WIXLIB

Big DataTougher, smarter,stronger, faster

ContentsForeword1Tougher regulation, policy and cost control2Smarter technology solutions4Stronger community engagement5Faster opportunities ahead6Summary7Contact us8References9This report has been co-authored by Justin Cheng, Chris Lewin, Tim Nugent,Rob Hillard (Enterprise Information Management, Deloitte Australia)

ForewordThe business case for Big Data analytics is improving,despite public and private companies discovering datamanagement weaknesses, uncovering skills gaps andencountering tough new regulatory challengesWhere are we now?Last year, we identified Big Data analytics as one of thekey technology trends to watch1. It certainly has beenhard to ignore. Yet for all the attention, local successstories are still few and far between.So why are Australian businesses finding it so hard tocreate value amid the hype?Our experience reveals that the reasons may be morestructural than strategic. The resilience of the Australianeconomy has dampened pressure on local businessesto dig deeper for insights and to invest in more datadriven processes. At the same time, companies havebecome more cost-conscious, anticipating a reversal ofour recent fortunes. Industry is becoming more wary ofinvestments in new and unproven technologies.The business case for Big Data analytics has sufferedmore than most. Almost certainly because its benefithas only been articulated as better insight profferedwith trust that better insight translates to businessvalue. In reality, businesses have thus far struggledto prove a definitive link between investments in BigData systems and financial performance2.Here’s the rub - things are about to change in Australia.As our economy rebalances and restructures andcompetitive pressures intensify, it is vital for all industryand government leaders to ask more searchingquestions about strategy, markets, customers andoperations3.The business case for Big Data analytics is about toreturn to the investment agenda in a very big way.Until now, a laggard approach to Big Data has been amixed blessing. Technology shifts have left first moverswith write downs on their early investments. Firms thathave pushed the boundaries on personal data use andretention have found themselves increasingly under theburden of government scrutiny.But as the tools mature, experience proliferates andaccepted practices emerge, risks to adoption forindividual firms will diminish. This can tip the balancefor Big Data projects, particularly in cases wherethe business benefits are clear and quantifiable ahead ofimplementation.The immediate focus for firms should be to invest indata management capabilities and to build data talent.These critical success factors have too easily beenshelved alongside fledgling Big Data projects.Investments in better data management and talentdevelopment offer a far better return than pickingwinning products. Not only is strong data managementa necessary condition for success with any new dataventure, but it is also transferable across point-in-timetechnologies. And while Big Data analytics continuesto hold great potential, it is unlikely to yield anysolutions that will reduce our dependence on humanmanipulation, insight and experience.Big Data Tougher, smarter, stronger, faster1

Tougher regulation, policy and cost controlFor all companies interested in Big Data, it pays to beacross new challenges appearing on the horizon. We’veidentified three key areas that we think will pose thebiggest hurdlesGovernment Regulation and PrivacyAs firms strive to retain more data and share it morewidely, they continue to be confronted by legislationand regulation. The Australian Government’s own‘Big Data Strategy’4 released last year, identified thirteenlegislative and policy controls which it had to consider.Much of this regulation was drafted more than 20 yearsago and can hardly be expected to keep pace with therequirements of modern corporate data use.Yet the legislation cannot be ignored. From March 2014,the Australian Privacy Principles (APPs)5 will replacethe National Privacy Principles and Information PrivacyPrinciples. This change will broaden the powers of theOffice of the Australian Information Commissioner(OAIC) to investigate and publicise privacy breaches,including civil penalties of 340,000 for individuals and 1.7 million for businesses6.Increased regulation will pose bigger barriers toadoption of Big Data analytics in some industries morethan others – particularly those that are required tohold and process sensitive customer information aspart of everyday operations. Conversely, industries suchas mining, where data is typically sensor-based andrarely contains private customer details, will be freer toexplore the Big Data opportunity.2

