2 ms·
My advice as a manager is as follows: 1) You always have time for tests. Your estimates should include tests and testing. If someone asks for a feature and you
by Jemaclus 3y ago
My advice as a manager is as follows:
1) You always have time for tests. Your estimates should include tests and testing. If someone asks for a feature and you think it would take 3 days to test, half a day to write tests, and half a day to do manual testing, then it will take 4 days. That's your estimate. You always have time for tests. Tell your team this with every ticket. Don't sandbag the estimates, but make sure you are planning time for your team to ensure high quality.
2) One of your many jobs as a manager is to protect your team from external pressure. Think of yourself as the submarine (too soon?) that keeps your team safe from the pressures of the business. The team should be focused on shipping high quality features, not whether the CEO is asking for something or not. If a feature needs more time for testing, it's your job to take the heat for it.
Both of these will give your team the breathing space to actually write tests and actually test their code, without fear of reprisal and without fear of losing their jobs. Right now, they probably have very little trust in this process. It's your job (another one!) to prove to them that they can trust you to protect them as they do their jobs. Over time, the team will develop a culture of testing and releasing higher quality products.
You must remain consistent. You should hold the line, hold your team to a higher standard than exists now, and hold your external stakeholders at bay.
Additionally, I personally don't believe in deadlines. With few exceptions such as legal/court orders or certain contractual obligations, deadlines are arbitrary dates made up by someone for some arbitrary reason. From a financial or company survival standpoint, there's very little difference in real terms between releasing a feature on Tuesday and on Wednesday or Thursday or even Friday. I try to think of these as target dates. We will do our best to meet them, but if we need an extra day to test, then we'll take it.
Hope that helps. Good luck.
- odle 3y ago1) and 2) I can't agree more. However, your guess is right, the team have little trust in the process. Now that there aren't really business critical new features, I think I'm successully not giving them pressure. But, as already happened in other teams/context, when the time for an important business feature comes, the director/business could "override" me and giving the team extreme pressure. About the deadline, I agree with you as a professional, but in this company they like to organize launch keynotes and stuff like that so we need to ship within the keynote date. (vent mode) You wonder how the keynote date is set? I don't really know, with my boss I really don't see any transparency in that