chin fan soProgram Manager| Galaxy EntertainmentTaipa, Macao
In Agile projects, timeline cannot be determined before running a couple of sprints, then how do you persuade the project sponsor to let you start a project without timeline? Do you ever come across a situation like this?
Any comment would be greatly appreciated! Saving Changes...
In any project (not just an agile one) it takes time to validate whether actual progress matches your estimate. That doesn't stop you making an estimate from the beginning. If you plan to deliver in time-boxed iterations then long-range estimation is simplified a little: your estimates only need to be accurate to the nearest sprint and it's normally safe to assume that any dependencies can be worked out by appropriately sequencing the backlog. As for persuading your sponsor, the sponsor should appreciate that since the delivery is iterative the "end" point matters very little. Business value can be delivered at every sprint and the final sprint should deliver the least important things. Saving Changes...
Regardless of the delivery approach, if you are asked prematurely for a time or cost estimate you will likely be providing a low confidence guess.
Your funding approach needs to align with your delivery approach. If the sponsor is willing to pay for a few sprints (assuming you are following an iteration-based delivery approach) to help the team explore the scope at a high-level, to come up with some candidate architectures and to do some actual construction work focusing on work items of higher uncertainty then the team would increase their confidence in the solution approach and estimates for a preliminary release.
The other alternative is to take a cost/time-boxed approach - the sponsor sets limits on those and the team works to deliver at least an MBI or MMR which fits in those. However, if they have never worked together and have little understanding of the domain or product, the risk is that these limits will be set too low to deliver something useful.
Kiron Saving Changes...
Stéphane ParentSelf Employed / Semi-retired| Leader MakerPrince Edward Island, Canada
If I had such a question, I would ask: how much money do you want to spend?
Assuming your iteration costs are fairly static, you can then tell the sponsor how many iterations her budget will purchase. Saving Changes...
Thomas WalentaGlobal Project Economy ExpertHackenheim, Germany
Actually in many cases these are separate but connected projects.
Beginning with a pre-project, and having the objective to find out more about the problem and it's solution (scope, time and cost) up to an acceptable level. The result is a project proposal, a business case. Then chartering the implementation project, based on the business.
And probably you also should have a post-project, once the project product is handed over, some call it hypercare, rollout or maintenance.
To your question: you might propose to the sponsor to charter a pre-project first. It reduces their risk, gives them more security about what lies ahead and shows that you are a caring project manager. Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
The timeline is your spints. For example, if you have 3 spints 4 weeks each then your timeline is 12 weeks. In waterfall based project is the same the problem is people do not understand that estimations must be reviewed after different phases because the amount of infomration you have (that what extensely demostrated by Barry Bohem). With that said, there is a trap in what you state. When you work using an agile based method you are publishing that you will have a deliverable ready for using in a date. So, is not valid to say that you will not know if you do not run spints. Sprints are created and defined to get a goal. What you do by running spints is to adjust the way you go for the goal. Saving Changes...
Sometimes it is not up to the sponsor to make such decisions. Project sponsors are not always project Gods. In some companies there is a rigid process needed to initiate a project that requires approvals. If you don't have a timeline the project may not be approved and the sponsor may not get the resources needed.
I agree with Thomas you may need a pre-project or some other sort of activity meant to find out more about the problem so that you can come up with the timeline for the actual project.
Agile does not work very well with projects when a plan for a longer period of time is needed. Saving Changes...
