The pace of business activities is so fast today that every nanosecond can potentially redefine strategy and have a make-or-break impact on key operations within an organisation. Decisions and actions need to be made at the speed of thought, in real time, to stay competitive and thrive.
Paul Scott-Murphy, regional CTO, TIBCO Software, explains why event-driven analytics may be the answer.
A growing number of organisations are taking advantage of event-driven analytics to help identify event patterns and their impacts on business trends and key performance indicators. Savvy companies are tracking and acting quickly on such information via complex event processing to position themselves ahead of competitors or in some cases to identify and act quickly on customer or operational issues they’ve uncovered.
Event-driven analytics provides organisations with the ability to analyse information as it arrives, understand that information, correlate it to anticipate future events, decide how to respond and monitor the outcomes in a closed loop. This approach ensures that nothing slips through cracks.
Read related articles from Business Review Australia:
For companies to achieve these types of capabilities and act on event-driven information in real time they need to develop event-driven architectures. This includes a sense-and-respond environment that lets decision makers identify emerging market trends and act quickly on behalf of customers and ahead of competitors.
The outcomes for organisations include deeper customer engagement at a more personal level despite the increasing volumes of available information. Organisations can use event analytics to get closer to their customers, responding to expectations in a more agile way.”
Here are Paul’s five ways to make event analytics work for your organisation.
1. Combine automation with human intelligence
Human intelligence is a critical part of the process of capturing events, identifying patterns and responding. Without the contextual understanding provided by human intelligence, an organisation’s response will not be able to be optimised. A response can always be automated but without the greater context this response will not be valuable, continuous or repeatable and the quality of response will be degraded. Providing context makes event analytics effective rather than just efficient.
2. Automate analytics with stream-based processing
Capture event information in real time to identify patterns and respond rapidly rather than collating data and analysing it later. This is done by making analytics a part of the business process rather than something you do to a business process.
3. Do not sample
Event analytics provides the ability for all data to be analysed, eliminating the need to sample data. By capturing and processing all of the data organisations have a greater ability to understand the data and identify patterns that will make their response more effective.
4. Be tolerant of erroneous data
Event analytics takes into account the value of all data regardless of quality. The process needs to be tolerant of errors rather than filtering the data or removing it in advance.
5. Correlation not causality
The outcome should not be looking to understand events through cause. Event analytics is about looking for evidence that the event exists and useful correlations that can be made to decide on a response rather than the reason behind them.