Help people predict what the system will do.
01 / UNDERSTAND
A mental model is someone’s working explanation of how a system behaves. It lets them predict consequences before they act: whether sharing a link grants access, whether archiving removes a message permanently, or whether two database views contain the same records. A familiar icon can help, but an accurate prediction matters more than familiarity alone.
People build these explanations from labels, prior experience and the feedback a product gives them. Problems appear when the interface suggests one rule and the system follows another. A person may assume a filtered view is a separate copy, for example, and delete records that still matter elsewhere. Research the explanation people are using, then make ownership, scope and consequences visible. If the system must behave unexpectedly, explain the difference before a consequential action rather than waiting for an error to teach it.
02 / RECOGNISE
Examples of how the principle can appear in an interface.
Google Drive’s sharing dialog presents people and access controls alongside the link. This makes a crucial distinction available at the decision point: possessing a URL and having permission to read or edit the file are different things. Test whether users can explain that distinction before sharing.
Notion allows a database to appear as a table, board, calendar and other views. Switching the view changes how the records are presented, rather than creating an independent copy. A clear shared database model helps people predict how editing an item affects the other views.
03 / CONNECT
Investigate mistaken predictions, unintended changes and successful explanation of permissions, rather than measuring familiarity alone.
Possible influences to investigate, rather than guaranteed outcomes for every product.
04 / APPLY
Ask users to predict an action before they perform it
Show the scope of sharing, editing and deletion
Explain exceptions beside the affected control
05 / EXPLORE WITH UPSLIDE
Use a practical guide, read a deeper explanation, or examine a related Upslide project.
Ask people to predict what the system will do and compare that prediction with actual behavior.
Investigate domain language and expectations before choosing labels and workflow concepts.
Examine how a conversational service communicates its purpose and next actions.
CONNECT THE IDEAS
Keep exploring
Keep essential information available instead of asking people to remember it.
