task will be bringing together the different pools and linking them by common themes and field names. This facilitates analysis and ensures that when the business is discussing its information, everyone is talking the same language. ACHIEVING CONSISTENCY A transport provider is developing consistency in risk reporting across the business by taking a top-down and bottom- up approach. The risk manager gained senior management support by collaborating to theme or 'bucket' the key risks of the business against publicised strategic priorities. This is creating risk data categories that should resonate at all levels and across all functions. The risk manager is subsequently producing key risk indicators and scorecards linking the main activities of each unit to the established themes. By linking risk data to the overall priorities and operational activities, consistent data collection will be built into all day-to-day processes. POOR QUALITY DATA A quarter of risk managers think the data they hold is of limited quality. Data needs to be of high quality in terms of the data producer, the person or system that inputs/ generates the data and the data user, who will view or manipulate the data. High quality data will be accurate, timely and appropriate in nature and volume to its use. Appropriate data verification and access checks will be key. Risk managers must establish who collected the data, and why. Data is often collected with a specific theory in mind and there can be a tendency to make assumptions based on limited information. Risk managers should assess the data collected in terms of its relevance to the question being asked, and reflect on any confirmation bias that may have crept in. 5 The barriers to data use DRIVING THE DATA DIVIDE | 6

• Cover