In today's always-on age, with real-time data as its engine, decisions are a continuous, non-stop process, more connected and complex than ever before. Real-time decisioning marries together the opportunities of live data and analytics within the boundaries of an organisation's decision-making process, with a laser focus on prioritising tangible, organisational value.
This type of real-time decision making doesn’t happen by chance, and it’s more than just clever technologies and lots of data. It’s actually about using real-time data and analytics to understand and advance your organisation’s purpose, optimise the time to actionable insights, and make better decisions faster.
Here are 10 principles which help shape today’s world of real-time decisioning:
1. Decisions, not data, create value
It’s the action taken from a decision that creates or protects value. Having real-time data, analytical tooling, and advanced technologies doesn’t enable meaningful, tangible value if you are unable to get a handle on the decisions that need to be made.
2. Organisations needs to respond to the world as it is
The datasphere can be an overwhelming place. In an event-driven age where every sensor update, ATM interaction, swipe, tweet, and click bombards applications with a continuous stream of data, organisational responsiveness is the new advantage.
Thriving in this event-driven age comes down to an organisation’s ability to predict and detect events as they happen, with organisations deploying contextual, real-time data and historical insight to spot where events are happening and ‘decision windows’ opening.
3. Not all decisions are real-time
Organisations need to make multiple kinds of decisions – from the ad hoc strategic discussions around boardroom tables to the millions of continuous and automated sub-second event-driven decisions that underpin 24/7 operations. Between these two extremes are broader decision types, with different characteristics and ways to leverage data and analytics.
Whilst boundaries between decision types are not always clear, the time constraints usually are. Not all decisions require a real-time response. Understanding the decision time constraints – specifically the frequency of the decision and the decision window – determines if there's a business need for real-time decisioning.
4. ‘Right-time’ analytics drives real time decisioning
While a decision may need to be real-time, it doesn't necessarily mean the analytics, data and tooling driving that decision always needs to be. Often, when people are talking about needing real-time analytics, they are actually referring to the ability to predict, detect or respond to real-world events and take action in a time appropriate manner1.
To capture the value of data, organisations need to marry the speed of analytics and the data available to the available time of the decision window. This means that understanding the decision at hand and the acceptable latency is key to determining the ‘right-time’ analytics for your real-time decisioning.
5. Decision-delay, value-decay: the last responsible moment
The lean software development principle "decide as late as possible" has a mixed reputation due to its justification for all sorts of whacky decision making. However, in real-time decisioning, the concept of the last responsible moment is an impactful lens to explore the trade-off between speed, accuracy and value optimisation.
Understanding a decision’s 'last responsible moment' requires an organisation to consider the decision from a 'decision-delay, value-decay' perspective to evaluate the opportunity cost of delaying a decision whilst seeking more information, or making a decision prematurely without complete information.
6. Humans versus the machine: the balancing act of real-time decisioning
Automating decisions is similar to automating any other business process – you codify a set of rules that link data with decision choices, and a continuous feedback loop provides a self-learning, self-correcting system.
Of course, not all real-time decisions can be automated, and there will always be exceptions and higher-order, judgment-based decisions that require human intervention. Nevertheless, automated decision making fuelled by data instead of human expertise is the fashionable solution, but real-time decisioning is about the sweet point between the two.
Real-time decisioning requires organisations to be alert to scenarios suitable for automation, the risk and reward of doing so and understanding why and when human intervention is needed so the decision isn’t just a good one, it’s the right one.
7. Decision driven first, and then data-driven is the trick to optimising the value of data
Data ages fast, and in the case of real-time data, really fast. An organisation seeking the ultimate advantage from this highly perishable data often find itself over collecting and scrambling around to find a purpose for this data.
Real-time decisioning encourages organisations to shift their mind-set to one where you start with a decision and find data for that purpose. The approach emphasises the importance of not just making do with the available and known datasets; instead, the focus is on what data is needed to protect or create organisational value. The subtle change from what data we have to what data we need enables organisations to prioritise and quantify the value of data and analytics initiatives.
