Could you give some recommendations on first pilot projects that tries to exercise a new development process ?
I have an example:
Which recomendations would you give to a customer ( with very little process maturity ) if he decides to subcontract a provider ( with high maturity process ) that will attack a huge project , and this project is the first one in which the customer will use formal software process ?
This is the explanation: I have a customer that has a lot of resources, his is just starting to know about process, we recommend them to go in slow and tried to applied the different trainning ( Requirements, PM, etc. ) in a project with low risk and medium to small size. He refused to apply the process in a minor project, he will attack instead a multimillion project as his first pilot project, and his strategy for mitigate the risk is to go international and subcontract a big company with high process maturity ( CMM3-4-5 ), that can provide better chances to finish on time, on budget and with high quality.
1- What would you recommend if they insist in doing this ? Could you visualize possible risks or dangers in going for this option ?
2- Supposing they are already on the road, what aspects would you take more care, emphasize, prevent, etc. ? Saving Changes...
Sort By:
Michael WoodProject Manager / Business Analyst / Business Process Improvement Guru| Independent ContractorGig Harbor, Wa, United States
The first thing that I would be concerned about is the cultural impact on the organization. The time lines and disciplines may not be fully understood at this point. By phasing the project you can keep risk low. Maybe the software project can go through a series of discovery - requirements - design - contstruction - deployment cycles. Focus first on getting the database designed to at least the table and attribute level (not full schema and views)
Use lots of prototypes
Remember that CMM levels is theoretically wonderful but adaptation to the culture and what can work is best.
The size of the project should be balanced by the payback and associated risk. Keep both on the table
Good Luck Saving Changes...
Stephen MayeSenior Vice PresidentVa, United States
There are several things going on here... The discussion about the client's desire to develop process maturity (I'm referring to practical maturity that may or may not be demonstrated via a CMM level...) needs to be taken in light of the off-shore direction that you mentioned. The decision to take large projects off-shore should be part of an application develop strategy--in support of a larger business strategy (I'm starting to sound like Michael...). Anyway, my point is that if your client intends to move toward oursourced development of large applications as a strategy, then their need for mature development processes in house is obviously impacted. More to the point, like any other project, your client needs to do a serious requirements analysis for what their process needs really are. They may find that their primary focus (going forward) will be on requirements, vendor, and contract management and very little on traditional development processes. (If I decide that I'm not going to do my own home maintenance and upgrades, I'm not going to spend a lot of time understanding plumbing details. What I will get good at is: finding, sellecting, contracting, and managing my contractors. You know what I mean...)
Back to the original question... IF the customer has a genuine need for developing, implementing, standardizing, and maturing development processes in house, then there are some things to keep in mind.
By going off-shore (I assume that's what you mean by "international"...), the opportunities for knowledge transfer from the service provider to the customer are greatly limited. My experience also tells me that the off-shore shop will not be as consulting-oriented as some other providers would be and thus further limiting knowledge transfer opportunities. If you hire a high maturity shop but intend to continue to do your own development, one of the benefits you should get from that is learning how they do it--i.e., benefit long term from their process.
In regard to pilots, etc... I think who is on the pilot team and who sponsors it is as important as the project profile. What you are looking for is the chance to demonstrate value, learn something, and dissiminate it more broadly. Look for a project with the best (and best respected resources), a high-profile influencial sponsor, and high bang for the buck (visibility, business benefit, etc.). But, your right, don't experiment too much on the core policy writing application for the world's largest insurance company first time out of the shoot.
I hope this at least spurs a thought or two... Saving Changes...
