What Is Planning Poker?
Planning Poker is a consensus-based estimation technique where team members independently estimate the effort of a user story using numbered cards (typically Fibonacci: 1, 2, 3, 5, 8, 13, 21). After everyone reveals their card simultaneously, the team discusses any large discrepancies and re-estimates until converging.
It works because:
- Independent estimation prevents anchoring bias (the loudest person's estimate doesn't dominate)
- Discussion of outliers surfaces hidden complexity that one person spotted and others missed
- Convergence builds team alignment on what "a 5-point story" actually means
How to Run a Remote Planning Poker Session
Before the Session
- Prepare the backlog — have 5-10 stories ready with acceptance criteria. Don't estimate more than 10 stories per session.
- Set the scale — Fibonacci (1, 2, 3, 5, 8, 13, 21) is most popular. T-shirt sizes (XS, S, M, L, XL) work for less technical teams.
- Invite the right people — everyone who will work on the stories should estimate. Managers and PMs participate in discussion but typically don't vote.
During the Session (30-45 Minutes)
Step 1: The facilitator reads the story and acceptance criteria (2 minutes).
Step 2: Everyone asks clarifying questions (2-3 minutes). No estimation discussion yet.
Step 3: Everyone selects their estimate simultaneously. Online tools handle the "card flip" — all estimates revealed at once.
Step 4: If estimates are within 2 adjacent Fibonacci numbers (e.g., 5 and 8), take the higher number and move on.
Step 5: If estimates are more than 2 apart (e.g., 3 and 13), the highest and lowest estimators explain their reasoning. Then re-estimate.
Step 6: After 2 rounds, if no consensus, the facilitator decides. Don't spend more than 5 minutes on any single story.
Common Mistakes
-
Estimating time instead of effort — story points measure complexity and uncertainty, not hours. A story that takes 2 hours but has high uncertainty is a bigger point value than a 6-hour story with zero unknowns.
-
The "1-point everything" trap — if every story is 1 or 2 points, your scale isn't providing useful information. Recalibrate by finding a reference story that the team agrees is a "5" and estimate everything relative to it.
-
Anchoring on the first estimate — this is why cards must be revealed simultaneously. If the senior developer says "3" first, everyone else adjusts toward 3.
-
Skipping the discussion — the value of Planning Poker is the conversation, not the number. When a frontend developer estimates 3 and a backend developer estimates 13, the discussion reveals integration complexity nobody else considered.
Connecting Estimates to Business Value
Traditional Planning Poker stops at story points. But points without context are just numbers. When you know your team's velocity (points per sprint) and your engineering cost per sprint, every point has a dollar value.
Example: If your team delivers 30 points per sprint and your monthly engineering cost is $60K (2-week sprints = $30K/sprint), each point costs $1,000. A 13-point feature costs $13,000 of runway.
This changes the prioritization conversation from "is this a big story?" to "is this feature worth $13,000 of runway?"
How BurnRateOS Helps
BurnRateOS Planning Poker is free, real-time, and cost-aware:
- Multiple scale options (Fibonacci, T-shirt, custom)
- Simultaneous card reveal with outlier highlighting
- Velocity tracking across sessions
- Cost-per-point modeling connected to your runway data