Esimation in SAFe

24/04/2025

In SAFe 6.0 (Scaled Agile Framework), estimation techniques are used to forecast the effort or size of work items at various levels (Team, ART, and Portfolio). Estimation helps teams and stakeholders make informed decisions, plan effectively, and align on priorities.

Here are the primary estimation techniques used in SAFe 6.0:

👕 T-Shirt Sizing in SAFe – The High-Level Estimator

What it is:
T-Shirt Sizing is a quick and intuitive way to estimate large work items like Epics or Features. Instead of assigning story points or hours, teams use relative size categories like XS, S, M, L, XL.

Why it's used:
When you're evaluating big-picture work early in the Portfolio Kanban (like analyzing an Epic), you don't yet know the full scope. T-Shirt sizing lets teams make fast trade-offs between initiatives and gauge the overall effort without wasting time on fine-grained estimation.

Where it fits in SAFe:
Primarily during Lean Portfolio Management activities, such as:

Epic intake and prioritization
Evaluating feasibility or forecasting rough capacity
Quick sizing during epic hypothesis creation

How it works:

Epic Owners or Product Managers briefly present the work item.
The group discusses it for context.
Everyone gives a quick T-shirt size estimate (either silently or by discussion).
Use historical mapping to translate sizes to rough point ranges if needed.

Example mapping (optional):

Small (S): ~30–50 story points
Medium (M): ~50–100
Large (L): ~100–200
Extra Large (XL): 200+

Think of it like a rough sketch before an architect draws up the blueprints.

🧩 Affinity Estimation in SAFe – The Fast Sorting Method

What it is:
Affinity Estimation is a collaborative, fast-paced estimation technique used to sort a lot of stories or features into groups based on relative complexity or effort, and then assign story points.

Why it's used:
When a team or ART has a large backlog (like before PI Planning), Affinity Estimation helps the group estimate dozens of items quickly and with broad alignment—without going into lengthy planning poker sessions.

Where it fits in SAFe:

During PI Planning to estimate features
In iteration planning or backlog refinement sessions
In team settings or across multiple teams in an ART

How it works:

Write user stories or features on cards (physical or digital).
Silent Sorting – Team members take turns placing cards from left (smallest) to right (largest) on a wall or board, without discussion.
Collaborative Adjustment – After everyone's placed some cards, team discusses outliers or adjusts positioning.
Story Point Overlay – Once the order is agreed upon, overlay a Fibonacci scale (1, 2, 3, 5, 8, etc.) and assign points to each group based on relative sizing.
Sanity check – Review for consistency and adjust if necessary.

Key advantage:
It's fast, often letting teams estimate 20–30 backlog items in 15–20 minutes, while still maintaining consensus.

🧠 So which should you use?

Use T-Shirt Sizing when you're:
- Evaluating Epics and Features early
- Needing a rough, high-level estimate to guide prioritization or feasibility
- Working at the portfolio or large solution level
Use Affinity Estimation when you're:
- About to enter a Program Increment (PI) and need to size work
- Planning multiple features or stories across teams
- Looking for a fast, team-inclusive estimation method that leads to actual story points

🃏 Planning Poker ?

Planning Poker, also called Scrum Poker, is a consensus-based estimation technique used by Agile teams (including SAFe teams) to estimate the effort or complexity of stories or features. It combines expert opinion, analogy, and discussion into a game-like format that encourages participation and alignment.

In SAFe, Planning Poker is typically used:

At the team level (Agile teams)
During Iteration Planning or Backlog Refinement
To assign Story Points to User Stories (not Epics)

🎯 Why Planning Poker?

Encourages engagement and prevents anchoring (where one person influences the whole group)
Surfaces disagreements or misunderstandings about requirements
Helps teams reach consensus quickly and fairly
Works well for cross-functional teams

🛠️ How Planning Poker Works (Step-by-Step)

Preparation:
- The Product Owner presents a user story to the team.
- Acceptance criteria and story details are clarified.
Discussion:
- The team discusses the requirements, risks, and complexity.
- Everyone ensures they understand what's being estimated.
Select a Card:
- Each team member privately selects a card from a deck representing story points (typically Fibonacci numbers: 1, 2, 3, 5, 8, 13, 21…).
Reveal:
- All cards are revealed simultaneously to avoid influence.
Discuss Outliers:
- If there is a wide variance (e.g., someone said 2, someone else said 13), the team discusses why.
- Usually, the lowest and highest voters explain their reasoning.
Re-vote:
- After discussion, the team votes again.
- Repeat until consensus is reached (or near consensus, depending on team norms).
Record the Estimate:
- Once agreed, the story gets its official point value.

