methodatlas

Agile Decision method

Affinity Estimation

Affinity estimation lets a team group items silently, discuss differences, and produce consistent relative size views.

Affinity estimation places tasks relative to known reference items and condenses the assessment into clusters. Shared sorting uses pattern recognition and makes outliers visible for focused clarification.

Core question
How can the team quickly order many comparable items by relative size and discuss only genuine uncertainty?
Plan Affinity EstimationMethod guide
Start comparison
Give feedbackIs this method useful to you?
Participants
3-12
Time
30–60 minutes for 20–50 comparable items; analyse disputed items afterwards.

The practice is repeated across comparable work items.

Paper illustration for Affinity Estimation.

  1. Frame

    Comparison set

    Set comparable items and one shared meaning of size.

  2. Anchors

    Size anchors

    Place a known small, medium, and large item as relative orientation.

  3. Sort

    First sort

    Team members silently place items relative to anchors; an item may be moved once.

  4. Conflicts

    Clarification pile

    On a second move request, put the item in discussion; resolve differences in scope and risk.

  5. Calibrate

    Relative estimate

    Name groups, inspect boundaries, and split or refer large and unclear items.

Add new evidence at review and version estimates with the prior state.

A good fit when

Wenn viele Aufgaben zügig eingeordnet werden müssen, sortiert sie nach wahrgenommenem Aufwand und Ähnlichkeit. Sie reduziert den Aufwand für Feinschätzung bei großen Mengen.

Ideal for

  • Large backlogs
  • Release planning
  • Team calibration

Not good for

  • Single high-risk items
  • Very unclear requirements
  • Detailed cost estimation

Outputs

  • Relative estimate

    Relatively ordered and calibrated items with references, marked conflicts, assumptions, and split or clarification actions.

Affinity Estimation · working structure

Fictional example

Fictional values show the actual method fields.

Affinity Estimation · working structure
ItemIncludedReferenceFirst placementConflictOutcomeFollow-up
Delete tokenAdd evidence and rationale.Build, test, security, rolloutSmall anchor: validation ruleSmallNoneSmallReady
Save cardAdd evidence and rationale.Build, test, security review, rolloutMedium anchor: provider tokenMedium/LargeGuest users and migration unclearMediumScope fixed
Provider migrationAdd evidence and rationale.Auth, migration, fallback, certificationLarge anchor: new payment methodBeyond largeSeveral outcomesClarification pileSplit + spike

  1. 1. Included

    Delete token
    Build, test, security, rollout
    Save card
    Build, test, security review, rollout
    Provider migration
    Auth, migration, fallback, certification

  2. 2. Reference

    Delete token
    Small anchor: validation rule
    Save card
    Medium anchor: provider token
    Provider migration
    Large anchor: new payment method

  3. 3. First placement

    Delete token
    Small
    Save card
    Medium/Large
    Provider migration
    Beyond large

  4. 4. Conflict

    Delete token
    None
    Save card
    Guest users and migration unclear
    Provider migration
    Several outcomes

  5. 5. Outcome

    Delete token
    Small
    Save card
    Medium
    Provider migration
    Clarification pile

  6. 6. Follow-up

    Delete token
    Ready
    Save card
    Scope fixed
    Provider migration
    Split + spike

All values are fictional.

Reference · original teaching example

First step

Set item set, included work, and shared Definition of Done; remove unclear or unlike items first.

Open the method guide

Where to go next?

Story Splitting method illustration showing its working structure
Agile
Story Splitting

Splits items that remain too large or unclear into smaller slices.

Paper illustration for Planning Poker.
Agile
Planning Poker

Later calibrates rough groups with a more precise team estimate.

Deep dive

Purpose

When many tasks need to be classified quickly, it sorts them by perceived effort and similarity. It reduces the effort of fine-grained estimation for large volumes.

How it works

It has tasks arranged in comparison to already known sizes and condenses the assessment into clusters. The shared sorting uses pattern recognition instead of individual debate and makes outliers visible exactly where they actually need to be clarified.

In detail

Affinity Estimation transfers affinity grouping into estimation. Instead of discussing each item at length, the team sorts many cards relative to one another from small to large. Then the clusters are labeled with sizes such as T-shirt sizes or story point buckets. Discussion happens mainly where items are unexpectedly placed or where several people have clearly different assumptions.

Facilitation

Prepare all items as cards and a visible size axis. Let the group sort silently first to avoid long debates. Discuss only outliers, unclear items, and cluster boundaries, then document assumptions and slicing candidates.

The method guide walks execution through 5 phases, timing, 5 pitfalls, and clear stop criteria.

Open method guide

Your perspective

Does this method help with your work?

Your feedback helps shape the next improvements to methodatlas.

