What is innovation?

What is innovation? Approaches to distinguishing new products and processes from existing products and processes, Bruce S. Tether, 2003

In this paper, Bruce Tether from CRIC discusses various definitions of innovation and some tools for identifying innovations in research.

Tether argues that innovation is sometimes defined as an achievement, sometimes as the results or consequences of that achievement, and sometimes as a business approach - and that there is considerable confusion between the three.

Innovation as achievement

It can be difficult to know how easy it will be to develop a new technology, and the extent to which people will want to use it when it is available. For economists, these represent two types of uncertainties in the development of new technologies - technological uncertainties (will it work?) and market uncertainties (will it sell, how quickly, and will competitors quickly introduce their own versions of the product if it proves successful?). The presence of these uncertainties is often an indicator of innovation and will be important in various definitions of innovation.

Innovation as achievement (in the face of risk and uncertainties) can be two-fold:
1. Achieving a significant leap forward in the technological frontier
2. Re-conceptualising existing problems and thereby restructuring technological systems

Innovation as the consequence of achievement

'Great innovations' are primarily thought great because of the consequences of technologies, and not necessarily because of the novelty of the achievement itself, which in any case has usually transformed substantially from the original achievement through the accretion of little details.

What is important here is that innovation has unintended consequences that benefit everyone. Economists call these unintended consequences spillovers or positive externalities. (There can, of course, also be negative externalities.)

Innovation as dynamic capabilities

The conceptualisation of innovation - as a process - is becoming more widespread. Here, innovation is less associated with particular acts or achievements (and their consequences), and is more associated with an attitude of mind, and a whole ensemble of behaviours and practices associated with that attitude.

A truly innovative firm is not one that introduces a new product 'once in a blue moon', but is instead one that is continuously engaged in practices intended to enhance the probability that it will 'discover' new or better products or processes of making them. ... Central to this concept of innovation is being alive to change. Being flexible - being able to adapt what is done in different circumstances, such as to particular customers needs - is usually insufficient to constitute being truly innovative.

Firms that are innovative tend to have 'dynamic capabilities'. 'A dynamic capability is a learned and stable pattern of collective activity through which the organisation systematically generates and modifies its operating routines in pursuit of improved effectiveness.'
(M. Zollo and S. Winter, 2002). This requires a combination of strategic and organisational skills.

Innovation in this sense does not necessarily coincide with innovation in the form of introducing new products and services.

Product innovation

One conceptualization of product innovation (Saviotti and Metcalfe, 1984) distinguishes between technical characteristics and service characteristics of a product, and maps the two onto each other. (E.g. number of cylinders in a car's engine vs/ acceleration.) Innovation can then happen in 5 conceptually different ways. (In practice they are often interdependent.)
1. A change in the absolute values of one or more of the technical characteristics
2. A change in the mixture or balance of the technical characteristics
3. A change in the pattern of mapping between the technical characteristics an the service characteristics
4. A change in the mixture or balance of the service characteristics
5. A change in the absolute values of one or more of the service characteristics.
The problem with this approach is that it is practically impossible to measure service characteristics, such as brand value objectively.

Process innovation

The pattern of innovation in processes is likely to differ from that of products, and particularly from standardised 'mass produced' products. With standardised products, a new product is typically introduced (following processes of experimentation and prototype development) after which it will remain unchanged for some time. A few minor upgrades will be introduced over time, after which a bigger, generational change will be undertaken. ... The cycle is then repeated with the third and fourth generations of the product. Over time, the scale of improvement between the generations is likely to decline. ... Innovation is fairly easily measured, at least in principle, by the scale of the jumps, or steps between the products available.

By contrast, innovation in processes - and indeed in services as well as customised products - can follow a different path. Sometimes there is rapid learning immediately after the introduction of the process - this is the learning curve, initially the process is slow and inefficient, but gradually and on a continuous basis, minor improvements are introduced. The same pattern is repeated if a substantially new process is later introduced. Alternatively, after the initial innovation, there may be a slow process of continuous improvement, but after a while the scope for improvement diminishes.

The point here is that improvements, and innovations, may be much harder to identify in processes ... than with the archetypal mass-produced standardised products because improvements are less likely to occur in definite steps. In particular, it can be difficult to distinguish between variations and innovations.

Radical innovation and the hierarchical decomposability of technologies

The literature on innovation is replete with references to radical and incremental innovations, yet there is considerable confusion about what distinguishes an incremental from a radical innovation.

3 definitions:
Freeman (1982): Radical innovations are those that transcend the technical limitations (of the existing technologies)
Saviotti, Stubs, Coombs, and Gibbons (1982): Incremental innovation can be defined as a series of quantitative changes in known parameters or in the introduction into a given product of technical characteristics already used in some similar product. A radical innovation would be, instead, the appearance of a new technical characteristic.
Tushman and Anderson (1986) focus on the impact on the industry: Competence-destroying discontinuities are so fundamentally different from previous dominant technologies that the skills and knowledge base required to operate the core technology shift. ... Competence-enhancing discontinuities are order of magnitude improvements in price/performance that build on existing know-how within a product class.

Unfortunately, each case leaves significant room for disagreement in trying to categorize specific innovations.

Constant (1987) clarifies some of the issues by conceptualising technological systems. Ontologically, systems are composed of sub-systems which are composed of an immense variety of components. ... This hierarchical decomposability suggests the absolute relativity of all change: Whether a given change is perceived as radical or incremental depends solely on the hierarchical level chosen. A new valve, a new turbine material or fabrication technique may represent a revolutionary solution to a specific sub-problem at that level; yet the same change, viewed from the level or the total aircraft system may appear only as a typical incremental innovation.

Practical approaches to identifying innovation

Tether distinguishes between 'object based' approaches (the researcher identifies innovations) and 'subject based' approaches (the researcher asks firms about their innovative behaviour).

Several approaches are presented with a special emphasis on OECD's Oslo Manual, which is considered the international standard. Although it helps in specifying innovations, in Tether's opinion it doesn't distinguish enough between adoption of innovations and innovative activity. According to the Oslo Manual, a firm is innovative if it adopts a new technology - even if this adoption occurs without any learning, adaptation or risk-taking on the part of the firm.


Tether recommends a 3-dimensional approach to the evaluation of innovations.

For products: Conceptual novelty, technological uncertainty, and market uncertainty.
For processes: Conceptual novelty, technological uncertainty, learning & adaptation.


This could be a useful approach in justifying my choice of industry for case studies. A problem remains, however, in that the case studies will most likely be in the service industry. How do Tether's recommendations apply to services? The analytical dimensions of product innovations work for services. However, many will develop more along the line of processes with many incremental improvements, rather than large 'jumps'.

In the end, it might be more useful to work with the concept of innovation as a process/attitude/set of practices.

A footnote: Tether emphasizes that he is talking about innovation, not invention. The difference being that an innovation has been commercialised where an invention has not.

No comments: