It surveys how this term is defined in the published literature — what the sources agree on, where they diverge and what remains contested — and carries its own numbered bibliography. It does not describe what we run. Programme policy is on the core pages.
Our own use of the word is often narrower than general usage; the short definition and the boundary are in the glossary, under Innovation. The review date below matters, because the literature moves.
A definitional review. Citations follow IEEE style; see References.
Abstract
Innovation is among the most heavily used terms in economic and educational policy and among the least settled. The literature does not merely disagree about its meaning; it states plainly that no authoritative definition exists. This article sets out the definitions on offer, the term’s lineage, the one feature the sources agree on — adoption, which separates innovation from invention — the typologies in use, whether innovation requires a market, and the limits of a corpus that is largely grey literature rather than research.
I. Definition
There is no settled definition, and the field says so. Stenberg’s review states it directly: the range of definitions “often overlap and there is no clear and authoritative definition,” and without one “it will be hard to develop strategies to be innovative” [1] — not a marginal complaint, but the finding her paper is named for.
The candidates disagree about what kind of thing innovation is. It is described as a process organisations run [1], a practice of developing and marketing [2], a result — “significant positive change” [3], an enterprise ability to manage knowledge [4], and a career competency of “utilisation, adoption, and commercialisation” [6]. The simplest form is etymological: from Latin innovare, “into new,” yielding “doing something different” [1].
The fullest synthesis is Baregheh, Rowley and Sambrook’s multi-disciplinary definition. The original was not available for this review; it is quoted here as reproduced in Stenberg [1]:
Innovation is the multi-stage process whereby organizations transform ideas into new/improved products, service or processes, in order to advance, compete and differentiate themselves successfully in their marketplace [1, quoting Baregheh et al., 2009, p. 1334].
Stenberg judges even this “still not fulfilled,” since it “only cover[s] an organizational and business part of innovation” [1]. The field’s most complete definition is, on its own account, incomplete.
II. Origins
The conceptual apparatus in general use descends from twentieth-century economics: Schumpeter on organisational innovation and the reorganisation of whole industries (1934), Schmookler on the product/process distinction (1966), Abernathy (1978) and Porter (1986) on the incremental/radical split, and Drucker (2008) on innovation as an essential process of change for enterprise growth. Foster’s Innovation: The Attacker’s Advantage supplies the still-common framing of innovation as “a battle in the marketplace between innovators or attackers trying to make money by changing the order of things, and defenders protecting their existing cash flows” [5].
⚠️ Sourcing caveat. None of those originating texts is on disk. Each reaches this review as a citation inside a policy chapter [4] or a set of teaching notes [5], and should be read before the attributions are leaned on.
The vocabulary of innovation was therefore built to describe firms competing in markets, then applied far beyond that setting without being rebuilt for it — which is Stenberg’s central objection [1].
III. What distinguishes innovation from invention
This is the one distinction the sources agree on, and the only load-bearing claim this review can make with confidence. Adoption is the dividing line. An invention is a new thing; an innovation is a new thing that has been taken up and used. The agreement holds across sources with nothing else in common:
| Source | Where the line falls |
|---|---|
| Berkun [3] | “The act of creating something, even if it solves a problem, should perhaps still not be considered an innovation until it is adopted by other people” — before that, “it’s just an invention with the potential to be an innovation” |
| McKinsey [2] | developing and marketing breakthrough products and services “for adoption by customers” |
| Yataththawala [7] | “invention is the formulation of new ideas for products or processes”; “innovation is the practical application of the invention to a marketable product or service” — Berners-Lee invented the web, Zuckerberg used it |
| Baregheh et al., as reproduced in [1] | a multi-stage process ending in products or services placed in a marketplace |
| Collinge [6] | innovation “involves the utilisation, adoption, and commercialisation of new ideas, technologies and research” |
Two corollaries follow, and they are what the distinction is for: an idea is not an innovation, and a prototype nobody has used is not an innovation. Harrowfield turns the criterion into an exercise — the zipper was patented in the mid-nineteenth century but reached clothing only in the 1930s, so students are asked to “discuss when the zipper became an ‘innovation’” [5]. Invention and innovation are different dates, and only the second is contestable.
