Globally, data traffic is expected to grow at 78 per cent compound annual growth rate over the next five years — from 583 pera bytes a month in 2011 to 10,551 pera bytes a month by 2016. Much of this growth will be driven by emerging economies (source: Department of Telecom’s 2012 report ‘Envisioning the Next Telecom Revolution’).
Electronic and telecommunication innovations have significantly changed the input data used to record business transactions. Typically, there is a large volume of electronic inputs and they are generated virtually over the Web at different locations, such as Internet banking and electronic payment portals.
Increase in business volume also increases electronic input, due to the manufacturer’s integrated interfaces with suppliers and customers. For example, a company that uses supplier and customer management portals will get the order, plan the production, and send schedules to suppliers electronically through these portals. In large companies, thousands of goods receipt notes or sales invoices are processed every day. In IT companies, revenue contracts are based on a Master Service Agreement or Statement of Work — thereafter, resource allocation, time efforts, billing and so on are done electronically.
As a result, corporates and their auditors have to deal with huge amounts of electronic data. Data analytics is a powerful tool that can help companies manage their business intelligence systems, as also improve revenues, boost productivity, and create new businesses. However, effective use of this tool calls for new talent, investment in technology, and a significant change in mindset.
Jeff Immelt, CEO of General Electric, said in an interview that his company currently has more than $100 billion in revenue tied to service contracts, where it gets paid based on the product (such as a power plant turbine, jet engine, or locomotive) that is in service. The company uses analytics software to help customers avoid downtime and make the arrangements profitable.
Considering the size of data and its value impact, traditional audit procedures are not sufficient to reduce the audit risk to an acceptable level. It is not appropriate to audit huge amounts of electronic data using manual techniques such as checklists. For example, when auditing sales transactions, an auditor can now collate the entire sales register in an excel spreadsheet and view/ filter it using attributes such as time (date/ month/ quarter), product class, business segment, geography/ territory, cost and profit centres, and so on.
This helps auditors better understand the nature of the population, the statistical distribution/ concentration of data items, and align the risk assessment to the audit procedures.
Further, data analytics helps identify the data items/ patterns that do not correspond to the expected values/ data behaviour, which are typically called exceptions. To make efficient use of data analytics, an auditor should first understand
The client’s business — products, geographies, group structure, sourcing and distribution networks, and so on;
Account balance interdependencies, drivers — for example, the management’s initiative towards environment protection may result in expenses such as legal advice, material procurement, infrastructure, and consultant fee;
Level at which the data is captured and its availability;
Enterprise resource planning (ERP) system and data extraction mechanisms used;
Risks around the information made available — these are essentially front-end and back-end controls that are set/ customised in the ERP system.
Given these added requirements, audit teams in the future may have to include computer programmers, systems specialists and engineers, apart from auditors.
Hemant M. Joshi is Partner and Nikhil Kenjale is Manager, Deloitte Haskins & Sells