🧠 Tips for Planning Poker in SAFe:

Keep discussions focused; timebox each estimate if needed.
Use tools like Jira, Miro, or PlanningPoker.com for remote teams.
Encourage all voices—especially quiet team members.
Don't debate exact hours; focus on relative size and complexity.
Planning Poker works best for small- to medium-sized stories.

📌 When NOT to use Planning Poker:

For very large work items (like Epics or Features) → Use T-Shirt Sizing or Affinity Estimation instead.
When you're estimating hundreds of items quickly → Use Affinity Estimation.
During Portfolio level planning → Use WSJF and capacity planning techniques instead.

Multiple-choice questions (MCQs)

1. Which estimation technique is primarily used by Agile Teams in SAFe to estimate user stories?

A. T-Shirt Sizing
B. Planning Poker
C. WSJF
D. Monte Carlo Simulation

✅ Answer: B. Planning Poker

Explanation:
Agile teams in SAFe use Planning Poker to estimate user stories with story points. It's a relative estimation technique where team members discuss and converge on the size of the story based on complexity, effort, and uncertainty.

2. In SAFe, what is the primary use of the WSJF (Weighted Shortest Job First) method?

A. To estimate individual tasks
B. To prioritize the backlog
C. To plan daily standups
D. To determine team velocity

✅ Answer: B. To prioritize the backlog

Explanation:
WSJF is used to prioritize features, capabilities, or epics based on their cost of delay and job size. It helps teams and stakeholders decide what to work on next for maximum economic benefit.

3. In Affinity Estimation, what is the main goal of the process?

A. To define exact time durations for each user story
B. To compare stories to a baseline and estimate relatively
C. To group tasks by team members
D. To assign deadlines to user stories

✅ Answer: B. To compare stories to a baseline and estimate relatively

Explanation:
Affinity Estimation is about relative sizing. Stories are placed in relation to a baseline story and then clustered by similarity, making it fast and collaborative.

4. T-Shirt sizing is most often used in SAFe to estimate which type of work items?

A. User Stories
B. Enabler Stories
C. Capabilities or Epics
D. Iteration Tasks

✅ Answer: C. Capabilities or Epics

Explanation:
T-Shirt Sizing (XS, S, M, L, XL) is typically used for higher-level items like capabilities or epics in the portfolio or program level, especially in early stages when details are limited.

5. What is typically not a factor in calculating WSJF in SAFe?

A. Time Criticality
B. Risk Reduction or Opportunity Enablement
C. Cost of Delay
D. Estimated Hours

✅ Answer: D. Estimated Hours

Explanation:
WSJF uses relative estimation and does not rely on exact hours. It uses Cost of Delay (Business Value, Time Criticality, Risk Reduction) and Job Size, which is often represented by story points or T-shirt size.

6. During PI Planning, teams typically estimate features by breaking them down into what?

A. Stories
B. Tasks
C. Milestones
D. Capabilities

✅ Answer: A. Stories

Explanation:
During PI Planning, teams break down features into user stories so they can estimate them using story points and plan iterations accordingly.

7. Which of the following best describes Story Points in SAFe?

A. A measure of time
B. A relative measure of effort, complexity, and risk
C. A unit of cost estimation
D. A measure of business value

✅ Answer: B. A relative measure of effort, complexity, and risk

Explanation:
Story points represent relative effort required to complete a story, factoring in complexity, uncertainty, and risk—not actual time or hours.

8. Which technique is most appropriate for estimating a large number of backlog items quickly?

A. Planning Poker
B. WSJF
C. Affinity Estimation
D. Critical Path Method

✅ Answer: C. Affinity Estimation

Explanation:
Affinity Estimation is ideal for estimating many items quickly by sorting them relative to one another, especially useful during backlog refinement sessions.

9. What is the typical scale used in Planning Poker for story point estimation?

A. 1 to 10 linear scale
B. Fibonacci sequence (1, 2, 3, 5, 8...)
C. Decimal points (0.5, 1.0, 1.5...)
D. Percentage-based (10%, 20%, ...)

✅ Answer: B. Fibonacci sequence (1, 2, 3, 5, 8...)