⚠️ The agreement is broad but the sources are not of equal weight. [3] is a blog post; [2] a consultancy explainer; [7] an undergraduate e-magazine article; [5] teaching material for a wool-textile module; [6] an institutional competency guide. The only item with research apparatus is [1], and it reports the Baregheh synthesis rather than arguing for it. The convergence is evidence of an established professional convention, not a tested conceptual result.
The grassroots-innovation literature partitions the same word differently — by the economy the work is produced in and the innovator’s position rather than by adoption (see the companion review). The criteria are independent: a case can satisfy one while failing the other.
IV. The evidence
A. The standard typology has four types
The four-part typology in policy use is attributed to OECD (2005a) in [4]: product innovation (new or substantially improved goods or services), process innovation (a new or significantly improved production or delivery method), marketing innovation (a new marketing method involving significant changes in product design, packaging, promotion or pricing), and organisational innovation (creation or alteration of business practices, workplace organisation or external relations) [4]. The chapter records a limit: the product/process distinction “is fairly obvious at the level of individual firms or industries; however, it can be ambiguous when examined in terms of the overall economy” [4]. The typology is scale-dependent, not absolute.
B. Incremental versus radical is a spectrum, and the small end dominates
Incremental innovation “involves modest changes in existing knowledge and resources or the existing products and services in the market”; radical innovation “involves the development of new businesses or product lines, based on entirely new ideas or technologies” [4]. While policymakers often prefer radical innovation for its high return, “incremental innovation dominates the market” and may generate greater cumulative long-term impact [4]. Harrowfield reaches the same position from practice, using monumental for the radical pole and insisting both are “part of a continuous spectrum,” where “often categorisation will depend on the viewpoint of the observer” [5].
C. Whether innovation requires a market
Most definitions surveyed realise value through customers and marketplaces [1], [2], [6], [7]. One does not. Chapter VII of the policy guidebook opens by defining innovation as the ability of an enterprise to “manage knowledge creatively in response to market-articulated demand and/or other social needs,” attributing it to OECD (1999) [4].
⚠️ Attribution check. The attribution holds as made by the citing document: the chapter’s own first paragraph carries “(OECD, 1999)” against that wording, and its innovation-system diagram is likewise sourced to OECD, 1999 [4]. The OECD original is not on disk and was not read, so this is quoted-within, and the citing chapter itself carries no listed authors and no year in the bibliographic record.
That clause — “and/or other social needs” — is the only place in this corpus where a definition explicitly admits a non-market trigger. Collinge comes closest independently, directing innovation “to benefit society and the economy” under a principle of responsible innovation, defined via UKRI as a process that “takes the wider impacts of research and innovation into account” [6]. Neither displaces the commercial mainstream, but it is not unanimous.
D. What the word is actually understood to mean
Stenberg surveyed 200 respondents recruited through Facebook [1]. The most frequent associations were growth (40% of 349 answers), new thinking (~36%) and creativity (~34%); about half agreed innovation is problem-solving. On the statement “innovation leads to something important,” only 12% strongly agreed and 41% neither agreed nor disagreed [1] — popular usage attaches the word to growth and creativity far more readily than to consequence. The limits are considerable: a convenience sample from one platform, no instrument for testing why associations are made, and an internal inconsistency about method, the abstract calling it “a qualitative approach” where the method section opens “This is a quantitative study” [1].
V. Scope of adoption
The term carries a substantial institutional apparatus. Innovation is measured at national level by competing indices — the BCG International Rankings of Innovation Capability (2009, 110 countries), the INSEAD Global Innovation Index (2011, scored 0–100) and the WEF Global Competitiveness Report (2010, scored 1–7), all three reported side by side in [4] — and organised through national and, more recently, subnational innovation systems, the latter described as “a relatively new concept” [4].
It is also operationalised as an assessable capability. McKinsey lists eight essentials — aspire, choose, discover, evolve, accelerate, scale, extend, mobilise — reporting internal research that firms mastering them “generate economic profit that is 2.4 times higher than that of other players” [2]; Collinge’s university framework sets out ten competencies across five capability levels, from Potential Innovator to Innovation Leader [6]. Both are practitioner instruments, not validated constructs.
