Works with Claude, ChatGPT, Gemini, or your own local model: paste this prompt and your AI will co-facilitate this exercise with you.

Help me prepare and run the exercise below. First ask for my participant count, time, setting, access needs and intended result. Check these against the stated limits. Use the supplied rules and materials. Explain any proposed changes before using them, and label them as adaptations. Do not invent source claims or require personal disclosure. Give me one step at a time when I say start. I manage the people, physical activity and clock; do not claim to observe the room. Ask me what happened before choosing a next step.

Reference revision: 058f8c291aa6a86ec120d88b0b1b8680b0e9cae9af19467c38c273f4fbf27b92

# Custom Retrospective

Teams co-design or select tailored reflection categories that fit their immediate context, then silently brainstorm, cluster, vote, and select concrete action items.

**Time:** 50 minutes. UWT suggested standard sprint retrospective duration; the source describes a software workflow without fixed clock allocations.
**People:** 4 to 16. Solo reflection followed by plenary clustering, voting, and discussion.

## Choose this exercise

**Purpose:** Gather targeted feedback using custom prompts tailored to a team's specific working context.
**Use it when:** Use when standard sprint retrospective formats feel stale, misaligned with the current work context, or need a hybrid combination of existing reflection frameworks.
**Skip it when:** Skip when a team is conducting its very first retrospective and needs the simplicity of an established off-the-shelf structure.

## Materials and tools

- Physical sticky notes or digital cards (at least 10 per person)
- Markers or pens for each participant
- Whiteboard, blank wall, or digital collaboration board divided into 3 to 5 custom category columns
- Dot stickers or digital voting pins (3 per person)
- Countdown timer

## Before participants arrive

1. Before the session, the facilitator selects 3 to 5 custom category prompts based on the current sprint context, or asks the team asynchronously to vote on a tailored combination (for example, combining Sailboat and Start-Stop-Continue into: Anchors, Wind, Start, and Applaud).
2. Label column headers clearly across the physical board or digital canvas.
3. Ensure all participants have access to notes and voting markers.

## Say this to open

Welcome. Standard retrospective prompts do not always capture what our team needs to examine right now. Today we are using a custom set of prompts tailored to our current cycle. You will silently write observations for each column, then we will cluster similar ideas, vote on key areas, and agree on tangible actions.

## How to run it

### 1. Introduce custom categories (5 minutes)

Read the opening script. Walk through each column header. Clarify the boundary and intended meaning of each prompt using a concrete example. For instance, if using Anchors and Wind: explain that an Anchor is something slowing progress down (example: 'Waiting three days for environment approval'), while Wind is what pushed the work forward (example: 'Pair programming on the API migration'). Answer quick questions about scope.

### 2. Silent reflection and note writing (10 minutes)

Participants work individually and silently. Everyone writes one idea per sticky note or digital card and places it into the corresponding category column. Anyone who chooses not to submit notes may simply observe.

### 3. Group and cluster themes (15 minutes)

Give the room 3 minutes to silently scan all posted cards. Then, invite everyone to actively drag or move related cards together in parallel rather than reading each card sequentially. The facilitator reads aloud emerging groupings, merges duplicates with group consent, and writes a concise, descriptive label on each cluster. Stop clustering at the 15-minute mark.

### 4. Vote on priorities (5 minutes)

Each person distributes up to three votes across the clusters or individual cards they consider most important to address. Voting can be done by placing dots on physical cards or using an anonymous digital voting tool.

### 5. Discuss and assign actions (15 minutes)

Take the highest-voted clusters in order. Spend no more than 5 minutes on each top cluster to discuss root causes and concrete adjustments. For each chosen topic, write an action agreement with an assigned owner and completion check date.


## Debrief questions

- Did these custom categories capture aspects of our work that standard templates usually miss?
- Which action item from today will have the most direct impact on our next cycle?
- What categories should we keep, change, or remove next time we customize our retro?

**Finished output:** A board of categorized team observations and a prioritized list of agreed action items with assigned owners.

## Remote

Run via video conference using a shared digital board (such as Parabol, Miro, or Mural). Set category columns in advance. Keep cards masked or private during silent ideation, then reveal them simultaneously before clustering and dot-voting.

## Hybrid

In-person and remote participants must use the same digital board to ensure all cards, clusters, and votes are equally visible and editable in real time. Remote participants speak over the room audio system during clustering and action planning.

## Access and participation choices

- Participants may submit notes anonymously or choose to contribute verbally without writing.
- Any participant may pass on placing votes or speaking during the plenary discussion without explanation.

## If the session gets stuck

**Participants write vague or generic notes that fit every category.**

Ask the author or group to provide one specific incident from the sprint that illustrates the observation.

**The group spends all discussion time on the top issue and cannot cover others.**

Set a strict 5-minute timebox per prioritized cluster. Move to the next topic once a single actionable task is defined.


## Terms used here

- Retrospective: A recurring meeting where a team reflects on recent work to agree on process improvements.
- Clustering: Sorting scattered feedback cards into related topic groups.

## Sources and adaptation

Observed in product template documentation published by Parabol.

Retained the custom retrospective concept and retro phase flow (reflect, group, vote, discuss) from the Parabol template guide. Replaced platform-specific UI click instructions with standard, platform-agnostic facilitation steps, timing, concrete examples, and practical ground rules.

[Parabol](https://www.parabol.co/templates/sprint-retrospectives/custom-format/)

[Parabol](https://parabol.co/templates/sprint-retrospectives/custom-format)

[Parabol](https://www.parabol.co/templates/custom-format)

[Parabol](https://www.parabol.co/templates/sprint-retrospectives/custom-format)

Current run sheet: https://unitedwetransform.com/exercises/custom-retrospective/