Explanation:
Planning Poker often uses the Fibonacci sequence because the increasing gaps reflect greater uncertainty with larger items.

10. In SAFe, which level typically uses T-Shirt sizing to estimate work?

A. Team level
B. Iteration level
C. Program or Portfolio level
D. DevOps team

✅ Answer: C. Program or Portfolio level

Explanation:
T-Shirt sizing is a high-level estimation technique used for epics and capabilities, typically at the Program or Portfolio level when detailed story information isn't yet available.

11. Which of the following is true about WSJF in SAFe?

A. WSJF helps determine team velocity
B. WSJF is used only at the team level
C. WSJF ensures the most valuable work is done first
D. WSJF estimates the technical complexity of a story

✅ Answer: C. WSJF ensures the most valuable work is done first

Explanation:
By evaluating Cost of Delay against Job Size, WSJF helps prioritize features that offer the highest value in the shortest time.

12. Which SAFe role typically facilitates estimation and story splitting during PI Planning?

A. Product Owner
B. Release Train Engineer
C. Scrum Master
D. Solution Architect

✅ Answer: A. Product Owner

Explanation:
Product Owners work closely with the team during PI Planning to clarify features, split stories, and assist with estimation.

🔢 MCQ #13: Team Capacity in Hours

A team of 7 members is planning for a 2-week sprint. Each team member is available for 6 hours per day on average.
What is the team's total capacity in hours for the sprint (10 working days)?

A. 360 hours
B. 420 hours
C. 480 hours
D. 600 hours

<details><summary>✅ Answer</summary> 7 × 6 × 10 = **420 hours** ✅ Answer: **B. 420 hours** </details>

📊 MCQ #14: PI Velocity Forecast

An Agile team completes an average of 24 story points per sprint. If there are 5 iterations in the Program Increment,
what's the estimated total velocity for the PI?

A. 100 points
B. 110 points
C. 120 points
D. 125 points

✅ Answer : 24 × 4 = **96 story points**, Answer: **A. 100 points*

🔁 MCQ #15: Story Load and Capacity

A team has a velocity of 30 and selects stories estimated at: 5, 8, 3, 8, and 8.
Can this team take all of them in one sprint?

A. Yes
B. No
C. Only with overtime
D. Cannot determine

<details><summary>✅ Answer</summary> 5 + 8 + 3 + 8 + 8 = 32 Velocity = 30 → Slightly over ✅ Answer: **B. No** </details>

🧮 MCQ #16: WSJF Feature Prioritization

Given:

Feature A: BV=60, TC=40, RR=20, Job Size=10
Feature B: BV=40, TC=30, RR=10, Job Size=5

Which has higher WSJF?

A. Feature A
B. Feature B
C. Same
D. Cannot say

<details><summary>✅ Answer</summary> WSJF A = (60+40+20)/10 = 120/10 = 12 WSJF B = (40+30+10)/5 = 80/5 = 16 ✅ Answer: **B. Feature B** </details>

🛠️ MCQ #17: Planning Adjustment

A team overcommitted by 10 story points in a previous PI. Their average velocity is 28.
To stay on target, how many points should they plan for in the next sprint?

A. 18
B. 28
C. 30
D. Less than 28

<details><summary>✅ Answer</summary> If they consistently overcommit, they should **plan for less** than their velocity. ✅ Answer: **D. Less than 28** </details>

🧠 MCQ #18: Program PI Load

If an ART has 8 teams, and each has a velocity of 25, what is the total program capacity for a 5-sprint PI?

A. 800
B. 1000
C. 1200
D. 1500

<details><summary>✅ Answer</summary> 8 × 25 × 4= 800 ✅ Answer: **A. 800** </details>

🔢 MCQ #19: Sprint Carryover

A team planned for 30 story points in a sprint but only completed 22.
What is the percentage of work carried over?

A. 26.6%
B. 33.3%
C. 20%
D. 8%

<details><summary>✅ Answer</summary> Planned: 30, Done: 22 → Carryover = 8 (8 ÷ 30) × 100 = 26.6% ✅ Answer: **A. 26.6%** </details>

📈 MCQ #20: Cumulative Velocity

A team delivered: Sprint 1: 18, Sprint 2: 22, Sprint 3: 20.
What is their average velocity?