Corporate Data PolicyFirms are recognising that they need a centralised policyapproach to extract value from data – big or otherwise.In response to increasingly complex data handlingrequirements across the organisation, Chief Data Officer(CDO) powers and responsibilities have become moreprevalent in the executive suite.Although data processing tools are still predominantlyseen at the business unit level, larger organisations havebegun forming parent data committees at the grouplevel. For example, the Australian Government hasadopted a whole-of-government approach to Big Data,establishing a multi-agency working group and a DataAnalytics Centre of Excellence (DACoE) to focus on thedevelopment of capabilities.CostIn most cases, the cost of Big Data analytics is muchclearer than its benefits. And although the entryprice for systems continues to drop, Big Data assetdepreciation is still driven by obsolescence, notproductive output. The faster the technology moves,the more costly it appears on financial statements.Big Data technology continues to evolve faster thanmost.Big Data can also require investment in expansive,power-hungry infrastructure. At a time that datacentre managers are trying to reduce energy footprints,Big Data distributed architectures are demanding evermore expansive physical processing power.A more centralised approach to governing data willpose new challenges for business units and agencieswhich crave autonomy and responsiveness. Data policycentralisation is of course well-intentioned (a typicalfocus being simply to improve the quality and availabilityof information). But where does business dataownership stop and corporate data ownership begin?Can corporate data control be effective withoutimpeding front-line decision making? Such crunchyquestions will only become more challenging as newand valuable data sources are discovered in all parts ofthe organisation.Big Data Tougher, smarter, stronger, faster3

Smarter technology solutionsThe challenges are real, but Big Data products arecountering these forces in new and innovative waysIn response to excess demand for software developersexperienced with new Big Data programmingparadigms such as Map-Reduce, technology vendorsare choosing to build bridges to more traditional totraditional tools, like SQL and R. This convergencereduces barriers to adoption, especially the burdenof reskilling employees and hiring specialised labour.Competitive forces are driving vendors to take thesebarriers lower still. With Big Data technology beingcommoditised faster than usual (in part due to basetechnologies like Hadoop being open source) vendorsare no longer shielded from competition by theirintellectual property alone.Product suppliers are being forced to reposition anddifferentiate their offerings through leaner integrationwith existing technologies and by helping develop datatalent with better training.4Managed Big Data solutions are offering an attractivealternative to owning physical systems. The necessaryinfrastructure can now be leased for as little as adollar a day. Multiply these savings by the reduction indeployment and maintenance costs and the financialsdramatically improve. Organisations increasingly want tolook at options with Big Data solutions as an operatingexpense, reducing capital investment.The cloud model presents its own unique challenges,however. Storing and processing confidential datasetson third party platforms creates a shared responsibilitybetween customers and providers - but organisationsshould make no mistake as to which party is ultimatelyaccountable. Processing data across borders canalso give rise to complex multi-jurisdictional issues.Companies must ensure robust data governancecontrols are in place before migrating to the cloud.

Stronger community engagementThe technology community continues to be captivatedby the Big Data opportunitySelf-organised user groups are growing rapidly andhave started attracting the attention of large industryvendors. One such group in Melbourne experienced13% growth in attendance per meeting last year, andnow boasts more than 700 total registered members.This community plays a significant role in filling theskills gap left by under-investment in data managementcapabilities. It brings together practitioners fromdifferent (and sometimes disconnected) disciplines tocreate a rich network for self-learning and knowledgeexchange.Without this type of open engagement, Australianindustry risks an encounter with critical shortages indata skills. Australian universities are doing their part,but simply do not have the resources to meet the scaleof the problem. Every company that is serious aboutusing Big Data needs a plan to develop and accessexternal data talent.For businesses, there is untapped potential in thiscommunity beyond a source of recruitment. Byproviding the right resources, goals and incentives,communities can be directed to solve real businessproblems. For example, GovHack, a competitionwhich challenges participants to create meaningfulvisualisations from freely available government data,highlights the breadth of opportunity.Big Data Tougher, smarter, stronger, faster5