Data on its own is not enough - you need to analyse it and apply the outcomes to decisions. To benchmark how well this is happening within companies around the world, SAS carried out a recent survey. David Reed spoke to the company about its findings.
Wider use of analytics increases productivity, reduces risks and helps individuals to make faster and better decisions. Every reader of DataIQ would readily agree with that statement. But is there a benchmark for where organisations are applying analytics, what benefits they are seeing and how far up the maturity curve they have progressed?
To find out, SAS sponsored a survey carried out by Harvard Business Review Analytic Services among 731 executives, managers, and professionals across all industries. It found that at least 50% more organisations enjoyed improved financial performance from their analytics if they had widely distributed access to those tools.
Laurie Miles, head of analytics, SAS UK and Ireland spoke to DataIQ about the findings and shared some exclusive findings about the differences between the use of of analytics in the UK compared to the rest of the world.
DataIQ: The financial benefits of adopting analytics enterprise-wide seem clear from the survey, yet the majority of companies globally still lag behind . What do you see as the main obstacles to adopting a data-driven culture?
Laurie Miles: “A key obstacle is the long-held misconception that data analytics is simply an IT issue which should be managed from within the IT department. Inevitably, this hinders an enterprise-wide approach, which can only be overcome by investing in data analytics as a business requirement across the board. This means delivering data to business users in formats they can understand and work with as well as challenging cultural norms.”
“In some cases, the analytics function may not have done a great job at disseminating insights across the company. Improved visualisation tools now provide a powerful aid in the democratisation of analytics, which makes data analysis easier to perform at scale.”
“A final issue that frequently hampers enterprise-wide analytics adoption is a lack of data governance. Without focusing on the organisational, cultural and management issues associated with creating high quality, valuable data across the organisation, any analytics effort will fail to deliver the levels of success it otherwise would. Data governance isn’t easy and requires engagement at the highest levels, so these are programmes that require extensive thought and a time investment.”
DataIQ: The report references skills as a requirement, yet these are in short supply – are there obvious ways to fill this gap?
Laurie Miles: “Education is undoubtedly a key factor in enabling companies to realise the potential of big data. From school age driving interest in science and maths, on to universities, we are working to increase skills and knowledge through SAS Curriculum Pathways. The initiative is designed to help instil interest in science and maths at school and, as students progress to university, they will have the option of selecting degrees relevant to analytics and big data.”
“Curriculum Pathways is a set of resources aligned to the curriculum in England for students from age 12 through to A-level age. The initiative is already used to support 80,000 students and 10,000 teachers across the US.”
“Our Global Academic Programme is designed to build or expand SAS knowledge within colleges and universities. Through a variety of services, including curriculum consulting, certificate programmes, instructor training programmes and access to teaching materials, it aims to increase knowledge of big data and analytics, which can subsequently be applied in a corporate setting.”
DataIQ: A data-driven culture should ideally be enterprise-wide, but this can seem too big and scary. Where is the best place to start?
Laurie Miles: “To ease businesses into a driven-data culture, it is important to assess current maturity and, if necessary, be realistic in order to prove the ‘quick wins’. The use of analytics tools deployed on one closely-defined business problem with potential for significant return or cost saving is sometimes simple to justify. It also provides a great example to senior management and other areas of the company to help communicate the impact analytics can make.”
DataIQ: Can analytical KPIs be created to help identify the benefits directly, or do these rely on overall business (or function) performance improvement?
Laurie Miles: “There are certain KPIs which can be measured, such as the time it takes to reach a decision. This is becoming ever more important to organisations looking to leverage big data analytics to become more competitive. The ever-expanding capability of computing and concurrent fall in its unit cost mean analytics is getting faster.”
“Using SAS High-Performance Analytics, our customers are analysing billions of rows of data in minutes, an activity that might have previously been an overnight job or taken even longer. Inevitably, it is the traditional ROI-style KPIs in areas like customer churn or procurement spend reductions that continue to be most helpful when assessing benefits.”