A. 18
B. 20
C. 21
D. 22

<details><summary>✅ Answer</summary> (18 + 22 + 20) ÷ 3 = 60 ÷ 3 = 20 ✅ Answer: **B. 20** </details>

🧮 MCQ #21: WSJF Calculation

Epic C:

Business Value = 90
Time Criticality = 40
Risk Reduction = 30
Job Size = 20

What is WSJF?

A. 8
B. 7
C. 6
D. 9

<details><summary>✅ Answer</summary> WSJF = (90 + 40 + 30) / 20 = 160 / 20 = 8 ✅ Answer: **A. 8** </details>

⚖️ MCQ #22: Estimating with T-Shirt Sizing

An Epic is sized as "XL". Your team's reference shows XL = 150–250 story points.
With an ART velocity of 100 points per PI, what's the minimum number of PIs needed?

A. 1
B. 2
C. 2.5
D. 3

<details><summary>✅ Answer</summary> 150–250 points needed → 250 ÷ 100 = 2.5 ✅ Answer: **C. 2.5** </details>

🛠️ MCQ #23: Team Load Check

Your team's historical velocity is 26. In current planning, they've selected stories totaling 32 points.
What should the Scrum Master suggest?

A. Accept the risk
B. Plan for all 32
C. Remove or split stories to reduce scope
D. Increase velocity artificially

<details><summary>✅ Answer</summary> ✅ Answer: **C. Remove or split stories to reduce scope** </details>

🎯 MCQ #24: Feature Progress Tracking

If a feature is estimated at 40 points and 24 points have been completed after 3 sprints,
what % of the feature is done?

A. 50%
B. 60%
C. 75%
D. 40%

<details><summary>✅ Answer</summary> (24 / 40) × 100 = 60% ✅ Answer: **B. 60%** </details>

📊 MCQ #25: Program Predictability Score

An ART has 5 teams. Their commitments and actual delivery in a PI are:

Team 1: 30 / 30
Team 2: 25 / 20
Team 3: 28 / 28
Team 4: 26 / 24
Team 5: 20 / 18

What is the ART's Program Predictability Score?

A. 85%
B. 92%
C. 95%
D. 100%

<details><summary>✅ Answer</summary> Total committed = 30+25+28+26+20 = 129 Total delivered = 30+20+28+24+18 = 120 (120 ÷ 129) × 100 ≈ **93.02%** ✅ Closest Answer: **B. 92%** </details>

⚙️ MCQ #26: WSJF Tie Breaker

Two Features have the same WSJF score of 10. Feature X has higher Time Criticality.
Which should be prioritized?

A. Feature X
B. The one with smaller Job Size
C. Arbitrary choice
D. The one with higher Business Value

<details><summary>✅ Answer</summary> When WSJF is tied, **Time Criticality** often acts as the next priority signal. ✅ Answer: **A. Feature X** </details>

🧮 MCQ #27: Team PI Load

A team has a velocity of 24 per sprint. A PI has 5 iterations.
If they allocate 20% of their capacity to Enablers, how many points remain for Business Features?

A. 96
B. 100
C. 120
D. 80

<details><summary>✅ Answer</summary> Total PI capacity = 24 × 5 = 120 Enablers = 20% of 120 = 24 Business = 120 - 24 = **96** ✅ Answer: **A. 96** </details>

🛠️ MCQ #28: Uncommitted Load

A team commits to 26 story points and adds 4 points of stretch goals.
What % of their plan is uncommitted?

A. 13.3%
B. 15.3%
C. 12%
D. 10%

<details><summary>✅ Answer</summary> Uncommitted % = (4 / (26 + 4)) × 100 = 4/30 × 100 = **13.33%** ✅ Answer: **A. 13.3%** </details>

🚦 MCQ #29: Team Load Forecasting

A team's capacity is 480 hours for a sprint. They estimate their current backlog at 520 hours.
What is their over-commitment percentage?

A. 8%
B. 10%
C. 5%
D. 12%

<details><summary>✅ Answer</summary> Overload = (520 - 480) = 40 (40 / 480) × 100 = **8.33%** ✅ Closest Answer: **A. 8%** </details>

🔄 MCQ #30: Shared Services Capacity

A shared services team supports 4 ARTs. Each ART requires 10% of their time.
What % of their total capacity is allocated to shared support?

A. 40%
B. 25%
C. 30%
D. 50%

<details><summary>✅ Answer</summary> Each ART = 10% → 4 ARTs = 10% × 4 = **40%** ✅ Answer: **A. 40%** </details>