Digital products have come a long way and so many teams are more agile than ever before. With continuous delivery, quick turnarounds and back-to-back release dates, running into production bugs every now and then is natural. At Marvel we, the support team, have developed a fool proof bug reporting method to help you squash those bugs in double time. ????????
Investigate it ????
Firstly, we investigate the issue and try to replicate it on our side. If replication isn’t possible we aim to fill a report with as much useful information as possible, by asking a sequence of questions to whoever spotted the bug.
Write it ????
Next, we collate this information in a report. One of the things that helps us to resolve bugs quickly, is speaking to developers in a language they can understand, so we use the following template to help describe each issue:
Description of bug
We write a helpful description of the problem. The more information, the easier it is for the development team to get an understanding of the issue, but we like to keep this straight to the point, no waffling here!
Steps to recreate bug:
Create an ordered list of each step taken to recreate the bug. For example:
- Go to download menu
- Click on iOS
- Download does not start
How we expect this to work
How the feature is (mis)behaving
Include a link to the effected project/screen/account area
Reference to browser console and network errors
If we can replicate the bug, we’ll pop a screen shot/recording in
Environment, Browser and OS version:
The browser & operating system the bug is happening on
When checking this bug, we include information on how many people this is effecting – this helps when prioritising!
Prioritise it ????
Once we’ve written up a bug report, we can set the priority. To help with this we use a matrix system that explains the impact and severity on a scale of 1-5. We assign a priority rating of P1, P2, P3 & P4, with P1 being the highest priority and P4 assigned to smaller issues, we then submit the bug to be prepared for discussion.
At Marvel, only P1 or P2 bugs are considered when we decide what will be fixed as we find that P3 & P4 bugs tend to be picked up by regular feature updates. So, even though we still report these, we don’t discuss these with the wider team.
Discuss it ????
Each week, we meet with product managers and check over the P1 & P2 backlog to see what needs to be pushed into upcoming sprints. Sprints last two weeks and are full of important feature work that we like to ship as quickly as possible, so there is a lot to compete with!
To help decide what makes it into an already busy list, we try to stick to strict fix deadlines that are assigned to priorities. These are that a P1 bug should always break a sprint to be fixed, or, be put in the next immediate sprint (up to two weeks) and a P2 bug should not spend more than 2 or 3 sprints in the backlog (up to 6 weeks).
We find that having these set rules and working from a small list of bugs in handy dashboards, allows us wider visibility of what needs to be fixed first.
Fix it ????
Wahey, the bug is now in sprint waiting it’s turn to be exterminated!
Test it ????????
We trust our engineering team but to be sure that everything is working as expected, we do a run through test to check that all is resolved.
Ship it ????
Now we can ship the fix and share the good news with anyone who found or experienced the bug.
Need help with a bug or question? Drop us a message.