Photo Credit: Pixabay- WOKANDAPIX
“Plans are worthless, but planning is everything.”Dwight D. Eisenhower
Has any of this happened to your Scrum Team?
- The team regularly takes on more work than they can complete in a Sprint.
- Work is removed from the Sprint because it could not be completed.
- The quality of the final work is low, with many issues found after the Sprint.
- Stories grow in size during the Sprint as more work is discovered.
- Arguments erupt on the right approach to getting work done.
- Sprints looked successful, only to have the stakeholders say it doesn’t meet business needs.
These challenges and more can be traced back to poor Sprint Planning.
The Basics of Sprint Planning
Sprint Planning marks the start of each Sprint. It sets the overall tone for Sprint. Good Sprint Planning leads to a good Sprint. A Sprint that doesn’t generate value is creating waste.
Let’s start with a quick primer on the five Ws of Sprint Planning:
|What? (Definition)||A planning event to create the Sprint Backlog, which answers the questions: “Why are we doing the Sprint?”, “What work will be done?” and “How will the work be done?”.|
|Why? (Purpose)||To create alignment on the purpose and work needed to deliver on the Sprint Goal.|
|When?||Sprint Planning is the first event inside the container event of the Sprint. The Sprint starts when Sprint Planning begins.|
|Who?||The Product OwnerThe DevelopersThe Scrum MasterStakeholders and SMEs as needed|
|How long?||Two hours per week of Sprint|
Now we’ll unpack some of these to understand better and learn how not doing them could impact your Sprint…
Unpacking Sprint Planning
Why do we do Sprint Planning?
We do Sprint Planning to create alignment and purpose around the Why, What, and How of Sprint Planning.
- Why is it valuable? (Creating the Sprint Goal.)
- What can be done in the Sprint? (Populating the Sprint Backlog.)
- How will the chosen work get done? (Breaking down the work)
If we don’t answer all three questions, we don’t have true transparency of the work and how it connects to the whole.
The most common places I see Sprint Planning, and by extension the Sprint, fail is with the Why and the How parts of Sprint Planning.
There is no Why: No Sprint Goal.
The Scrum Guide defines the Sprint Goal as “the single objective for the Sprint.” To this definition, I add it is “a concrete stepping stone towards the Product Goal.”
If your Sprint Goal does not connect to your Product Goal, then you are not “maximizing the value of the product resulting from the work of the Scrum Team.” Without value, you can’t be profitable. Without profitability, you can’t be a sustainable business.
Defining a Sprint Goal at the beginning of Sprint Planning is critical. Defining the goal upfront allows all the work pulled into the Sprint to support the Sprint Goal, enabling the output and outcome of the Sprint to contribute towards the Product Goal.
There is no How: The work is not understood.
One of the reasons for Sprint Planning is to create alignment between the Product Owner and Developers. Alignment is not just between the Product Owner and the Developers. It is also between the Developers. Does the database expert know how the business logic person will implement their work? Is everyone clear on how the end product will be tested? How will we demonstrate this? If the team is not asking these questions in Sprint Planning, you are headed for rocky waters.
Many Agile teams treat the Events like secret Star Chamber meetings requiring a password to enter. Scrum works on the fundamental pillar of Transparency. Without transparency, everything else in Scrum will suffer and eventually fail.
With Sprint Planning, transparency is essential as it is the last chance to refine and understand the team’s work in the Sprint. Inviting the business stakeholder to provide greater context, or the Architect, to work through some design questions can be the difference between a Sprint Review where everyone says, “That was awesome!” and one where you hear, “That’s not what I wanted” or “You realize that’s not security compliant and you need to redo it?”
The Scrum Guide says:
“Sprint Planning is timeboxed to a maximum of eight hours for a one-month Sprint. For shorter Sprints, the event is usually shorter.”
Eight hours? That’s an entire day! Are you serious?
Let’s do a little math here. A one-month Sprint is, on average, 20 days or 160 hours. Eight hours is just five percent of the total time of the Sprint. Five percent!
Now let’s look at a typical waterfall project. Twenty years ago, the projects I managed were typically 12 months long. The first three months of that project were back and forth between product management and development on precisely what would be built with no code being written (for this release) for at least two months. Plus, the product managers had previously written requirements documents for at least two months.
Twelve months, with two months spent in “planning.” That’s 16 percent of the release spent on planning. And let’s be honest, that’s probably on the low side by a generous margin.
