Joseph FlahiffFounder| Whitewater Projects Inc.Bothell, Wa, United States
I would like to better understand the issues that people are experiencing in mixing agile into their waterfall PMOs or with other governance groups/models.
What are execs complaining about not getting? What are you doing about it? Saving Changes...
Sort By:
Joseph FlahiffFounder| Whitewater Projects Inc.Bothell, Wa, United States
I guess no one is reporting anything to management. Saving Changes...
Scott LilliePMO Manager| OPERSColumbus, Oh, United States
Hello Joseph, we have recently transitioned some our projects to Agile. The biggest struggle we encountered was defining and reporting on scope and value to our execs. The scope was not clearly defined in the beginning and we frequently heard the question of "What are we getting?" and "What does success of the project look like in the end?". We were able to define a MVP (Minimum Viable Product) for the project and eventually linked each user story back to features (deliverables) of the project. As the project progressed we provided a report listing each of the features and percent of user stories complete. The other challenge we continue to face is estimation of the MVP with out all of the requirements completed up front as in our waterfall methodology. We would like to be able to answer the questions "How much and how long?" next and provide a estimate that everyone buys in to. Saving Changes...
Joseph FlahiffFounder| Whitewater Projects Inc.Bothell, Wa, United States
Scott,
That is great! I can see a couple of tools that can help you out. they have to do with statistics (ewwww!).
First, you want to know how much work your MVP will be but you don't want to break down and estimate all of the stories. And with good reason. That is a lot of work and smacks of BUFD (Big Up Front Design)
That's OK. We don't really have to. if you have a good idea of your MVP. Take a look at the epics or features that make it up. Randomly pick a few (7-11) of them (it is important that it is random) and then break them down into the stories needed for the work. if you do that for somewhere between 7 and 11 epics you can estimate how many stories you will have in your whole MVP (heck the whole backlog) with a confidence of about 95%. Pretty dang accurate.
Next, you need to start working on it. get a few sprints in and we can have a look at your velocity trend. not a straight-line trend though, that you may have heard from some agile trainers. Use our old friend the Monte Carlo simulation to do determine the RANGE of dates that are possible. At that point you can have the discussion about risk tolerance. If they want a 99% accurate estimate OK but it is going to be out further. if you want a 90-95% confidence then it is a different date. it all depends on their risk tolerance.
