//

10 Ideas to Make Backlog Refinement Rock

10 mins read
Backlog Refinement

Before we go through the ten ideas, it’s important to take note of two really important things — backlog refinement is a team sport and backlog refinement is continuous — not just one session every couple weeks. So although most of the ideas are focused on the backlog refinement session, the intent is to introduce ways that we can increase participation, make it easier to facilitate, and hopefully, greatly improve shared understanding.

  1. Be Prepared. Probably the most critical part of making any refinement session successful is being prepared. What does this mean, well maybe a checklist is in order:
    • Identify the top-x stories that are not “ready” and need refinement. Ideally, your backlog is ranked, and it’s the next items on the list.
    • Identify the questions for the team and invited subject matter experts.
    • Prepare any artifacts that will help the conversation around the stories (e.g., designs, user feedback, 3rd party specifications, etc.).
    • Send out the agenda with the list of stories to be covered and identify the preparation steps for the attendees.
    • Make sure that if you have remote people that you have identified the technologies that you will use (e.g., web meeting software, full-duplex phone, online brainstorming tool, etc.) and you’ve tested these out.
    • If there’s someone new or the meeting will be largely remote, plan to either start early or use the first few minutes to get everyone connected — set this expectation ahead of time.
    • Identify and confirm the facilitator and the approach for capturing notes and observations. Let’s not have all the great ideas and things we learn to escape into the ether.
  2. Design Storm. A Design Storm is a type of workshop that is usually associated with human-centered design and or design thinking where participants are asked to create a low-fidelity concept around an idea or problem a group is trying to solve. For a Backlog Refinement session, you can use this approach to get people who might not normally speak up, to get engaged in the process. Generally, participants will uncover great ideas, and discussions will ensue around complexity or “de-complexifying” the backlog item(s).
     
  3. Lo-fi Prototype. Create a low fidelity prototype and pass it around the room or have multiples created. Something as simple as sketches of the screen(s) and when someone “clicks” a button or enters a field, you can show another sketch or discuss content. You don’t have to be an artist or use some amazing (a.k.a. complex) UX mocking tool. Just use post-its and or normal letter (or A4) paper. When you use this approach, encourage folks to be critical and question everything, even if it’s an area where you tried to mark out a mistake. If you are wondering just what a low-fidelity prototype is, check out this article by Laura Busche.
     
  4. Oblique Strategies. Oblique strategies are a set of statements that are meant to create a challenging constraint to help increase creativity through lateral thinking. Okay, this one may seem a little odd for a backlog refinement session, but I’ve been witness to many sessions where the ideas to solve the user problems are mundane and more of the same — often being offered by the same 1-2 people in the room, over, and over, and over again. Oblique strategies can help the group try something new. Do a quick search on the app stores, and you’ll find an app that has oblique strategies listed. NOTE – some oblique strategies are to help musicians and might not make sense — go to the next one.
     
  5. G-W-T. Given-When-Then (a.k.a., Gherkin) is the product of Behavior-Driven development. GWT is an approach where as a team, we can use a structured language to talk about how a user [or system] may interact with a tool or piece of software to achieve their desired outcome. If you’ve never used GWT, you’ll find the structure to help drive some conversations and give clarity to the behavior the system without dictating the solution.
     
  6. SQUID. This acronym stands for Sequential Question and Insight Diagram. SQUID is a great approach to allow the group to ask questions and record the answers. It can be done interactively with the facilitator(s) capturing items, or ideally, there are a series of brainstorm moments where the team writes questions, and the knowing team member poses answers. The process repeats until the team has clarity for the backlog item. See the GameStorming.com website for how to use this approach.
     
  7. Apply Lean Coffee. If you’ve never heard of Lean Coffee, it’s an approach to facilitating an open session of people. Generally based on a context, Lean Coffee provides a means by which to structure discussions through the use of a personal kanban board and time boxing (see leancoffee.org for more information). For a backlog refinement session, instead of brainstorming and dot-voting topics — use the backlog items to be discussed. Applying the Lean Coffee structure does two things — (1) focuses the discussion on one item at a time and (2) it helps keep the conversation moving instead of getting stuck on one item for too long. When I’ve used this question in the past, the question around extending the time box was “Can we estimate this one yet? If not, do we keep talking? Or, does it need to go back with the PO for more discovery?”
     
  8. Story-walk. One thing that brain-science has made clear, our noodles are more engaged and work better when we move, and there’s some variety. A story walk is simple, print out the story details as it’s known today — make it big, so it’s easily readable. Give each person a pad of sticky notes and a marker. Ask them to walk around writing and posting Questions (?), Assumptions (!), Actions (*) and Ideas ($). Make sure they use symbols that make it clear what type of post they are submitting. Do this for 20 minutes, and then ask the team to re-walk the stories — this time clarifying and grouping similar items. After ten more minutes, focus on the questions, and get them answered. Hopefully, by the end of the session, you can do some quick affinity estimation or white-elephant sizing. A similar exercise as the story walk is doing a story workstation, where you set-up multiple tables with handouts containing details about each story. The ask is the same, but instead, the participants are moving table-to-table.
     
  9. Triad-up (a.k.a., Three Amigos). Have three people take a backlog item and gain an understanding of it. The three people should be made up of one with business knowledge, another focused on the question of “how to test?”, and another person focused on the question of “how to build?” These three amigos should do their best to get the backlog item to a point where they believe the team should be able to estimate it together. Get the room into triads, divvy up the backlog items, and give them a time box to discuss the backlog items. Leave 15-20 minutes for shared discussion and estimation at the end. To learn more about the Three Amigos (in context of agile software development) check out this article by George Dinwiddie.
     
  10. Re-baseline Understanding of Refinement. One thing we often do as humans is that we often extend our level of understanding what we believe others know about a topic. Well, if a team hasn’t been doing something for very long or that something they are doing hasn’t been very effective, then it’s probably time to practice something new. All that said, take the time to discuss the concept of refinement, the team’s role in refinement, how a refinement session should work, and the desired outcomes.
     

Matt Badgley

With a career in Information Technology that has officially reached drinking age, I have one too many hats in roles from Systems Analyst to Programmer to IT Manager to Programmer to Director. I am a seeker of new ideas and learning new things. My passion has been working with product delivery teams to develop into their own identity and simply build valuable solutions by making great software. I have been applying Lean Agile ideas for a long time and leveraging their underlying practices with both a pragmatic and experimental purpose. At the end of the day, I believe in integrity, hard work, trial-and-error, people, and faith.

Leave a Reply

Your email address will not be published.

CommentLuv badge

Happy Testivus
Previous Story

Testivus for the Rest of Us: A Whole Team Testing Experience

Next Story

Using a Hero's Journey to Create an Inspiring Vision

Latest from Agile