In my opinion, its a fast and extreme practice in the Design phase. It’s also throwaway prototyping till the last one survives with highest votes. It’s putting down all the thoughts of how the UI can improvise on the piece of paper, review, biased or unbiased by other’s ideas, inspired apps replica, bring on the favorite app features, aspired by Chinese apps, creative thinking — draw it and propose your ideas so that they are voted.
Design sprints usually last for 5 days. They are used effectively during two scenarios — When a new product is being created, and when an existing product needs improvement. It goes from problem to solution in just one week, and both the client and the end user are involved, which makes it possible to validate the ideas quickly. We conceptualize together, design and test and gain valuable insights.
The final product of the design sprint is an interactive prototype which is tested by a group of users. It could be a click model or a simplified app interface. As long as it is obvious what it is we’re actually testing.
- All devices — phones, laptops away until the day’s goal end.
- Sticky notes of various colors, 3 sets of sketch pens, chrome labels of two colors for voting, a stopwatch.
- Team — Not more than 12 members.
Step 1: List the Problem statements from the metrics you track. Eg: Amplitude, Woopra, Mixpanel and KISSmetrics cohorts, etc. Retrospect the change in numbers for the past 6 months and infer the changes since then.
Step 2: Tackle one problem at a time by drawing the existing designs and analyze where exactly are the users dropping off. Suggest alternative designs and approaches for the same.
Each person in the room can come up with up to 4 approaches in the span of 8 minutes on the stopwatch. Once done, pens down and display all the sketches on the walls of the room. If one of them is an incomplete sketch, never mind since each of us will be pitching the idea to resolve the problem statement. Questions are asked once the narrator explains each of his solutions. At the end of his exercise, each person in the room is given 3 votes and 2 minutes on the clock. Note that a person cannot vote on his/her idea. Highest voted design wins.
Step 3: Reiterate the highest voted idea on the paper along with the overall flow to check if it connects all pieces of the app. The best idea may not be the winner but this is a second chance to pitch better from the displayed sketches. Stick to your thought until there are valid reasons to drop it off for a market or feasibility etc.
The main idea of the sketching in this attempt to check how well the new design ideas stick to the existing workflow. Will the existing users get the hang of it or only a new user can appreciate it? This decision is tested along the lines of ALL types of users.
The lifecycle of a user — User Definition
- Active user: A user who logged any active event.
- New user: An active user who logged any event for the first time. A user will be counted as new for one full interval after their first event is received by Amplitude. For example, if your usage interval is set to 7 days, and a user was new on January 1st, they will be considered new for any events they perform between January 1 — January 7.
- Current user: An active user who was active in the previous time period (day, week, or month). For example, a user who was active on January 1st and January 2nd would be considered a current user on January 2nd for a daily lifecycle chart.
- Resurrected user: An active user who was not active for the previous interval (day, week, or month), or longer. For example, a user who was not active on January 1st, but performed an event on January 2nd would be considered a resurrected user on January 2nd for a daily lifecycle chart.
- Dormant user: A user who did not log any active event but was active in the previous time period (day, week, or month). For example, a user who was active on January 1st but not active on January 2nd would be considered a dormant user on January 2nd.
One vote and the final decision will be an amalgamation of the best features hand-picked from highest voted designs. This is how each problem is addressed in a rigorous Design sprint.
Before jumping into the next problem statement we also list and share the ideas inspired from our favorite apps who have been tacking the problem statement excellent. Eg: Listing of 5 apps to resolve the user dropoff from the Onboarding screen.
Note that the solutions may be not a permanent one. We need to analyze the numbers again before we decide. If it’s not adding value, maybe we need to brainstorm again to find a solution. Remember to measure. Everything that’s measured on a regular basic, improves.