Incremental innovation happens by itself, because the customer asks for it and production can see it. Disruptive innovation gets discussed in committee and almost never gets funded. Adjacent innovation — moving only one of the two unknowns — has the best risk-to-return ratio in a mid-sized industrial company, and is the only one that never shows up unless somebody goes looking for it.
In an industrial company, "innovation" usually means one of three things, and they are worth keeping apart because they demand completely different resources, timescales and risk tolerance.
Same product, same customer, same technology. You gain on cost, performance or quality.
You move only one of the two variables: either the customer is new, or the technology is. Never both at once.
New customer and new technology. It is a different business under the same company name.
The observation that organises this piece: adjacent innovation has the best risk-to-return ratio in a mid-sized industrial company, and is the one that gets searched for worst. Incremental happens by itself, because the customer asks for it and production can see it. Disruptive gets discussed in committee and almost never gets funded. Nobody asks for adjacent, it does not appear on its own, and you have to go looking for it with a method.
Before arguing about whether an idea is good, it is worth knowing which kind it is, because that decides who runs it, how long it takes and what money pays for it. Two questions are enough.
Who would buy it?
What would we make it with?
The technical office runs it inside its own plan. It is justified by the saving or the performance, and measured in months.
What you already manufacture, for somebody who does not know you. The work is commercial and about approvals, not R&D.
Your existing customers, with a capability you have to acquire. The work is technical, and the customer is already there.
New customer and new technology at once. It needs its own structure, its own funding and tolerance for it not working out.
In both adjacent boxes one of the two unknowns is already solved. If the customer is the existing one, you know who to ask and what they pay. If the technology is yours, you know what it costs and how long it takes. In the bottom-right box you know neither, and that is why you pay for the whole learning curve.
SCAMPER is a list of verbs that forces you to look at a product from angles you never visit on your own. Here it is applied to yours: type in your product and the part or attribute you want to put through the list, and the seven questions rewrite themselves in your words.
The seven questions are not answered in a meeting: they are answered with data. "What if we substituted the material?" is answered with a datasheet and a cost calculation, not an opinion. The list is for opening the fan; closing it is engineering work, and what separates an idea from a line of work is that somebody has put numbers on it.
SCAMPER works on the product you already have. The other two work outwards, and they are the ones that actually produce adjacent ideas.
All three techniques give far more if you have first written down, without embellishment, which capabilities the company already has installed and paid for: processes, achievable tolerances, materials you have mastered, certifications, sales channel. An adjacent opportunity is precisely a new use of something on that list, and you cannot go looking for one without having written it down.
Coins "lateral thinking" as a counterpoint to vertical reasoning: instead of digging deeper along the obvious path, deliberately change the starting point. This is where the idea of forcing angles you never visit on your own comes from.
Formalises as an acronym the checklist of verbs Alex Osborn had proposed earlier in his work on applied imagination. Its value is that it is a list, not a method: it forces you through all seven angles even when six look absurd.
The origin of the two-axis quadrant. Its contribution is treating "new product" and "new market" as two independent risks that can be taken on separately, which is exactly what defines an adjacent move.
Published in Harvard Business Review. It proposes managing innovation as a portfolio with three horizons — core, adjacent and transformational — and choosing the split explicitly, rather than letting urgency decide it.
A method derived from analysing hundreds of thousands of patents. Its central idea is that technical problems repeat across sectors, and that the solution to one is usually already published in another. It is the theoretical basis for sector transfer.
Processes, tolerances, materials mastered, certifications, channel. This is not a reading reference: it is the document you need written down before applying any of the three techniques.