An Eight-Layer Model For Exploratory Testing - Steve Green
Exploratory testing is often represented as being unplanned, unstructured, random and unrepeatable, and that it mostly finds obscure edge cases with unrealistic user behaviour and data. Ten years ago this would have been fair comment, but it’s not now.
Developed over the last decade, our 8-layer model is a framework (not a process!) for exploration that finds bugs with the simplest sequence of events and most vanilla data that can cause them, minimising the need for lengthy diagnosis. These are generally realistic, repeatable bugs that project managers will want to fix.
When used in conjunction with appropriate note-taking, this approach provides a means of planning and tracking test coverage and conveying to other people what has and has not been done.