A new layer of the tech stack: The Insights Layer: Where algorithms meet speeding up the OODA loop
Three quick points:
- Capital is moving faster than ever before with implications on the half-life of competitive advantage (merely witness the increasing ‘topple rate of companies across industries – now averaging 14% / year) with the urgency to make “smart capital” decisions faster and more effective. Increasingly, as Frank Liddy, a sophisticated investor and operator put it, increasingly, capital managers are calling this the speeding up of the OODA loop. The OODA loop, taken from the military, refers to a process to Observe (the situation) : Orient (to that situation) : Decide (on options to take) and Act (move!). Speeding up the OODA loop, is critical to decide:
- Where to focus
- How to prioritize
- How to allocate capital / resources more effectively – given different risk profiles & alpha opportunities.
- More and more companies need insight into where to focus and how to prioritize given the blunt reality:
- That the pace of competitive movement (as evidenced by capital flow, the volume. & velocity of start-ups, and the speed of new product innovations) continues to accelerate, and
- The increasing concentration of economic profit by fewer and fewer firms continues apace putting more and more pressure on ongoing competitive sustainability. (Not incidentally, this double whammy at both ends – of fewer firms doing better and more firms struggling is exacerbated an increasingly fast ageing business model based on providing “more of the same types of consulting advice” – e.g., do what everyone else is doing – delivered in an increasingly out of date business model of experienced, “seen-it-before” experts
- The maturing and explosion of new tooling and methods to test, learn from and harness algorithms in service of specific questions to answer.
The implication? There WILL be a new “layer” of the technology stack resulting from the maturing of AI-driven techniques and architectures – an INSIGHTS layer, no less significant than the other layers that have developed and matured over time – with sets of tools and eventually market leaders driving them.
To take but one example:
20 years ago, building IT applications was (and some more argue still is) complex, with few consistent tools or framework to model much less build and test applications across much less within any particular department. In *any* point of such friction, new capabilities – and firms – will be developed. RationalRose, with its modeling tools, consistent frameworks and set of “build” tools tackled such IT “build friction” and became one of the world’s leaders in the space until it became absorbed into IBM years ago. Take any layer of the stack – and you easily tell a similar story.
Well, now, it’s time for an Insights Layer, and the time is now and moving quickly.
Insights are no more than the expression of a hypothesis or method of looking or answering questions. Typically, insights have been the primary purview of people – who obtain information, analyze such information, perhaps develop a hunch about it or based on experience, opine on the question asked. The process used was, and typically remains, a “carbon-based” (human-based) algorithm of reasoning through a topic.
Data-science based algorithms do the same – just faster and arguably differently for reasons it’s not necessary to explore here.
Systematizing insights, algorithmically, is in its early days. Yet, efforts are being made to create repeatable, scalable and learning tools – we call them “Primitives” – to systematize an extremely broad range of insights.
With two clear objectives:
- Diagnose quickly to accelerate the OODA loop with speed and scale
- To accelerate the build – and field – of the new layer of the tech stack – that of the Insights Layer.
And with that new layer will emerge the ability to answer “complex” non-obvious questions much more effectively and, oh yes, faster.
We are building such an Insights Layer which supports algorithmically-based decision – and strategies – in hours and days, rather than weeks, months or never across a wide range of domains.
Why? Because it’s time.
Want new articles before they get published?
Subscribe to our Blog.