13. Terminology
AARRR Metrics: Acquisition, Activation, Retention, Referral, Revenue - a set of metrics defined by Dave McClure in 2007 to determine whether online services are performing well as a whole.
BDD: see Behavior-driven Development
Behavior-driven Development: an agile software development practice that encourages collaboration among developers, QA and non-technical or business participants in a software project. Popularized by Daniel Terhorst-North from 2006 onwards.
CFD: see Cumulative Flow Diagram
Cumulative Flow Diagram: metric to observe flow in our team and spot impediments, typically implemented by plotting the amount of work in each state of our process over a period of time.
Estimating: in the context of this book, it’s the common style used by teams to predict how long a piece of work is going to take, relying on people’s informed best-guess. Usually based on techniques like story points, velocity, t-shirt size, ideal days, or similar.
Flow: movement and delivery of customer value through a process (Vacanti2015). It’s how consistently and continuously work moves through our process all the way to a customer.
Flow Efficiency: percentage of time that work spends being actively worked on, as opposed to the total amount of time the work takes, including wait time. A high flow efficiency means that work flows through our process as fast as it possibly can, with minimal wait time or delay.
Forecasting: the prediction of work completion, as well as many other business questions, based on the use of historical data to simulate what might happen in the future. The results are expressed as a list of possible outcomes with their likelihood.
Lead Time: the time that a piece of work takes to go from start (commitment point) to end (last state of influence) of a process. Note: we are using the names that seem to be most common in the Kanban community, but be aware that different people use different terms when meaning the same things. Most famously, Dan Vacanti calls it “cycle time” in his excellent book “Actionable agile metrics for predictability” (Vacanti2015).
Little’s Law: states that \(averageLeadTime = \frac{averageWIP}{averageThroughput}\). In simple terms it means that, on average, if you want stories to go faster (shorter lead time), you either have to increase the throughput (get more done) or reduce the amount of work in progress (work on less things at the same time).
Metrics: in the context of this book, the measure of some aspect of the team’s work. They help us gather insights, answer questions and enable the team to make better business decisions. We split them into value, quality and operational metrics.
OEM: Original Equipment Manufacturer.
Quality: conformance to requirements, as defined by Philip B. Crosby’s book “Quality is Free” (Crosby1987).
Pirate Metrics: see AARRR Metrics PO: Product Owner.
RAS model: Reliability, Availability, Serviceability - model developed by IBM (RAS) to look at different dimensions of quality.
Story Points: a measure of the perceived size of a user story. For example, something that is perceived as simple and quick to implement might be allocated a single story point, whereas something complex and long might be given ten story points.
Throughput: the number of work items that were completed during a particular period of time (e.g. in a sprint, or an iteration). Often also called delivery rate, or story count.
Value: why we do what we choose to do. Metrics for value help us focus on the things that matter, allowing us to not waste time and do something worthwhile.
Velocity: number of story points that the team completes in a sprint. Often used to make predictions about what the team can deliver, e.g. to decide how many stories they can include in the next sprint, or to determine how many sprints it will take to complete a given set of user stories.
WIP: see Work In Progress
Work In Progress: any piece of work that has been started but has not been completed yet. High amounts of work in progress typically results in a number of problems, including slower progress (Little’s Law), more time wasted in context switching, and lower quality.