Scrum is a framework not only for producing product, but also for constantly advancing the abilities of the team creating that product. From a process perspective, Scrum is all about improvement, which, in Japanese, is just “改善.” In its own process improvement effort, Toyota took kaizen as a slogan, usually written as “カイゼン.” The international community picked up the term and started using it in the sense they projected from Toyota’s programs: continued, relentless improvement. The next pattern, ScrumMaster, describes the role that owns the Scrum process in a Scrum team and which often leads the charge for kaizen. However, it’s important to remember that kaizen is everyone’s job: the ScrumMaster is just the team’s kaizen conscience.
Kaizen (改善) is the Japanese word for “improvement.” In business, kaizen refers to activities that continuously improve all functions and involve all employees from the CEO to the assembly line workers. It also applies to processes, such as purchasing and logistics, that cross organizational boundaries into the supply chain. It has been applied in healthcare, psychotherapy, life-coaching, government, and banking.
. . .
Kaizen is a daily process, the purpose of which goes beyond simple productivity improvement. It is also a process that, when done correctly, humanizes the workplace, eliminates overly hard work (“muri”), and teaches people how to perform experiments on their work using the scientific method and how to learn to spot and eliminate waste in business processes. 
The team can use kaizen opportunistically or to remedy known shortcomings. It is a Japanese workplace tradition that upon noticing a personal or team shortcoming, the team does hansei: that is, they take a posture of individual or collective regret for the shortfall. Here, kaizen is related to restoring self-worth and honor, and reflects the sense behind Greatest Value. There is a Toyota saying that “there is no kaizen without hansei.” So you would usually approach kaizen not with celebration, slogans, and excitement, but with humility and resolve. See Pop the Happy Bubble.
The term kaizen usually implies incremental improvement. Much of agile is based on taking small steps, as in One Step at a Time. The problem with taking only incremental steps is that you can land in a local optimum. Occasionally taking greater risk and making a different kind of change—often a larger change—can take you to higher ground. This is a kind of radical change, which in Japanese is kaikaku (改革): an extreme kaizen.
In fact, it is likely that most improvement comes from incremental change. The British bicycling team (Team Sky) popularized the notion of marginal gains in 2003, showing that many tiny improvements in bike design and training add up. Though marginal gains are overhyped, the approach certainly merits serious consideration in any improvement effort. Financial Times says:
...most productivity growth came from existing companies improving existing products, rather than new businesses or products. That was true in the 1980s and still is. 
Toyota’s employee suggestion box has yielded over 1,800 kaizens for factory line efficiency—a form of marginal gains. But they can become a kind of diminishing return, begging more radical change to attain the next level of excellence.  To keep moving ahead requires occasional discontinuity, and Financial Times adds:
Google may A/B test its way to greater profits [by fine-tuning web page color schemes—a small change that supported higher engagement and resulting higher profits] but the company’s success has been built on a leap forward in search technology in the late 1990s, and even more fundamentally on the publicly funded efforts to develop the web and the internet. 
They add that kaizen is likely to take care of itself, and hint that more radical change may require an external disruption. Great Scrum teams alternate radical and incremental change to sustain what might be called continuous discontinuity. Kaikaku changes often originate with management, as they may entail change to the corporate governance or structure—changes at the enterprise or system level. See Involve the Managers.
As an example of diminishing returns from kaizen, consider how Toyota treats poke-yoke (a “failsafe” mechanism that guides the work in a way that avoids errors; see Emergency Procedure). Each failsafe brings with it overhead which over time may add up to unacceptable waste—especially if it leads to complacency that dulls team members’ awareness and prevents them from suffering the occasional priceless setback that leads them into learning. The team must be circumspect about this tradeoff.
Some patterns that we might casually include in the set of incremental changes may in fact require kaikaku action. For example, it may be difficult to transition from functional teams or component teams to Cross-Functional Teams in a piecemeal fashion. Organizations might best be advised to put the right structure in place up front with management support. Experience shows that it does not happen of its own accord, especially in large organizations. Other patterns may be the same; each organization can judge for itself.
Russ Ackhoff, pioneer in the field of operations research, systems thinking and management science, admonishes us to look beyond “continuous improvement” to systems thinking that deals with the Whole. The alternative to continual improvement is discontinuous improvement, focusing on properties of the Whole, through creativity that breaks with what preceded it. 
 —. “Kaizen.” Wikipedia, https://en.wikipedia.org/wiki/Kaizen, 7 September 2018, (accessed 3 October 2018).
 Tim Hartford. “Marginal gains matter but game changers transform.” Financial Times, https://www.ft.com/content/6b5c046c-0fc0-11e7-a88c-50ba212dce4d, 24 March 2017 (accessed 17 December 2017).
 —. “Challenging Marginal Gains Mathematics: 10 Myths about How Marginal Gains Add Up.” https://www.youtube.com/watch?v=28E9lzSmFvI, 12 May 2017 (accessed 17 December 2017).
 Russ Ackhoff. “If Russ Ackoff had given a TED Talk...” Video recording, https://www.youtube.com/watch?v=OqEeIG8aPPk, 1994 (accessed 25 December 2017).