So, yes, two hours per week of Sprint is not only reasonable, it’s responsible. We’ve already looked at the two common failure patterns above. Lack of sufficient time is not only our third failure pattern, it also causes the first two.
And let’s remember; the Scrum Guide says “a maximum of.” If the team finishes before the timebox, no one will complain and demand you all stay until the timebox ends. Schedule the time so you have it. Nothing is worse than having a two-hour meeting and finding out you needed two hours and fifteen minutes.
In the next section, I’ll show you my agenda for a two-week Sprint.
Agenda for an Effective Sprint Planning Event
The following is a recommended agenda flow for a two-week Sprint and has a four-hour schedule.
|Set the Stage||Welcome attendees and set expectations||~5 min|
|Why||The Product Owner describes a possible Sprint Goal and then discusses it with the Scrum Team||~10-20 min|
|What||Final Refinement and Estimation occur on Product Backlog Items (aka stories) that the Product Owner desires for the Sprint||~15-45 Min|
|What||The team reviews their availability and past performance to determine their potential capacity for the Sprint||~10-15 Min|
|What||Populate the Sprint Backlog with the Product Backlog Items that will meet the Sprint Goal||~15-30 Min|
|How||The Developers plan how they will complete the items in the Sprint Backlog||~2 hrs|
|Close||Take a final confidence vote, thank the team, and end||~5 min|
1. Set the Stage: In their book Agile Retrospectives, Esther Derby and Diana Larsen say, “Setting the stage helps people focus on the work at hand. It reiterates the goal for the time the team has together in the retrospective.” The above is good advice for all the Scrum Events, and I start Sprint Planning by thanking everyone for their time and reminding them of the purpose of Sprint Planning.
2. Establish a Sprint Goal: We’ve already discussed how failure to create a Sprint Goal or even creating a Sprint Goal after deciding what to do is one of the top reasons Sprints fail. If Sprint Planning sets the tone for the Sprint, the Sprint Goal sets the tone for the planning.
The Product Owner should come prepared with a recommended Sprint Goal. This Sprint Goal can result from the Sprint Review the day before or based on other strategic work. To be effective, the Sprint Goal connects directly to the Product Goal. The team then discusses, refines, or changes the Sprint Goal.
3. Final Refinement and Estimation: You should already be doing regular Backlog Refinement to keep your Product Backlog topped off with regular work. That said, Sprint Planning is the “last responsible moment” to do refinement. Check out the Scrum Alliance article linked above for more on Refinement.
Generally, I see three kinds of refinement happening during Sprint Planning. From most common to least, they are:
- Reestimation of work
- Updated Acceptance Criteria
- New, usually urgent, work
Reestimation can happen at any point up to when the team starts working on the item. Say that the last refinement session was the Thursday before. Since then, Sunil has been rummaging around the data layer and has discovered several new connections to other processes. Sunil returns to the team and says, “I think we underestimated the complexity.” The estimate for the item could change right there in Sprint Planning.
Acceptance criteria are how the Product Owner can shape the value of the work. The PO may want the new login page to be 25% faster than the last one. Martha has been looking left, right, and sideways at the code, and she doesn’t think the team can squeeze more than 15% out of the current system. The PO then decides if getting the new login page now is more important than getting it to load faster. The PO wisely decides now is better and agrees to change the AC to a 10% improvement.
And finally, remember that the Product Backlog is dynamic and emergent. New items can shoot to the top of the backlog overnight. The Product Owner may bring a brand new item to the team and ask them, “Can we refine this and include it in this Sprint?” While this should be uncommon, it should also be possible.
4. Determine Sprint Capacity: How much work can the team do on average? Will anything impact the team’s ability to work this Sprint? A great example is what to do when a major holiday falls in the Sprint. Instead of the team shortening the Sprint, they consider how this will impact them. If the entire office shuts down for a week, then the team’s capacity is 50% of what their velocity says they can do.
5. Populate the Sprint Backlog: Once you have a capacity guide for the Sprint, the team can fill the Sprint Backlog with work that supports the Sprint Goal. There are many ways to do this. Just remember that the Developers pull work into the Sprint. The Product Owner decides if the work to do has value and contributes to the Sprint Goal; the Developers decide how much work they can do in a given Sprint.
6. The Developers plan how they will complete the items in the Sprint Backlog: This is the most frequently forgotten — and therefore problematic — part of Sprint Planning.