Adoption is uneven in kind, not only in degree. Firms in developing economies of Asia and the Pacific “are not pushing the frontiers of technology; instead, their innovations consist of introducing modified products and processes nationally, or even subnationally, and are mostly application-oriented” [4]. The same chapter cautions that the comprehensive innovation-system approach “may not always work with less developed countries in the region” without a well-targeted and manageable plan with reasonable resources [4]. The term travels; the model it presumes does not.
VI. Limitations
The definitional non-closure is the field’s own finding, not an outside criticism. Stenberg states there is no clear and authoritative definition and that strategy suffers for it, and judges the fullest synthesis available to cover only the organisational and business part of the concept [1]. Anything built on “innovation” as a defined term inherits that gap.
Measurement is unresolved and partly circular. Three international indices of the same construct sit in one table with three different scales and three different rankings [4], and at least one draws on survey judgements of “capacity for innovation” — perceptions of the thing being measured. McKinsey rejects activity proxies, noting outcome measures are “often more meaningful than looking inward at measures of activity, such as the number of patents secured,” but its alternative, the “green box,” is a proprietary framing, not an independent metric [2]. Berkun’s threshold — “significant is a 30% or more improvement in something” — is asserted rather than derived [3].
The word is inflated in professional usage, and the complaint is on the record. Berkun names “we innovate every day,” “chief innovator” and “innovation pipeline” as inflations, argues that using the word lightly shows “hubris in the present and ignorance of the past,” and recommends naming the specific problem solved instead [3]; Stenberg records the same neutrally, the word having “become a buzz over the last decades” [1]. Where a term is applied to every activity it stops discriminating between them — which is why the adoption criterion in Section III matters, since it is the only test in this corpus that anything can actually fail.
The corpus is largely grey literature. Of the six documents read here, one is a consultancy explainer [2], one an unattributed policy chapter [4], one a set of module teaching notes [5], one an institutional user guide [6] and one a student e-magazine article [7]; [3] is a blog post fetched separately. Only [1] has research apparatus, and it is an undated paper reporting a convenience sample. Three of the seven records carry no year and one carries no author. The convergence in Section III is real, but it is a convergence of professional convention; a reader needing a citable scholarly definition should go to Baregheh, Rowley and Sambrook (2009) directly, which this review could not.
References
[1] A. Stenberg, “What does Innovation mean — a term without a clear definition,” Dept. of Business, Economics and Law, University of Halmstad, Halmstad, Sweden. Undated — no year or DOI in the record; the title page gives only author, title and affiliation. Quotes Baregheh, Rowley & Sambrook (2009), Management Decisions, vol. 47, no. 8, pp. 1323–1339, not read for this review.
[2] McKinsey & Company, “What is innovation?,” McKinsey Explainers, Aug. 2022. Practitioner explainer, not research; cites internal research whose method is not reproduced.
[3] S. Berkun, “The best definition of innovation,” scottberkun.com, Apr. 3, 2013. [Online]. Available: https://scottberkun.com/2013/the-best-definition-of-innovation/ Blog post; wording confirmed by fetch, 2026-08-03.
[4] “Innovation and technology,” Chapter VII, pp. 131 ff., in a policy guidebook on SME development in Asia and the Pacific issued by the UN Economic and Social Commission for Asia and the Pacific. The record carries no authors and no year; the parent volume is identified from the chapter’s own text and was not retrieved here. Internal citations run to 2012. All OECD, Schumpeter, Schmookler, Abernathy, Porter, Leifer, Lundvall, Drucker and ADB material attributed to this source is quoted-within.
[5] B. Harrowfield, “An introduction to innovation,” module notes, Innovations in Wool Textile Technology. No year in the record; the document identifies the author as a wool industry consultant. Teaching material, cited for framing only. Quotes R. N. Foster’s Innovation: The Attacker’s Advantage, not read for this review.
[6] K. Collinge, “Innovation Competency Framework: User Guide,” Edinburgh Innovations and the University of Edinburgh, 2024. [Online]. Available: eil.ac/InnovationCareersHub Built from interviews and workshops with innovation professionals; not a validated instrument.
[7] P. Yataththawala, “Are innovations and inventions the same?,” student e-magazine, University of Moratuwa, Sri Lanka. No year in the record; the masthead reads “First edition | May 2021”. Undergraduate magazine piece, not peer-reviewed.