Streaming data analysis and in-memory processing have increased the opportunities for using information in near real time. However, such applications still come with a cost overhead, and the parameters of speed, reach, and intelligence are a good framework against which to assess the potential benefits. Organizations should assess whether information systems are capable of delivering useful and actionable information (intelligence), within an appropriate timeframe (speed), and to the people who require it (reach).
This is a very different approach to assessing benefits for most transactional applications. Much of the benefit of these systems has been founded on process automation, and should (at least in theory) be easier to measure: if the costs of operating that process are reduced, for instance through handling purchase invoices with fewer staff or by reducing the levels of raw material held in stock, then there is a tangible outcome.
A significant proportion of IT investment is now going into analytical applications, and determining an equivalent quantifiable outcome becomes more difficult. For applications such as business intelligence, content management, and enterprise collaboration, we recommend an approach that includes these measures of speed, reach, and intelligence in addition to direct financial benefits.
In the case of speed, measures should focus on how quickly information moves through the organization: time taken for operational information to be available to a decision-maker, and time taken from decision to action. Both can be expressed as an "information velocity ratio" between current operation and the anticipated enhanced performance. For example, if a business intelligence application reduces the time taken for a product manager to receive an analysis of a marketing campaign from two weeks to two days, then the velocity ratio is 86%. Similarly, if a content management system reduces the time taken to create a new website from 10 days to seven days, the velocity ratio is 30%.
For reach, measures should highlight the potential to distribute or receive information both internally and externally to the business. This should take into account the size of the recipient audience of employees, customers, suppliers, or combinations thereof compared to the total potential audience. For example, consider a portal that is designed as a self-service HR facility for employees. Out of a total of 10,000 employees, you calculate that around 8,000 will be able to access the portal: the reach is therefore 80% of the potential audience. Similarly, for an online store application, your research suggests that 15% of orders will arrive via this channel.
For intelligence, measures must concentrate on the value of the information processed. Here there are several possible approaches, which can be used in conjunction. For some projects, it is feasible to gauge the proportion of existing cost or revenue categories to which the information relates. As an example, consider a call center automation application that is deployed to reduce the costs of handling incoming customer enquiries; if your total call center direct costs are €2m per annum, out of a total cost base of €40m, then the proportion of costs that are targeted by the system is 5%. Different types of intelligence metric can target specific elements of the organization's knowledge capital: the proportion of core business processes to which the system is applicable, the relative effect on organizational learning and innovation, or the influence on the company's brand and market perception.
Speed, reach, and intelligence, when combined with direct financial outcomes, give a better framework for assessing the value of information-centric systems to the organization. The idea is not to calculate an absolute worth, but to provide a comparison of how different investments may impact the business. You will quickly find that certain types of project are biased toward one or other of the areas: while this is inevitable, we would question projects that do not show positive benefits in at least three of the four categories.
Straight Talk is a weekly briefing from the desk of the Chief Research Officer. To receive this newsletter by email, please contact us.