Innovation is not invention: three tests for the word "breakthrough"

A member of staff hands over a proposal with the line "a breakthrough solution, nobody has done this". What should the owner ask before signing off the money? The answer starts with two ideas that get mixed up, invention and innovation. Three scales of innovation pull them apart and test the label word by word.

How does innovation differ from invention?

An invention produces an idea. Innovation puts something new to work, and that does not need new science. Schumpeter (1947) treats the defining mark of the entrepreneur as "the doing of new things or the doing of things that are already being done in a new way" and calls that innovation (Schumpeter, 1947, p. 151). The new thing need not be spectacular or of historic importance. It could be Bessemer steel or the explosion motor, but it could also be "the Deerfoot sausage" (Schumpeter, 1947, p. 151).

He separates the two roles explicitly. The inventor produces ideas; the entrepreneur "gets things done", which "may but need not embody anything that is scientifically new" (Schumpeter, 1947, p. 152). An idea alone has no economic weight. Greek science had probably produced all that was needed to build a steam engine, yet neither Greeks nor Romans built one, and Leibniz's suggestion of the Suez Canal had no effect on economic history for two hundred years (Schumpeter, 1947, p. 152).

Ooni offers a recent illustration, in the company's own words. Its story page says the founder, Kristian Tapaninaho, found the wood-fired ovens on the market very expensive and bulky, and that he "wasn't an inventor, an engineer or a product designer". After sketches, prototypes and backyard testing he built a portable oven, and the first Ooni project went to Kickstarter in November 2012 (Ooni, n.d.). The company also calls it the world's first portable pellet pizza oven. That is a marketing claim I could not check.

Which three scales test an idea?

How large the change is, who it is new to, and whether it disrupts the market or reinforces it.

The scales are not equally firm. The third, sustaining or disruptive, rests on a source I have read in full. Christensen and colleagues describe sustaining innovations as improvements along the dimensions mainstream customers care about, letting incumbents sell more to their best customers at higher margins. Disruptive innovations are rarer: at first they are worse than existing products on the accepted measures, but they bring a new mix of attributes that appeals to customers at the fringe, such as being smaller, cheaper, more accessible or more convenient. Disruption has occurred once mainstream customers start adopting the entrant's offer in volume (Christensen et al., 2018).

The first two scales, size and novelty, are familiar management vocabulary, and neither source I read defines them as scales. Treat those two rows as prompts for questions, not as tested concepts.

Christensen and colleagues give the label-maker two warnings. First, "disruption" is used as a synonym for any technology or start-up that wants to shake up an industry, and people who rely on such renditions may apply faulty ideas. Second, disruption does not happen everywhere: in hotels, low-end entrants have historically struggled to move up and challenge high-end chains such as Four Seasons (Christensen et al., 2018).

Three named examples: which labels hold up?

Public sources for each example answer only part of the question, and the gaps matter as much as the answers.

Two readings follow. First, Netflix and Blockbuster is the classic pair in this library, yet Wikipedia shows that Blockbuster fell for several reasons at once (Wikipedia, 2026). Christensen and colleagues also separate a firm leaving the market from a firm losing industry leadership, and disruption theory cares more about the second (Christensen et al., 2018).

Second, Rolls-Royce is a change in how the product is sold; the release does not say whether the engine technology was new. On my reading, the release describes a change in how the product is sold and priced, aligning the interests of manufacturer and operator (Rolls-Royce, 2012), which is the business-model slot in types of innovation.

A hypothetical proposal: where does "breakthrough" land after three questions?

This example is entirely invented: the firm and the description are mine, made up to illustrate.

A rice mill in An Giang puts the line "breakthrough drying technology" into a bank loan application. In fact it is a heat-pump dryer that many others already use, bought to replace a husk-fired dryer. Three questions give this table:

The right label is "process improvement, new to the firm, no sign of disruption". That is no insult: cutting drying costs creates real benefit, and by Schumpeter's account the new thing need not be large (Schumpeter, 1947, p. 151). But the word "breakthrough" has no evidence behind it, and a loan officer has reason to doubt the other lines in the file.

A working rule: if the proposal can only answer the first question, the word "breakthrough" has not been earned. To claim "breakthrough", you need evidence for the first (a large change) and the second (nobody in the industry does this). To add "disruptive", you need evidence for the third (rivals' customers are switching, with numbers and dates).

When do the scales fit, and when not?

They fit when you must read someone else's label: an internal proposal, a funding pitch, a partner's presentation. Where a question has no evidence yet, write "unverified" rather than deleting or believing it.

They do not fit when you want to know which idea will succeed. Schumpeter stresses that creative response can practically never be understood in advance from the pre-existing facts, only afterwards (Schumpeter, 1947, p. 150). The scales sort; they do not forecast. They also do not fit when the label changes no decision.

Going deeper: where the scales come from, and what this article did not read

The library's types-of-innovation framework lists four places where something can change: product, process, service and business model, and credits Schumpeter (1947) as its origin. That article contains no such list. Where it does classify, Schumpeter lists enterprise that introduces "new" commodities, technological novelties in the production of "old" commodities, new commercial combinations such as new markets or new sources of supply, and the reorganisation of an industry, for instance by making a monopoly of it (Schumpeter, 1947, p. 153). The two lists are related but not identical, so I treat the library's four as its own arrangement and do not attribute it to Schumpeter.

Christensen and colleagues wrote an intellectual history, not an empirical test of the examples above. The authors themselves say later empirical work has rarely engaged with the theory's key arguments (Christensen et al., 2018). I use the paper because it defines disruption clearly and warns directly against stretching the word, not because it proves Netflix is the standard case.

Limits of my reading: I read Schumpeter's 1947 article and the 2018 Christensen review in full, but not their earlier works (including Christensen's 1997 book). The library's Netflix page is only a press site with no company history, so the Blockbuster details come from Wikipedia, a secondary source that can change. For Ooni and Rolls-Royce the only sources are the companies' own pages and release, with no independent check on the figures or on the word "first".