Planning poker is an effective estimation technique for software development teams to estimate how long project tasks will take. I’ve personally used this technique throughout my 15+ years of software development. It’s a fun, collaborative activity that utilizes the collective knowledge of your team to generate accurate estimates. It’s also especially helpful if your goal is to improve your project planning and estimation accuracy.
What is Planning Poker?
Planning Poker is an Agile project management estimation technique. I’ve personally used many times to help development teams estimate the effort required for various tasks or user stories. It’s an excellent combination of expert opinion, group consensus, and game-like elements to ensure accurate time estimates.
Planning Poker was introduced by James Grenning in 2002 and popularized by Mike Cohn in his 2005 book “Agile Estimating and Planning.” Here’s a brief history
- 2002: James Grenning introduces Planning Poker
- 2005: Mike Cohn popularizes it in his book
- 2009: Mountain Goat Software trademarks Planning Poker
A typical session requires the following:
- A deck of cards with values
- A product backlog of user stories
- Team members who will estimate
- A moderator to run the session
Planning Poker is a core part of Agile and Scrum. It allows teams to break down a large project into small tasks. It’s also an effective way to encourage collaboration and utilize the wisdom of the crowd to make more accurate estimates.
Estimation Card Values and Their Significance
Planning Poker uses a modified Fibonacci sequence for card values because the sequence captures the inherent uncertainty in estimating larger items. I’ve found this to be helpful as it allows us to more accurately capture the nonlinear nature of software development effort.
Common card values are:
0, 1, 2, 3, 5, 8, 13, 20, 40, 100
Each value represents:
- 0: Task is done or requires no effort
- 1: Very small, well defined task
- 2-3: Small tasks with low uncertainty
- 5-8: Medium complexity task
- 13-20: Complex task with some uncertainty
- 40-100: Very complex task that probably needs to be broken down
Some teams use different card decks. T-shirt sizes (XS, S, M, L, XL) are a more abstract scale. The ? card represents a task that still needs more details before estimating.
These values give you a framework for discussion. They force team members to analyze the effort of each task critically.
How to Conduct a Planning Poker Session
Conducting an effective Planning Poker session is key to accurate estimates. Here’s a step-by-step guide that I have found to work well:
- Gather the team and materials.
- Present a user story or task from the product backlog.
- Allow the team to ask questions for clarity on the item.
- Each team member selects a card that represents their estimate.
- All team members turn their cards over at the same time.
Discuss why team members picked their cards, particularly if there are any outliers.
Re-estimate the item until team members agree on an estimate.
Record the estimate and move onto the next item.
The role of the moderator is to facilitate the discussion and ensure everyone has a say. Participants make estimates based on their own judgment and experience.
Session Performance Data:
- Average session length is 2-3 hours.
- Teams can estimate 20-30 backlog items per session.
- It takes an average of 2-3 rounds to estimate an item.
For remote teams, modify Planning Poker by using a video conference and digital estimating tools so that people can participate from anywhere.Advantages of Story Point Estimation
There are several reasons why Planning Poker is superior to other common estimation techniques. I’ve personally experienced these advantages in many projects:
More accurate estimates: It harnesses collective intelligence, which produces more accurate estimates. A study by K. MoløkkenØstvold and N.C. Haugen discovered that Planning Poker estimates were more accurate than estimates from individual experts.
Reduction of anchoring bias: By revealing estimates simultaneously, you eliminate the influence of the first estimate on all subsequent estimates.
Team collaboration and knowledge transfer: During the discussion, team members will share insights and concerns with each other. This ensures every team member has a more comprehensive understanding of the project.
Early risk identification:
- Hidden complexities within tasks
- Knowledge silos among the team
- Task dependencies
- Potential bottlenecks
The study from IEEE also found groups using it delivered estimates that were about 30% higher than estimates from unstructured group estimation. This indicates Planning Poker helps capture many of the often forgotten complexities.
Common Pitfalls in Planning Poker and How to Avoid Them
Planning Poker is excellent, but there are some drawbacks to using it incorrectly. Here are the main pitfalls I’ve seen:
Over indexing on accuracy: Estimates are estimates, not exact figures.
Letting a few strong voices dictate the entire session: Every team member should have an equal opportunity to submit their estimate.
Not discussing the rationale behind each estimate: This step is crucial to avoid misinterpretation later.
Forgetting to adjust the estimate as more information becomes available: You should periodically revisit and update your estimates based on new information.
Optimal Strategies for Agile Estimation Sessions
To get the most out of it, here are the best practices:
Prepare the product backlog in advance: Make sure your stories are properly prepared and ready to estimate.
Define a clear “done” for each item: This will help ensure everyone is on the same page and produce more accurate estimates.
Promote open discussion and questions: Build an environment where team members aren’t afraid to speak their mind.
Review and optimize the process: Always look for ways to improve your Planning Poker as a team based on the feedback and results.
Tools and Resources
There are a few other tools you can use to make your experience more enjoyable:
Physical Planning Poker card decks: Physical card decks are useful for in-person sessions and add a physical element to the process.
Digital tools for remote sessions:
- PlanITpoker
- Scrumpoker-online
- Pointing Poker
Integration with project management software: Many Agile tools, such as Jira or VersionOne, have a built-in Planning Poker feature.
Other resources to learn and implement Planning Poker: Agile estimation books, workshops, and online courses.
Comparison of Top Estimation Card Game Tools
Here’s a quick summary of the best digital Planning Poker tools:
- PlanITpoker:
- Real–time collaboration
- Customizable decks
- Jira integration
- Free with paid plans for additional features
- Scrumpoker-online:
- Simple interface
- No registration required
- Supports various estimation methods
- Free
- Pointing Poker:
- Create custom rooms
- Timer
- Mobile friendly
- Free with premium features
Each has its own set of features, so select the best for your team based on their particular needs and preferences.
Planning Poker is an essential part of sprint planning, helping teams estimate their work effectively. It’s also closely related to the concept of story points, which are commonly used in Agile estimation.
To Sum It Up
Planning Poker is an excellent tool for agile teams because it combines collaborative estimation with bias reduction to make estimates more accurate. The use of the modified Fibonacci sequence and a structured process result in more realistic project timelines.
The technique also improves team communication and helps teams identify risks early. Remote teams can use it by utilizing online tools. The more you practice and discuss retrospectives, the better you’ll become at using Planning Poker. However, keep in mind that the goal is relative sizing of tasks, not nailing down perfect time estimates.
