Zack KanterVerified account

I'm been running on informal (less than one hour of t-shirt sizing every ~2 months) software dev estimates for about a year now, having an estimate on total work from around 50-70 days of work, and we have consistently landed in the range every time. 1/

Recently I've been back into several situations where I've been asked for specific development plans, where having a highly granular development roadmap with associated epics/tickets/etc is viewed as a proxy for reliable and functional SDLC. 2/

However, all I see on that path--at this point, at least--is the painful overhead of building those roadmaps and keeping them up to date and handling adjusting them as certain things take less or more time than expected, and then managing people around those. 3/

I have a hard time taking what feels like it would be a 10-15% tax on development to maintain those when the weekly demo and frequent merges to trunk and availability of anyone to run the trunk application proves effective SDLC more than anything else (IMO). end/