Many with an interest in technology will have been watching developments in artificial intelligence, and in particular generative pre-trained transformers. A much smaller group may have performed some experiments of their own, but the release of ChatGPT in November 2022 brought this technology to the general population. In fact, just two months after launch ChatGPT was estimated to have over 100 million monthly active users making it the fastest growing consumer application in history. In the following few months, a new version, based on Open AI’s GPT-4, was released extending the capabilities even further.
It is currently unclear what the impact of this technology may be on businesses, and the societies they serve; it might be limited or profound, and governments are considering what, if any, constraints they will impose. What is clear is that few would dispute that ChatGPT does not represent substantive innovation. Some, including Nvidia CEO Jensen Huang, have drawn comparison to the launch of the iPhone.
This post is not going to discuss the impact of this technology on businesses, though that may be significant. A previous post, however, discussed being adaptive and good security architecture does not presume what the future state may be. What we will discuss is innovation and the attribute Innovative. Most people can give some examples of innovation, or recognise when it hits the market, but how can we define innovation.
Firstly, we should recognise that innovation is generally considered as distinct from invention. Invention is focussed on the concept, the idea, whereas innovation leverages invention to provide something of use in the real world. Greg Sattle, in his book Mapping Innovation1, defines four types of innovation.
A lot of innovation that occurs could best be categorised as Sustaining Innovation. This is where we seek to get better at what we’re already doing – to improve existing capabilities in existing markets. We have a fairly good understanding of the customer base, the problems that need to be solved and the approaches and means we can employ to solve them. Traditional R&D activities and competitor acquisitions are usually effective.
The second type identified is Breakthrough Innovation. Here the problems are well understood but the solutions are not and tend to be complex. We may need to explore unconventional skill areas and consider the problems within the paradigms of an adjacent business domain. Examples here could be the breakthrough in manufacturing processes from the assembly line, or the transformation of business operations introduced by Uber.
The concept of Disruptive Innovation was introduced by Clayton Christensen in his book The Innovator’s Dilemma2. This is the innovation type that leads to the kind of big ideas many will have in mind as what signifies innovation. These disrupt a market and render existing solutions redundant and obsolete. They transform the value proposition attracting previously marginal customers. A common example here is the iPod, which changed the way that music is bought and consumed.
The final type of innovation Sattle defined is Basic Research. This is needed when neither the problem space (or market) nor the solution is well enough. The path breaking innovations do not appear fully formed. Basic Research is the application of science, the discovery of new knowledge, new principals and new concepts. If you like, this is the realm of invention.
Attribute definitions should be easily, even implicitly, understood by their audience. Let us restrict ourselves to breakthrough and disruptive innovation. We can now define Innovative.
Innovative: Driving top line business growth by creating, and exploiting, new business models and markets.
While this reflects the intent of innovation, it could be viewed as missing the reality that a significant proportion of innovation projects fail. Or rather, they fail to have an impact on the business’ top line financials. When it comes to measuring the performance of our Innovative attribute it is important to consider the value that can be drawn from even these failed project – what was learnt. This applies not only to knowledge gained but also to optimising the innovation processes themselves to ensure when we fail this is recognised as early as possible.
Measurement should also reflect the different stages of an innovation project. Whichever of the following questions we’re focussed on will naturally lead to different measures.
- Is the problem worth solving?
- Does our solution solve a customer’s problem?
- Will the customer buy our solution? What model will they accept?
- Can our solution scale?
- How can the business adopt our solution and transition to BAU?
Our metrics also need to identify the best projects to invest in, e.g. those that are aligned to the business strategy, while not promoting activities disproportionately targeting continued investment over innovation progress.
As you can see, the real challenge here is not so much to define what being innovative is, but to measure it. Innovation does not easily map to financial accounting methods – the assets created and refined are often intangible and at least in early stages hard to quantify in monetary terms.