The original article was published by IDC. You can find the article here.
At IDC, we are seeing increased business focus on accelerating decision velocity – an interest that extends well beyond the traditional domain of IT executives. With greater frequency – and greater urgency – business leaders, including the C-suite, are now taking ownership of initiatives that address decision velocity.
In my recent presentation at IDC Directions 2023, I presented a video clip of a pit stop in a 2019 Formula race in Brazil to illustrate the impressive result of high-velocity decisions by highly trained professionals – four tires changed in under two seconds (yes, the Formula One driver – and his supporting Red Bull team – won that race). While most businesses must contend with less controlled, more volatile environments, Formula One offers an extreme, but nevertheless instructive example, demonstrating how these teams extract and analyze data from every race and apply those learnings to improve performance.
The Impact of C-Suite Turnover
In a recent IDC survey, when we asked organizations about impactful changes over the last few years, they cited a familiar set of challenges, including changing (remote and hybrid) work models, increased regulation, and higher than usual employee turnover. But, 30% of organizations also cited changes in the C-level executives.
Why is that important?
Well, turnover in the C-Suite is often a leading indicator of new, impactful projects and initiatives, as new executives ask questions, address problems, and propose solutions. New projects, particularly ones that involve data analytics and artificial intelligence, will typically kick off within six to 12 months of a major change in the C-Suite. And, as these projects are implemented, business leaders must contend with the three familiar Vs of data – velocity, variety, and volume.
There’s more data than ever before… it’s faster, and it’s more varied. Data is now increasingly distributed geographically, across hybrid and multicloud environments, creating additional complexity. The only way to thrive in these highly volatile and uncertain environments is to improve decision velocity. How organizations do that requires an understanding of what it means to make big decisions and what the decision-making process looks like.
In the 1970s, Colonel John Boyd, USAF, defined one of the best-known decision processes, the OODA loop, which was used by fighter pilots in its earliest iterations (Colonel Boyd later became a consultant to many of the largest organizations in the U.S.). While the OODA loop can become quite complex as additional feedback loops are added, the fundamental model relies on four steps – observe, orient, decide, and act. The goal is to move through the loop as quickly as possible, but there is still a need for control. In fact, that’s how we define decision velocity – the delicate balance between speed and control
Some organizations understand this delicate balance; many don’t and don’t invest accordingly. We found that digitally mature organizations assign importance to both speed and control with twice the frequency of less mature organizations.
We can see the impact on economic output over time, as data dense products grew to comprise two-thirds ($17.3 trillion) of U.S. GDP in 2022. In contrast, low data density products and services grew by only by one trillion since 2006, suggesting that data and decision velocity will be the key differentiating factors for achieving desired societal goals, such as addressing issues of aging populations, food supplies, and energy usage.
And, we have made progress. In 2002, $290 billion was spent on big data and analytics – that is, data warehouses, links, data integration and machine learning and other related tools and supporting infrastructure.
Are we seeing a strong return on that investment? I would argue that we’re not quite there, yet. Challenges persist, as many organizations still struggle with data silos, data quality, data analysis, and ultimately getting data to the right decision makers.
There are also technical challenges. Today 77% of organizations say that data intelligence is a challenge, which translates to a lack of data lineage and an inability to understand where data is and who has access to it. IDC analysts Stewart Bond and Phil Goodwin discussed these challenges, data logistics, and the need for a unified control plane in greater detail in their respective IDC Directions 2023 presentations.
Only 26% of streaming data is analyzed in real time before it makes its way to a repository, such as a data lake. In keeping with the Formula One theme, this is analogous to driving around a track at 200 mph only to make a pit stop that takes two hours – rather than two seconds – to change the tires.
We see this happen time and time again… and the real impact takes the form of data waste. 34% of surveyed executives have indicated that they often don’t get around to using the data they receive, obviating all the investment in data capture, analysis, cleansing and presentation. And, then there’s the issue of data decay. Half of 1,000 organizations surveyed globally indicated that their data loses value within hours; 75% of respondents said it loses value within days. If organizations don’t have the required decision velocity to act, they are simply wasting the data.
In addition to control, organizations must factor in three decision types – situational, scenario, and portfolio. IDC’s research shows that 13% of use cases require situational decisions to be made in seconds. Examples include financial services companies assessing and blocking potentially fraudulent credit card transactions. Just about two-thirds of use cases revolve around scenario decisions, which can be made within a few hours. For a financial services company, a scenario decision may involve the review and approval (or rejection) of a loan application. Portfolio decisions – 22% of use cases – require the most time. Examples include hiring a new chief risk officer or making an M&A decision.
While each decision type requires organizations to maximize decision velocity, the approaches across decision types can differ significantly. As Dr. Hannah Fry, Professor of Mathematics of Cities at the Centre for Advanced Spatial Analysis at University College London has discussed, sometimes organizations need to be data-driven; other times they need to be data-informed. It’s critical to understand the difference because many low-level situational decisions can be data-driven and fully automated. On the other hand, it’s highly unlikely that we’re going to see a ‘ChatCEO’ capability to automate executive-level portfolio decisions any time soon.
We don’t see enough of this nuanced view, yet, but organizations have been making progress toward improved decision velocity – 62% say that automation across the decision-making workflow has increased and 64% reported that metadata is growing faster than raw data, suggesting that data is being organized and structured in ways that make it available for consumption.
So… we’re moving in the right direction.
Just about 40% of organizations are prioritizing budgets for streaming data analysis as they invest in decision velocity. But, organizations need to do more to keep the flywheel of innovation in motion. They need to invest in new technologies and assess new categories of opportunities, including decision intelligence, a category of solutions that empower organizations with greater situational awareness while helping them recognize alternatives, assess risks, and perform simulations and to provide decision makers with better recommendations. And, despite calls for more self-service capabilities, what organizations really want is just the opposite – full service. They want data that they can use with their tools and applications when needed.
A third category is knowledge networks, a new generation of knowledge management tools that are becoming critical to improving decision velocity. As part of this focus on knowledge networks, organizations need to address the issue of humans learning from each other, especially in an environments characterized by high employee turnover. We’re seeing a lot of progress in this space, but a lot more needs to be done.
Finally, organizations must continue to invest in enterprise digital twin technology, bringing it from traditional design and product engineering, where it has been successfully deployed, to the business domain. Easier said than done, but this should be the goal, as organizations leverage decision velocity to improve overall enterprise intelligence.