Here’s how I explain this to students:
“When Sprint Planning starts, you are coding! Not all coding is hands-on keyboard typing. There is whiteboarding, design conversations, coordination, and more.”
If the developers need to look at the code to determine how they will build something, then look at the code. Scrum Masters, this is one of the places you are most needed. Ask questions, and seek clarity. Does the team understand what they will do, or are they planning to “just figure it out as we go?” That’s not Agile; that’s just lazy.
The How phase is the most time-consuming part of Sprint Planning. Done well, it prevents rework and conflicts (people, process, and product). It allows for a greater chance the team will deliver a usable increment of value (AKA done work the customer cares about) at the end of the Sprint.
Here are some things Developers should be doing in Sprint Planning:
- Break work down “into smaller, more precise items.”
- AKA create individual tasks that represent the technical work needed to complete the Product Backlog Item. (UI work, DB work, code review, create automation tests, etc.)
- Whiteboarding the design
- Sketch the user interface
- Review existing code
- Anything required to plan how to complete the work
Breaking work into tasks is vitally important and should never be pushed out with a “we’ll figure it out later.” Tasking allows for:
- Remembering the work: Five days from now, will you remember what you discussed in the Sprint Planning meeting? And if you didn’t even talk about it at Sprint Planning, do you remember when you did talk about it?
- Alignment: How will the UI work fit into the Business Logic work?
- Track Progress: If you break work into tasks of no more than four hours, your burn-down charts will reflect progress toward the Sprint Goal more accurately.
- Create stronger teams: When the team breaks work into bite-size pieces, it often allows someone with less skill to pick it up and complete it. This benefits the team in creating more cross-functional members and benefits the individual who gets to improve their mastery.
The Product Owner during the How
What does the Product Owner do during the How part of Sprint Planning? Unfortunately, the typical pattern is “See you all later; let me know how it goes.”
“The Product Owner is accountable for maximizing the value of the product resulting from the work of the Scrum Team.”
– The Scrum Guide.
The work is still all about value. If the Product Owner isn’t present, important questions might go unanswered, acceptance criteria can’t be tweaked, and problems not detected until too late.
The PO should be present and available. Their presence will ensure a continued alignment to value.
7. Close: Closing can be an under-appreciated part of good Sprint Planning. In Agile Retrospectives, the authors say, “End the retrospective decisively: don’t let people (and their energy) dribble away.”
Remember the maximum timebox. If the team is still going strong at three hours and fifty minutes, the Scrum Master can ask the team what’s needed to wrap up. Maybe they didn’t get all the work planned out. What will they do about that? Do they need to de-commit some work before the Sprint starts?
One of the things to do here is to collect a confidence vote. Now that the Developers have planned the work, how confident are they that they can complete it all in the Sprint?
Using the Fist to Five consensus framework is an excellent way to check on confidence.
Working with multiple teams
What happens when a Product Owner is supporting multiple teams? This is not immediately a bad thing. The Scrum Guide even has a suggestion:
“If Scrum Teams become too large, they should consider reorganizing into multiple cohesive Scrum Teams, each focused on the same product. Therefore, they should share the same Product Goal, Product Backlog, and Product Owner.”
The critical point here is that to be effective, the Product Owner must only have one Product Goal at a time. The next question will be, “How many teams can one Product Owner support?”.
My answer is 2
There are other answers and arguments, and we’re stepping into scaling conversations now. My experience is that a Product Owner becomes exponentially less effective beyond two teams. Once you move to four teams, you need to insert a scaling practice to have someone focused on the longer term while Product Owners continue to be sleeves rolled up and engaged with no more than two teams.
In practice, the way I’ve coached Product Owners supporting two teams is as follows:
- Shared Backlog Refinement
- Not everyone has to be in refinement sessions. Ensure you have representation from both teams and that all skills needed to complete are represented.
- Shared Why and What Sprint Planning
- This allows alignment, horse trading of work, managing dependency, and more.
- Individual How
- Do these at the same time with the PO floating between both teams.
- Individual Daily Scrums
- Stagger them so the PO can go to both
- A representative from each team should go to the other Daily Scrum
- Individual Retrospectives
- The PO rotates attendance every Sprint
- Every quarter or so, do a combined retro on overall ways of working.
A good Sprint Planning Event creates Value and Quality
The concept of “Starting on the right foot” is not just a trite saying. If you want to improve the quality of your work, improve the quality of your Sprint Planning.