Learning through experience - experimentation is based on David Kolb’s model of how adults learn.

Experimentation can lead to innovation. Trying different variants and making unorthodox connections can in effect give an innovative solution.

During experimentation making mistakes becomes a teaching factor. However, a significant amount of mistakes can be of value, if these mistakes appear in the experimentation process instead of the process of implementation.

The founders of start-ups usually focus on looking for mistakes or clients’ needs that they could solve or satiate, expanding the service or product to generate revenue in the future. They test ideas for businesses with multiple groups of potential users, to better meet them and figure out how they can adequately answer the needs of these potential users. To that end they create prototypes - these are among other things dummy products, mock-ups of web tools, and simulations of services.

Prototypes are low-cost and contain the bare minimum of the characteristics of the planned product or service to create a context in which potential users interact with it. The Lean Startup method is based on cyclical tests of multiple aspects of a business idea, created by Eric Riese, it gained great popularity in the worldwide start-up and investor environment. 

Experimentation allows one to check how something (a device, method, recipe, algorithm) works in practice. It allows one to check a lot of theses in practice.

Prototyping and experimentation are widely used in the service design industry which designs, among other things, client interactions with both on and offline services. Tools available on the Service Design Tools website may be helpful in the experimentation process.

Experimentation can also have a cognitive character, not used to check specific theses but allowing for the learner to experience something new - contact with a new program, working in different conditions, etc.

Companies can lead experiments on a small scale e.g. ones related to ways of knowledge management. Testing multiple ways of working in groups or leading team meetings (e.g. unconference, short standing meetings, Pecha Kucha) and periodic measurement of results can verify hypotheses referring to the effectiveness of the flow of information and knowledge in a company and in larger perspective influence a change of the whole organization's culture.

An experiment can be a form of diagnosis. For example, someone can be put in a specific situation which allows for them to see how they feel in it and whether they can deal with it or not.

In undertakings that have a goal of social development the Human Center Design method works well, it is based on testing possible system solutions with active user participation from local communities, groups in danger of exclusion, etc. Experiments have the goal of showing the source of a problem or a mechanism the modification of which may influence the search for a long-term solution. The testing of hypotheses also pertains to the method of implementing the solution e.g. the method of vaccine distribution or taking payment for clean water.