Nine in ten deep-tech innovations fail to commercialise.
That number gets quoted a lot and there’s been some push back on it, but whatever the exact number might be, more or less… it’s still too many. What gets discussed less is when the failure actually starts. For most of my career working in innovation, I had a nagging feeling I couldn’t quite name. I’d watch technically brilliant teams sit down with a Business Model Canvas or Lean Canvas and struggle. The tools weren’t wrong. They’re excellent tools, two of the most important contributions to innovation practice in the last twenty years. And the teams weren’t incapable.
The further upstream you go, the bigger the misalignment gets.
An observable pattern began to emerge: teams were stuck because they were being asked to answer questions that were genuinely unanswerable at that stage. Who’s your customer? What’s your revenue model? How do you capture value? At the fuzzy front end of deep-tech, those aren’t questions with answers yet. They require a different kind of work first: surfacing assumptions, naming uncertainties, building a shared language before any commercialisation logic can meaningfully begin.
The BMC and Lean Canvas were designed for a stage where that groundwork already exists. At the fuzzy front end, it doesn’t. So teams fill in the boxes anyway. They produce answers that look coherent.
Canvases get completed. But complete isn’t the same as ready.
The valley of death gets framed primarily as a funding gap. That’s not wrong. But a lot of the time, the funding gap is a consequence of something that started much earlier. Investors walk away not because the science is weak, but because founders haven’t demonstrated commercial logic. That gap opens in the fuzzy front end, long before a funding conversation begins.
This is what drove me to pursue my MSc in Sustainability, Innovation and Technology, and ultimately to develop the Impact Innovation Canvas.
Let’s be 100% clear: the IIC isn’t a business modelling canvas. It’s an articulation readiness tool, designed for the stage before a business model can be meaningfully built. Its purpose is to make teams genuinely ready for the BMC and Lean Canvas work, not to replace them. Think of it as the missing first chapter in the Alex Osterwalder lineage.
The workshop structure for doing this work is the 3D Method — a framework built around three dimensions: Discover, Define, and Deliver. The IIC sits at the core of the Define phase, giving teams the shared language they need before the commercial questions can meaningfully begin.
My thesis evaluated the IIC in a live accelerator environment, at Fraunhofer AHEAD, one of Europe’s leading deep-tech accelerators, with real teams under real conditions. The pre/post data showed a 19% improvement in perceived articulation across all ten canvas dimensions.
But the more important finding was the pattern underneath that. The largest gains came precisely where baseline clarity was lowest: Ethical Dilemmas, Impact Metrics and Channels. The canvas was doing its most significant work exactly where teams were most stuck.
After the workshop, the accelerator team decided to move the IIC earlier in their programme pipeline. Teams will now encounter it up front, before progressing toward the Lean Canvas in the months that follow. That decision wasn’t made because of my argument. It was made because it matched what experienced practitioners had been observing for years, and finally had a structured response to.
Completing this research at Tomorrow University of Applied Sciences has been one of the most transformative experiences of my career. It gave rigour to what practice was already showing. The evidence is there. The tool exists.
The question now is how far it can travel. Corporate labs, innovation hubs, university spinout programmes, wherever the impact-driven innovation conversation is happening.
Before the business model, there is work to be done.
The structured front end isn’t a nice-to-have. It’s the essential bridge.
If you’re running an accelerator, innovation lab, or tech transfer programme and this reflects something you’ve been experiencing in your cohorts, I’d welcome the conversation. You can also download the Impact Innovation Canvas free at sequoralab.com/canvas (licensed under Creative Commons).