Decisions, decisions, decisions. We make them every day: bad ones, good ones, seemingly inconsequential ones that have profound implications. It’s no wonder “analysis paralysis” plagues so many of us, even if the stakes are as low as picking a brand of ketchup in a supermarket aisle.
When the stakes are higher – say, running a multibillion-dollar company – businesses need to strategise. Many years ago, data collection meant hiring consultants, conducting thorough A/B testing and in-depth market analysis. More recently, the explosion of big data and smarter analytics software has allowed for “business intelligence”, or BI: really deep, granular analytics, helping organisations to carve out the correct course of travel based on the information they hold about the past and present.
While business intelligence has proved enormously valuable to organisations of all stripes, a new practice is emerging: decision intelligence. Put simply, this is the commercial application of artificial intelligence (AI) to make better business decisions. It aims to take data-led insights a step further. It’s been called the “new BI”, but really it means broadening the capabilities established by business intelligence in a much more intuitive, smarter way.
All the data in the world is useless if it doesn’t teach you anything. Just look at the NSA, which busied itself by diligently collecting all that it could on American citizens, rendering its surveillance systems ineffective, according to whistleblower William Binney. Useful data needs to be separated from the junk, and businesses need to draw insights from it.
Decision intelligence could offer a solution. Merging data science with social science, it promises to democratise analytics, applying machine learning to data and helping people make decisions. The social science aspect is crucial. While machines are getting much better at doing the heavy lifting for laborious manual tasks, they still lack the nuance and understanding that humans possess.
A decision intelligence initiative might mean examining sets of data, running potential outcomes through machine learning models, and presenting decision makers with potential courses of action to take, all in plain English.
This isn’t mere buzz from vendors. Analyst firm Gartner believes a third of all large organisations will employ analysts that practice decision intelligence, including decision modelling, as soon as 2023. That’s partly because it’s becoming far more complex to make business decisions: a recent Gartner survey found that over half of respondents (65%) believed the decisions they made were more complex than two years ago, while 53% said there was more pressure to justify or explain their decisions.
Big tech leadership
As is often the case, big tech is shaping the trends. Pioneering companies with the resources to spearhead experimentation, such as Google, are rethinking their approach to decisions and data. In 2019, Google hired its first “chief decision scientist”, technology evangelist and influential data scientist Cassie Kozyrkov, to help meld data-led AI tools with behavioural science.
IBM calls the practice “Decision Optimization”, with its customers including logistics firms, warehouse providers, financial services organisations and energy companies. Meanwhile, Manchester-based company Peak recently drew $75m in Series C funding from Softbank’s Vision Fund, and Quantellia, the business founded by machine learning pioneer Lorien Pratt, who has written a book on decision intelligence, counts major players like Cisco, SAP, and RBS among its customers. Typical use cases span industries like finance, transportation, pharmaceuticals, utilities and others where supply chain optimisation is critical to maintain competitiveness.
Traditional BI relies on static data, meaning organisations are looking in the rear-view mirror at a point in time, says Claire Rutkowski, CIO at software company Bentley Systems. While it provides a wealth of information that can be used as the basis of better decisions, the user must know what data to look for and how to interpret it, she explains.
“They need to work with a data office or similar function to figure out how to extract the data, manipulate it, create filters, then validate the results before the report or dashboard is published.”
Decision intelligence is “much more powerful,” Rutkowski adds, because it allows for natural language querying and can answer deeper questions, offering insights on the “why” and not just the “what”.
At present, however, the framework for decision intelligence is loosely defined. Krishna Roy, senior research analyst for data science and analytics at 451 Research, says decision intelligence uses automation and machine learning, which isn’t usually the case with plain old business intelligence.
This seemingly small divergence has significant business implications, which are already proving a powerful tool for some businesses.
The role of AI
For beverage multinational Molson Coors, there was a need to gain better insight into its vast, complex operations, continually improving how these were managed at a speed and scale that wouldn’t have been possible without AI.
In mid-2020 this led the company to Peak. Together, they set about assessing potential areas of Molson Coors’s operations where the platform could be deployed, speaking with key stakeholders within the business in an effort to ascertain how their roles could be improved. By Christmas, they’d agreed on a trial, with work starting in February 2021 and rolling out nationally in July.
Mark Elston, digital solutions controller at Molson Coors Beverage Company, explains that the company has a huge technical services team operating in the UK. Decision intelligence helps capture insights into those touchpoints and quickly translate them into service improvements.
One of the brewer’s major commercial projects this year was rebranding one of its flagship lagers, Coors Light, to Coors, says Elston. This multimillion-pound investment meant replacing 20,000 dispense points with newly branded versions, a “massive operation where a decision intelligence solution really came into its own”.
The system takes information from engineers’ schedules, like planned jobs, most likely travel routes, skill specialisms and the equipment they have in their vehicles, then calculates where they are able to conduct additional rebrand work without significant disruption, Elston says.
“Because this is all being done by a computer, it happens live, making the whole process incredibly dynamic and responsive. The platform was instrumental in getting the work done efficiently and at pace, enabling our customers to be up and running with newly branded dispense points faster, with minimal disruption.”
Stories like this are becoming more common, with decision intelligence making a difference to businesses right now.
Increased adoption will likely change the pace and quality of decision-making, enabling organisations to speed up the entire process, quickly glean insights and develop plans to either act as countermeasures to their discoveries or drive opportunities harder, says Rutkowski. Bentley Systems is currently layering decision intelligence tools into its data technology stack to draw additional insights from its data.
“Companies will be nimbler with their strategies, which should drive better performance and better results. Corrective action will be more focused when we understand the ‘why’ and will likely be more successful,” she adds. “When companies discover positive insights on a given data set, they can further leverage strategies and initiatives that are driving growth, expanding it even further.”
And for businesses struggling to recruit professionals to interpret the data they generate and draw insight from it, decision intelligence could provide some reprieve for alleviating the data skills gap, which the British government has acknowledged is an issue in the UK.
Decision intelligence “promises to bring analytics to the masses”, explains Roy. “It’s all about providing insights to answer business questions, without the user requiring smarts in analytics or data.”
Businesses have created self-serve data platforms for employees, meaning they don’t need to be data specialists to draw insights and make better decisions. Decision intelligence products could simplify this process, making analytics available to everyone that would benefit from it.
“Analytics has yet to be democratised, so people without skills in using analysis tools or understanding struggle to get the insights they need to do their job,” says Roy. “Decision intelligence is one of the latest ways vendors are looking to address the skills gap – by automating the delivering of business insights, as well as explaining them in a way a nontechnical user will understand.”
Organisations that want to get started should first research how AI and decision intelligence could fit into their business, advises Elston. Key stakeholders should be brought into the conversation early to hear how the technology will benefit them. Molson Coors spent months on this initial research and conversation phase, which “paid dividends” in achieving buy-in later on, says Elston, and allowed the company to hit the ground running.
Start simply, Elston adds. Early deployments should be easy to implement. They should deliver value quickly to build confidence with leadership teams as well as those that are using the solution.
“Introducing this technology effectively requires the relevant stakeholders to embrace change, and there is no better way of getting buy-in to that than a successful proof of concept,” he says.
Decision intelligence will probably be a standard component for businesses one day, but it will be “critical for competitiveness” in the meantime, says Roy. Businesses would do well to explore the field today and take at least some of the pain out of decision-making.