3 ms·
Their conclusion is a bit off. The WBS and GANTT charts are used for resource planning. I don't work for a unicorn so resource constraints mean that I actuall
by myohmy 5y ago
Their conclusion is a bit off. The WBS and GANTT charts are used for resource planning. I don't work for a unicorn so resource constraints mean that I actually need to plan, prioritize, and schedule work. If I do a bad job of that then my developers get laid off, we don't get paid for slack.
One thing they miss is that managers manage expectations both up and down. People only get angry when you step outside of their expectations. I work in a PMP setting and I am up front in my project plan about how we're using PERT estimation (so they always look high), that we're dedicating time to testing, automation, and documentation, and explain why. Its straight up in scope. The client signed it. The CIO signed it. Problem solved.
The one thing I find is that developers chronically underestimate. And that's something that I found going from a programmer to a project manager. I always thought about my code and never the back-and-forth communication with stakeholders, and the change requests, and all the little BS that inflates estimates. That's why I teach my programmers PERT estimation, because it forces them to think about what the doomsday scenario is. We've gotten pretty darn good at estimation these days, which is good because we are a consulting shop, so if we underestimate we don't get paid, or have to get a change request.
- NAHWheatCracker 5y agoThere are managers like you, who understand estimation. Then there are managers like my former manager, who took a 2.5 year estimate from me and sold it to leadership as something we can do in 6 months. When he tells me the project got a green light, I pointed out that I also listed several risk factors that would be major blockers in the first 6 months. This manager was also the scrum master on the project, and proceeded to contribute nothing for the next 3 months. He was waiting around for a promotion to some sort of leadership/project management position. Luckily his replacement was a massive improvement. A year later and we were right on track for my original estimate but I lost all credibility in the org despite busting my butt to keep the project on the rails. It's not just engineers who are bad nor is it just managers. Both sides have to do their part and respect each other or it can go sour.