🟢 Why classic innovation frameworks fail in our current chaotic markets

Most industrial innovation units still operate with ideas and tools from the likes of Harvard Business Review.

Key concepts such as 'core competence', 'the ambidextrous organization', 'Blue Ocean Strategy', the '70-20-10' portfolio split, 'lean experimentation', or lately 'jobs to be done.' After twenty years of watching these frameworks meet markets that behave less like a staircase and more like stormy weather, it's quite obvious they don't work. At least not out of the box. The pattern is consistent enough: classic innovation frameworks fail in chaotic markets because their authors' reasoning was calibrated to a pace of change that most sectors, from automotive to pharma, no longer have.

Core competence strategy assumes the market moves at a predictable pace

OK, mostly, this is sound advice. For as long as a technology travels in something close to a straight line toward a known application, a condition innovation sadly rarely (if ever) respects once it actually leaves the lab.

In My Big Technology Framework, the space a new technology passes through before settling on a market form is what we call the cone of incertitude, a widening zone of diffusion and hype that only narrows once a genuine bottleneck, energy costs, structural reliability, valuation math, compresses it into a handful of viable applications.

A battery chemistry breakthrough does not know, at the moment of invention, whether it will innovate consumer electronics, automotive, or grid storage. As such, a core competence defined before that bottleneck resolves will oftentimes commit a company to the wrong branch of the cone.

The ambidextrous organization and the 70-20-10 split assume exploration can be scheduled

These two concepts, linked together, are an administratively convenient answer that survives every reorganization precisely because it is convenient rather than because it holds up.

The 70-20-10 split assumes again that the underlying zones stay put. Breaking news: they generally do not! A position that looked like a comfortable seventy percent core bet in January can behave like a genuine frontier bet by autumn once a Grey Swan moves through it. Risks and opportunities that sit on the map from the start get mostly discounted because nobody in a three-year planning session wants to be the pessimist. This is not the only reason why the split doesn't work, but it's clearly the main one in 2026.

In contrast, our portfolio work treats positions as flowing zones rather than fixed budget lines. A beachhead position and a frontier bet are not fixed addresses, but states a project moves between within a single year, sometimes slowly, sometimes they just get upgraded to another position overnight. And if there are some quantitative splits in our approach as well, they are much more stochastic and strategically spread out.

Blue Ocean Strategy and disruption theory assume disruption arrives from one direction

Real markets behave more like a fractal Romanesco cauliflower than a two-axis chart. Zoom into any sub-segment of a supposedly uncontested space, diagnostics within medical devices or lithium chemistry within energy storage, and the competitive complexity you find there roughly matches the complexity of the whole market it sits inside. But the usual competitive space mapping is mostly built around products, such as said medical devices or energy storage units.

What if a competitor developing AI-augmented pattern recognition to sort automotive spare parts decides he can give a shot at your medical imagery market? Or a company in humanoid robotics decides to apply its battery tech to large-scale storage? You'll be cooked, incapable of seeing them coming, or worse, leveraging side opportunities for yourself. Here again, tools like Real Options circumvent these issues. They're built to avoid trying to predict what's next and where opportunities and threats will come from, but rather each zone of a portfolio gets treated as a hedge that can be exercised or walked away from as the picture clarifies, favoring both preparedness and adaptability.

Lean Startup and Jobs to Be Done assume the problem is already knowable

Admittedly, both work reasonably well for a discrete consumer problem that can be tested at low cost with known customers in established markets. They work less well for a tier-one automotive supplier, where the job a product is meant to do is rarely defined at the level a product team is actually testing, and where, as I argued in a piece on systemic bottleneck analysis, most of the return on innovation effort comes from scouting the upstream signal rather than from the speed of the downstream test.

On the innovation canvas we use with clients, separating perceived pain points from invisible ones is critical. The perceived pain points are exactly what a lean experiment validates quickly; the invisible ones, shaped by a demographic shift or a regulatory response further up the value chain, are generally the one that decides whether the product survives.

What does all this tell us?

I still sit in enough of these innovation programs to recognize the same slide reappearing with a new logo in the corner, the same split, the same canvas, the same template. To me, it says less about the people in the room than about how these frameworks get inherited, through a consulting deck, a business school syllabus, or a predecessor's slide library, long before anyone gets a genuine chance to question them. They are also conveniently understood (at least in broad terms) by any executive committees. They feel like solid, reassuring ground.

But the ground right now is anything but solid. And it's past time we collectively upgrade to tools and mindsets that do work in chaotic times.