DataIQ: With ever more data being generated and considered in organisations, can existing analytical solutions cope?
Laurie Miles: “There is no doubt that big data is ever-increasing in terms of volume, variety and velocity. Current methods are struggling to process the volumes of data and are failing to unlock the most business value now available. This effect is heightened as data gets bigger, and decision-making windows decrease.”
“The greater variety reflects the growing amount of unstructured data, eg, content from social media, call centre logs, emails, loan applications, etc. It's up to 90% of all business data. Text analytics allows you to get insights from electronic text data regardless of format or location. Our solution uses sophisticated linguistic rules and statistical methods to evaluate text like a human mind would, minus the inconsistency and ambiguity. It enables businesses to interpret customers' opinions, improve products, optimise services, streamline processes and make proactive, fact-based decisions.”
“To keep up with growing data volumes, organisations are practicing new techniques for big data analytics. Our High-Performance Analytics suite offers a number of distributed processing options to deal with ever-increasing volumes of data in a matter of minutes and hours, as opposed to days or weeks offered by previous alternatives. “
“There are different options available depending on the demands of the business. These include In-Memory Analytics, where big data and analytics computations are processed in-memory and distributed across a dedicated set of nodes to produce insights to solve complex problems in near-real time; In-Database, where data integration and analytic functions are executed inside the database, enabling better data governance and speeding time to insight because there’s no need to move or convert data; and Grid Computing, where jobs are processed in a shared, centrally-managed pool of IT resources, which promotes efficiency, lower cost and better performance.”
“The solution type that best fits an organisation will depend on a wide number of factors including data volumes and formats, the type of decisions being made and current maturity.”
UK analytics more widespread, but still lagging
UK practitioners are slightly ahead of the global analytics maturity curve in some aspects of their adoption. Among the 47 UK-based respondents, use of analytics by the function for setting direction or making decisions scored 4.46 overall, compared to a score of 4.32 across the 731 total answers (see Chart One). This was reflected in the high level of importance which peers place on data to achieve results, which scored 4.1 in the UK against 3.9 globally.
There is also significantly higher enterprise-wide adoption of formal processes for decision-making (see Chart Two), with 40% identifying this compared to 27% globally, although analytics is often at a departmental level (40% compared to 29%). This sometimes forms the starting point for broader adoption by the enterprise. But the UK lags in metrics, with only 38% using quantitative measures compared to 52% overall, while 28% rely on judgement against 18% overall.
Where the UK most lags is in the use of analytics for real-time decision making at 21% against 34% globally and in the admission by 13% of respondents that their function does not use analytics. Just 4% of the global sample said this. Scored for maturity, analytics in the UK rated 2.36 against a global average of 2.66 (see Chart Three).
In its report on the findings, SAS and HBR Analytics Services spoke to a number of leading organisations globally about the benefits they were achieving. “In consumer products, you might think there isn’t much reason to invest in another new laundry detergent, but consumers respond strongly to innovation,” said Filippo Passerini, group president of global business services and chief information officer of Procter & Gamble.
He added: “What’s different now is the tools allow me to see what was important last year, last quarter and last week so I can understand what will happen tomorrow, next month and next year. That is a huge conceptual shift in thinking. We’ve used data analysis for 50 years, but we’re just beginning to develop predictive ability through business models to anticipate what’s coming.”
Jim Bander, national manager of decision sciences in the risk management department of Toyota Financial Services, said: “The point of becoming a data-driven company is to become a wiser company by making better decisions. And that isn’t simply a matter of data, but of fitting analytics into your corporate culture.”
He added: “For example, Toyota has a culture of continuous improvement and respect for people, including consensus building. My job is to fit analytics and data-driven decision making into that kaizen framework. An organisation with a different corporate culture - whether a mass-production manufacturer or a Silicon Valley start-up or a government agency - would find a very different way to integrate analytics into its own decision-making processes.”