Inspiration for building an app for the Atlassian Marketplace
The inspiration for building a Cloud solution for the Atlassian marketplace came back in December 2020, when I came across an article about Atlassian new cloud platform Forge.
Once I learned about Forge, its Function-as-a-service backend, its CustomUI front-end framework, and simple command-line interface (CLI) - hosted by Atlassian - I knew this was the opportunity I had been looking for. With Forge the barrier to entry had been lowered, allowing small teams to bring new products to life on the Atlassian marketplace.
Why on earth do we need yet another Planning Poker app?
Interesting question because lot of solutions already exists. The best solutions are solutions that truly understand the job people are trying to get done. I felt that current tools on the market didn't properly understand the job of leading a planning/refinement session. I say this because working with existing Planning Poker apps I wanted a solution that:
- Didn’t ask users to create and manage a separate list of stories for every meeting. Why manage a second list of stories when I already have the backlog?
- Didn't ask users to name the planning session and invite team members because other tools are perfect for this, i.e. Slack, calendar, etc.
- Is available when I open a story. That is, doesn’t take me away from the actions I need to update the story and record key decisions
- Easy to use and supports one-click estimates without opening a new window
- Works for remote as well as co-located teams.
- Allows participants to see who has given an estimate and who has not. Just as when the team is in the same room
Furthermore when I noticed a reoccurring theme in online forums discussing current Planning Poker apps on the market:
- “My teams are struggling with estimation tools, we have tried everything with little to no success”
- “We abandoned this idea, there were no tools that actually worked out”
- “I haven’t found a tool that doesn’t slow the process down”
I understood there must be a way to simplify and improve on these problems.
What Smart Guess for Planning Poker does
Smart Guess for Planning Poker removes 20+ steps from your story point estimation - compared to other tools on the market. Saving time and making it easier than ever for teams, remote/co-located, to run planning poker estimation during planning and refinement meetings.
Smart Guess doesn't need any effort to setup before planning. While other tools need Scrum masters to create a ‘planning game’, name it, create an ‘estimation backlog’, and so on. With Smart Guess you skip all that, allowing you to focus your time where it’s needed.
With Smart Guess you run planning and estimation solely from the Jira backlog. Allowing teams to keep the Jira issue as the single source of truth about the story being discussed. By running your meetings using the Jira backlog you can easily update, split, reprioritize stories as the discussion takes place. With current planning poker tools, this becomes so more complicated - as users notice.
During story discussion when your team has enough clarity on the story and is ready to estimate, with Smart Guess giving an estimate is always one click away. To record decisions during story discussion is easy because all the actions are readily available from the Jira backlog.
How Smart Guess was built
In the product discovery phase, just to set the baseline, work started by mapping out the job steps of a regular Sprint planning meeting for a team using physical planning poker cards. Then map out the job steps of current planning poker tools on the market. Comparing the two workflows showed a stark difference with how current tools work and hinted a simpler solution might be possible.
A decision was made to build out planning poker estimation within the Jira issue panel, allowing users to give estimates directly within the Jira issue and the Jira backlog. Given the fact Jira issues already contain a lot of information, a decision was made to limit the space it takes on the screen. After all story estimation is only part of what happens during Sprint planning and refinement meetings - not the main focus.
Using a whiteboard a rough outline of how the app might work was drawn out in the beginning and iterated during implementation in discussion with potential users. It made sense to build on top of Atlaskit UI components as Jira users are already familiar with them. Therefore Atlaskit components are used wherever possible.
On the development side, running through the examples on how to get started with Forge gave the confidence everything was in place to start development. Having the Atlassian developer community was invaluable to find answers and get feedback on issues that came up. Reporting issues and getting direct feedback from Atlassian developers to resolve problems further increased the confidence with the Forge platform.
Getting feedback during development was essential. Understanding what potential users found confusing helped hone in on improved texts (app copy) and the real-time interactions allowing team members to understand which team member selected an estimate, who just revealed all the estimates, and so on.
Challenges
The main challenge during implementation was how to support real-time interactions between users taking part. That is, once a team member joined, selected an estimate and so on you want everyone else on the team to see who joined, who selected an estimate, and so on. Just as when everyone is sitting around the same table. Asking about this, the Forge platform didn’t support connections to external servers, enabling this kind of functionality. Soon Forge evolved, allowing Smart Guess to build out real-time interactions critical for the solution to work.
What we learned
The Atlassian Forge platform is a true game-changer for small teams bringing new products to market. Truly lowering the barrier to entry by hosting the solution on top of Atlassian (AWS) infrastructure and making it easier to bring strong solutions to market. Atlassian does this by providing great resources:
- User community - what are users struggling with?
- Developer community - helping developers finding answers to questions
- Design system - Making it easier to give Atlassian users consistent user experience
- Marketing guidelines
Need an idea for your product? First, understand what users are struggling with. Check out online user forums, Quora, Reddit, etc. The possibilities are endless, so many things can be improved 😉🚀
What's next for Smart Guess for Planning Poker
As Smart Guess has just recently been made available there are a lot of improvements yet to be made. As an example current version:
- Doesn't work perfectly for mobile users
- Only supports story point estimation
- Is only available in English
In addition, there are a lot of ideas on how to make the solution stronger. To understand what problems are most important feedback from actual users is essential to clarify the next steps. Be in contact with any questions or feedback you might have.
