Activation

The first observable event that shows a new account has completed the product’s core useful job.

Example: for an invoice-import product, activation could be a user importing a file, resolving errors and exporting the approved result. Creating an account is not activation.

Do not confuse it with onboarding: onboarding is the path and assistance around first use; activation is the useful event reached through that path.

Churn

The loss of a customer, subscription or recurring revenue during a defined period.

Example: a customer who cancels and stops paying has churned. A customer who skips a week in a monthly workflow has not necessarily churned.

Do not confuse it with low engagement: engagement describes product activity; churn describes the end or loss of the commercial relationship.

MVP

A minimum viable product is the smallest product version that lets an intended user complete the core job well enough for you to evaluate the workflow and offer.

Example: an import tool that accepts one supported file format, reports invalid rows and exports a corrected file may be an MVP. A clickable mock-up that cannot process a file is not.

Do not confuse it with a prototype: a prototype explores an interface or technical idea; an MVP supports a real end-to-end use.

No-code

A way to build software mainly by configuring visual components, data models and workflows instead of writing the application in a general-purpose programming language.

Example: connecting a form, database and approval workflow in a visual builder is no-code. Writing a custom backend and merely using a visual page editor is not.

Do not confuse it with low-code: low-code expects configuration plus custom code for parts of the product; no-code aims to keep custom code optional.

Pricing

The rules that determine what a customer pays, when they pay and which usage or access changes the amount.

Example: a monthly charge per active client account is pricing. Calling one plan “Professional” without stating its charge or billing rule is packaging language, not pricing.

Do not confuse it with packaging: packaging groups access, limits and service into offers; pricing assigns money and billing conditions to those offers.

Value metric

The unit used to connect a customer’s charge or plan boundary to the amount of useful work the product handles.

Example: charging by active managed location uses locations as the value metric. Charging more only because a user wants dark mode does not.

Do not confuse it with a price point: the value metric is what the charge varies with; the price point is the amount charged.

MRR

Monthly recurring revenue: the recurring subscription revenue normalized to one month, excluding one-off setup fees and other non-recurring payments.

Example: a monthly subscription contributes its recurring monthly amount to MRR. A one-time migration fee does not.

Do not confuse it with cash received: cash received follows payment timing and can include annual prepayments or one-off fees; MRR normalizes recurring revenue to a monthly view.

RLS

Row-level security: database rules that decide which rows a request is allowed to read, insert, update or delete.

Example: a policy can restrict a signed-in user to records belonging to their account. Hiding another account’s records in the interface is not RLS.

Do not confuse it with authentication: authentication establishes who made the request; RLS applies data-access rules to that request.

Automation

A defined sequence in which software performs repeatable steps after a trigger, with explicit inputs, outputs and failure handling.

Example: validating an uploaded file, creating a job and notifying the owner when processing fails is automation. A reminder for a person to do all three steps manually is not.

Do not confuse it with integration: an integration connects systems or exposes data; automation decides and executes the sequence that uses those connections.

AI feature

A product capability that uses a model to generate, classify, extract or transform an output whose exact result is not fully predetermined by fixed rules.

Example: drafting a reply from a support ticket is an AI feature. Sending a saved template when a status changes is deterministic automation, not AI.

Do not confuse it with automation: automation follows defined steps; an AI feature introduces model output that needs its own quality checks and fallback.