Sources and topics

Topics
EstimationBacklogRelative sizing
Sources
Affinity Estimationmountaingoatsoftware.comMountain Goat Software by Mike Cohn; describes Affinity Estimation as a fast, largely silent estimating technique.
In playbooks
Improve Estimation and ForecastingMove from intuition to shared sizing, uncertainty, and delivery forecast.
Artifact templates
Affinity Estimation: WorksheetSet subject, data, roles, flow, and review.
markdown

affinity-estimation-working-template.md

Set subject, data, roles, flow, and review.

Affinity Estimation: Worksheet

Question

To be confirmed

Desired outcome

To be confirmed

Scope

Practice run Time: 30–60 minutes for 20–50 comparable items; analyse disputed items afterwards.

Complete this worksheet on paper or in your own document during the work.

Preparation for this scope

Method setup

  • Materials: Comparable refined items, shared Definition of Done, relative scale, reference items, and discussion pile
  • Roles: Delivery team sorts · product owner clarifies · facilitator protects silent flow
  • Advance information: Set item set, included work, and shared Definition of Done; remove unclear or unlike items first.
  • Overall time needed: 30–60 minutes for 20–50 comparable items; analyse disputed items afterwards.
  • Setup: Prepare a workspace with Frame · Anchors · Sort · Conflicts · Calibrate; separate example from live data.

Prepare for the work steps

Frame: Comparison set

  • Set item set, included work, and shared Definition of Done; remove unclear or unlike items first.
  • Comparable refined items, shared Definition of Done, relative scale, reference items, and discussion pile

Anchors: Size anchors

  • Comparison set

Sort: First sort

  • Size anchors

Conflicts: Clarification pile

  • First sort

Calibrate: Relative estimate

  • Clarification pile

Affinity Estimation · working structure

ItemIncludedReferenceFirst placementConflictOutcomeFollow-up
Delete token
Save card
Provider migration
  • Delete token: Add evidence and rationale.
  • Save card: Add evidence and rationale.
  • Provider migration: Add evidence and rationale.

The blank worksheet contains prompts only.

Reference · original teaching example

Frame: Comparison set

Expected artifact: Comparison set

Set comparable items and one shared meaning of size.

Entry:

...

  • Do all items include the same work types and Done standard?

Anchors: Size anchors

Expected artifact: Size anchors

Place a known small, medium, and large item as relative orientation.

Entry:

...

  • Are anchors understood and from the same work domain?

Sort: First sort

Expected artifact: First sort

Team members silently place items relative to anchors; an item may be moved once.

Entry:

...

  • Was placement independent before discussion or numeric anchoring?

Conflicts: Clarification pile

Expected artifact: Clarification pile

On a second move request, put the item in discussion; resolve differences in scope and risk.

Entry:

...

  • Does discussion expose different work models rather than preferences?

Calibrate: Relative estimate

Expected artifact: Relative estimate

Name groups, inspect boundaries, and split or refer large and unclear items.

Entry:

...

  • Are neighbourhoods plausible and open items assigned a next step?

Open questions and next steps

...

Method guide: https://methodatlas.meierhoff-systems.de/en/methods/affinity-estimation/run-sheet

Alternatives and neighbors

Fitting alternatives with decision guidance, followed by other related methods.

All methods
Paper illustration for Planning Poker.
Agile
Planning Poker

Instead of Affinity Estimation, choose Planning Poker when you want to calibrate estimates relative to each other and surface shared assumptions.

Paper illustration for Wideband Delphi.
Decision Making
Wideband Delphi

Instead of Affinity Estimation, choose Wideband Delphi when you want to make estimates more robust with several rounds of experts.

Paper illustration for Bucket System.
Agile
Bucket System

Moves backlog, sprint work, and team flow toward a concrete result through "prepare buckets and scale", "place reference items", and "flag large or unclear items".

Paper illustration for Story Points.
Agile
Story Points

Moves backlog, sprint work, and team flow toward a concrete result through "choose reference stories", "define the scale", and "interpret velocity only within the team".

Paper illustration for T-Shirt Sizing.
Agile
T-Shirt Sizing

Sorts backlog items roughly by reference sizes and flags early on which items will later need splitting or clarification.

Paper illustration for Ideal Days.
Agile
Ideal Days

Moves backlog, sprint work, and team flow toward a concrete result through "explain the ideal day unit", "calibrate with a reference item", and "do not confuse it with calendar days".

Constraints and origin
MediumWorkshop30-90 min

Profile

Method form
Decision method
Participants
3-12
Time
30-90 min
Maturity
Established
Origin
Agile Community
Story Splitting method illustration showing its working structure
Next up: Story Splitting