Sample frameDay 240·Innovation · Architecture

Good morning. Same components, same suppliers, same physics. Only the arrangement changed — and that's the one you'll miss.

Innovation.

Architectural Innovation

5 min read·Apply by lunch

The question

When something new turns up in your market and you hand it to the person who knows your product best, what comes back? I can guess: they'll tell you it's the same components you already use, assembled a little differently, with nothing in it you could not build yourselves inside a quarter — and every word of that will be true.

The idea

Rebecca Henderson and Kim Clark — Henderson then at MIT's Sloan School, Clark at Harvard Business School — published Architectural Innovation in Administrative Science Quarterly in 1990, and it explains a failure that looks like incompetence and isn't. They split product knowledge in two: knowledge of the components, and knowledge of how the components fit together. An innovation can overturn either. Change the components and everybody sees it. Change only the architecture — the same parts, connected and sized differently — and the change is close to invisible from inside an established firm, because architectural knowledge has stopped being knowledge at all. It has been built into who sits with whom, which teams review what, which numbers appear in the weekly pack. Their evidence came from photolithographic alignment equipment, where the market leader changed at nearly every generation, and where one leader examined a rival's new machine and reported back that it was a copy of its own. The failure isn't stupidity or complacency. It's a filing system: an organisation whose channels and reports were shaped by yesterday's architecture will file tomorrow's under nothing new here, accurately, component by component.

The framework

01
Two kinds of knowledge, and only one of them is written down.
Conway's Law says your communication structure shapes the product you build. Henderson and Clark supply the sequel, and it's worse: that same structure then blinds you to a rival's rearrangement of your own components. Component knowledge stays explicit — specifications, named specialists, documents you could hand a new starter. Architectural knowledge stops being knowledge at all after a few successful years and becomes the shape of the firm instead: who shares a floor, which reviews happen together, what the weekly pack reports. That's efficient while the architecture holds. It's also why the filters that would have flagged the change were never documents anyone could revise.
02
What did your expert actually evaluate?
When a new entrant appears, the honest instinct is to hand it to whoever knows the technology best, and that person will assess it the way they assess everything, part by part. On components they'll usually be right, and the report will come back saying nothing new. So change the question you give them. Not is any part of this new, but what does this arrangement do that ours cannot, for whom, and at what cost. The second question survives an architectural innovation intact. The first one doesn't, and it was never designed to.
03
The people who should evaluate it are the ones who own no component.
Architectural blindness is a property of position rather than intelligence, so the fix isn't to think harder, it's to change who looks. Give the assessment to people whose careers aren't staked to one part of the current architecture: a recent hire still inside their first ninety days, someone from a neighbouring industry, a customer who buys the outcome rather than the parts. Ask each of them to describe the new thing without using your internal names for anything, because your vocabulary is itself a piece of the old architecture. What survives that translation is usually what your specialists could not see.
04
Write down the arrangement, then attack it on paper.
Draw your product, or your whole business model, as blocks with lines between them: the parts, what each hands the next, who owns each hand-off. Then work the lines rather than the boxes, because that's where the change lives. Merge two blocks into one. Send a hand-off straight to the customer and delete the block it passed through. Move a block outside the company, onto a partner or a platform. Reverse the order of two. Nearly every architectural innovation in your market is one of those four edits, and an afternoon drawing them yourself is cheaper than reading about them later.

Members only · 4 principles + template + AI mentor

364 more frameworks are waiting.

Architectural Innovation is Day 1 of 365. One framework every morning for a year — across strategy, sales, negotiation, leadership, and more.

$1 /day

Billed $365/yr · cancel any time

365 frameworks

12 topics · templates · AI mentor

Reserve your place — $365/yr →

One payment a year · the whole library · no auto-renewal.

More Innovation frameworks

All 24 innovation frameworks →