Data gathering is a necessary step if a decision is unable to be made on the data which is already held.
However, for organisations that treat their data as assets once freshly gathered data has met its purpose, there's an opportunity to sweat their data asset further and take a data-driven approach. A secondary data-driven approach can extend the shelf life of highly perishable real-time data highlighting opportunities to enhance existing decisions or expose new insights, simply by considering the full spectrum of decisioning options across different time frames.
8. Managing the hype of cognitive technologies to level up real-time decisioning
A real-time decisioning capability is dependent on continual collection, processing, and analysis of incoming data that underwrites the decision-making process.
Whilst organisations need to be responsive, the exponential data growth in volume and variety that organisations grapple with daily requires something more than traditional approaches. This is where cognitive technologies come into play, exploiting advances in high-performance computing with intelligence technologies and data analytics techniques.
Augmenting cognitive computing technologies with right-time analysis amplifies opportunities for innovation, with value creation taken to a whole new level. But, like all new technologies, managing the hype, so it delivers on its promise is the skill.
9. Data quality always trumps quantity, especially with highly perishable data
An oldie, but a goody: poor data quality is a major contributing factor in the breakdown of information trust and lost opportunities. Data quality measures how fit for purpose a data set is and if it's good enough to support the outcomes it is used for. Real-time decisioning demands organisations gravitate around the decision to be made and find data for that set purpose.
Quality assurance of data and analytical models remains one of the most effective ways to exploit the potential value from data. From an economic perspective alone, as data quality is optimised, so is an organisation's balance sheet, as poor quality data represents risk, regulatory liability and missed opportunities.
10. Empower and protect for real-time
Real-time decisioning is about a mind set and culture change, as it is technology and data. It's not about implementing batch-style strategies in real-time, more designing systems and processes to predict, detect and respond to real-world events and enabling and empowering their people to work in this new way.
In a real-time decisioning environment, you can blink, and with a refresh of data, the decision you had just made has turned out to be the wrong one. Or, if you are continuously bombarded with data and decisions, there's never a sense of accomplishment and getting away from the data smog.
Psychologists have coined the term "information fatigue syndrome" to describe the effect of information overload. In a real-time decisioning organisation, leaders have a role in creating safer environments, one where a learning mind-set is encouraged, and individuals believe and trust their mistakes won't be punished.
About the author
Holly Armitage is a Principal Data Strategist at BAE Systems Applied Intelligence
- Bringing data to the party. Caroline Bellamy is on a mission to transform how the UK Ministry of Defence uses data. She tells Mivy James about her 30-year career in industry and why data holds the key to smarter and faster decision-making across Defence
- Delivering data dividends. Mivy James examines what needs to be done to help the military be more data centric
- We need to talk about data engineering. Organisations across the public and private sectors are increasingly prioritising the role of data engineers – and rightly so, says Alex Richards
- Tuning up data trust. How can governments generate greater trust when it comes to data? Nicola Eschenburg says it can be done, and the sooner the better
- Homing in on data-driven government. Statistical models, data, and analytics have always loomed large in Andy Gregory’s in-tray, but now he’s putting his expertise to good use at the UK’s Home Office. He tells Dylan Langley about his eclectic career and adjusting to life as a senior civil servant.
- How a new model of collaboration can detect data risk. Data sharing has many advocates but it is fraught with ethical risk. Richard Thorburn says technology can help fuel greater collaboration that safeguards data by detecting risk earlier and faster
- Operationalising big data analytics and machine learning. Harriet Barr examines the challenges faced in moving from research to operations and explains why machine learning is actually the easy bit
1. In most cases, the activity and context to enable analytically-driven decisions happens long before the decision itself is required, with data and analytics being pulled in and applied at numerous points across the decisioning process.