Free text
Words with three or more characters are matched against name, category, method form, summary, purpose, ideal-for, and tags.
How your recommendation is built
You describe your situation in the recommendation and get a ranking of matching methods. This page explains which signals are used, how they turn into a score, and how to read and improve the result.
Looking for the usage guide instead? Start with Help: Recommendation and Playbooks.
What you enter
Six signals describe your situation. All of them are optional, but every additional signal makes the ranking more precise.
Goal
Each goal type brings search terms and preferred categories.
Phase
Discover, Define, Design, Deliver, Operate, and Grow add phase terms.
Format
Workshop, Async, and Both are compared with the requested mode.
Complexity
Low, Medium, and High can be preferred directly. Any leaves all levels open.
Team size and time
Structured duration and group-size ranges are checked against your situation. Unclear values remain unconfirmed.
An example before it gets technical
Your input
Goal: Decide or document architecture. Free text: Architekturentscheidung dokumentieren. Format, complexity, and time stay open.
What happens to Architecture Decision Record
- The category Architecture is one of the preferred goal categories.+8
- Your free-text word hits the method profile.+4
- The goal terms architecture and documentation hit name and tags.+6
- Open format and complexity add five base points together.+5
Category, matching terms, and constraints contribute rank points. Confirmed constraints take priority; the exact score depends on your input.
The rule base in detail
All methods are scored with the same rule base. This table is the complete reference for all point values.
SignalScoreLogic
Goal category matches+8Strongest anchor when the method belongs to one of the preferred goal categories.
Goal category misses-3Small penalty when the goal has defined categories and the method sits outside them.
Free-text hit+4 per hitUser words carry high weight because they express the concrete situation.
Goal terms+3 per hitTerms from the selected goal type, such as discovery, roadmap, architecture, or incident.
Phase terms+2 per hitTerms from Discover, Define, Design, Deliver, Operate, or Grow.
Format matches+3Both matches Workshop and Async. Any keeps the mode open.
Format misses-2Penalty when a concrete format was selected and the method works differently.
Complexity matches+2Applies when Low, Medium, or High was selected, or when Any is active.
Complexity misses-1Small penalty because a higher or lower complexity may still be a conscious choice.
Team size fits+2The method must cover the full selected team-size range. Research samples do not count as team sizes.
Timebox fits+2The full known duration must fit the budget: short up to 60 minutes, half-day up to 240 minutes, multi-day without an upper limit.
Unknown time value0Unknown, partially matching, or conflicting durations add no duration points.
How to read the result
- Only methods matching the category, goal, or free text and scoring above two rank points appear.
- At most 12 methods are shown.
- Methods with fully matching constraints come first, followed by rank points and method name.
- Rank points describe relative ordering. They are not a probability of success.
- Visible reasons show the most important hits, not every internal scoring component.
- Warnings identify unconfirmed duration, team, format, or complexity constraints. The strict filter hides these methods.