Pen & paper
I started, and now we do sketch the UI but only two scenes; the main screen with some notes on it about behavior and possible paths to other screens and a screen of choice. Showing them in a video session or onsite face-2-face to gather some meta feedback; meta feedback such as beautiful UI or huge feature list? Feature-set completeness or due dates? Time or money?
Together we write down a wishlist, a list of thoughts, expectations, and wishes. Getting answers the questions like "How will this software looks like in the year 3000?". It's an overthought massively overdrawn picture of something that will never come to live. But we try to remove all borders and limits to allow each developing ideas for the future ready to be used tomorrow. Our customers quickly jump on the bandwagon and sketch some ideas about how many users, how to access data, what shall be automated, etc. , etc.
Tools We put everything into Confluence and Jira by converting handwritings into digital paperwork. Yes, at the end we do software development the Atlassian way. But it turned out it is the simplest way to connect with all peoples involved into the process.
We tried Enterprise Architect for a couple of weeks but failed to get something out of it. A high price tag for this learning lesson.
Architecture We stripped down each point of the wishlist into a traceable item in Jira and Confluence. We meet with our customer and negotiate about prioritization of each item aka feature or feature set. After the prioritization, we talk about milestones, due dates, and costs. By talking about time (implementation and future usage) and costs, we decide about the tech stack. The customer decides indirectly by agreeing on our rough time and cost plan. With the tech stack given, we plan to draw a picture of data flows, state and state changes. We translate it into a Confluence diagram and share it with the team.
From UI over resources to the architecture. We do this because we can't tell our customers that something is impossible because the chosen tech stack or parts of it doesn't allow us to implement certain features.
Denny Trebbin
Lead Fullstack Developer. Experimenting with bleeding-edge tech. Irregularly DJ. Hobby drone pilot. Amateur photographer.