Check out https://www.dataopsmanifesto.org/ for some ideas. I haven't had a chance to implement it, but it sounds like the direction I want to go.
I think general SDLC is a good foundation, but I think good data development is more standardized and should therefore use more guidelines and LESS requirements.
The best I have seen is the POSMAD model presented by Danette McGilvray, Granite Falls Consulting. POSMAD refers to Plan, Obtain, Store/Share, Maintain, Apply and Dispose. Attached is a link to her website which provides some free downloads of the model. Her book Executing Data Quality Projects: Ten Steps to Quality Data and Trusted Information. is an excellent book, if you don't have it on your shelf you should pick one up.
Beth Steffey, MBA, CDMP
I've built a data-centric lifecycle that was used to complement the company SDLC. The idea was that plenty of time was going into developing software, but not enough time to developing data management components. People didn't really understand what they needed to do. However, the concept are so similar. You still need to plan, design, test, etc.
Wow - Thanks for all the replies, everyone - after this question
was sitting out there for over a month I was beginning to think no
one was visiting the forum any more!
Mark: Yes, POSMAD is a pretty good model, and McGilvray's book that you cite is one of the best books on data (not just DQ) that I've read (and I'm a tough audience). The only drawback is that as a term "POSMAD" is (IMHO) a bit clunky.
Beth: The DM-BOK data lifecycle is compatible with POSMAD and is good. And, again, as a term, "data lifecycle" doesn't really convey much about what it is.
I come from a manufacturing background and have been toying with the idea of "Data Supply Chain Lifecycle" - DSCL. I think the idea of a "supply chain" covers all the POSMAD/data lifecycle phases. And the acronym - DSCL - is an inside-out version of SDLC, making it both memorable and a more directly-relatable complement to SDLC. (I know - it's kind of corny, and maybe I'm putting too much weight on a name, but I often think: if it ain't catchy, it won't catch on.)
Gail McAuliffe, M.Sc -CIM
I like the idea of a data supply chain. Supply chain management is something that non-data people can readily understand. Although as a Data Management person in the IT department, I can more readily gain traction with IT colleagues by referring to Lifecycle Management. In fact, as I type this I realise that we use lifecycle management to refer to asset acquisition, maintenance, and disposal. So perhaps lifecycle management is relatable to to the entire organisation.
Why not an agile methodology? You would need to know the tasks involved in the project and load the backlog, and you would need to know the approximate ordering of tasks so they can be effectively pulled into sprints. While data projects can be categorized (i.e., data warehouse, MDM, big data analytic cluster, etc.) with each having a roughly consistent set of tasks to do, no two projects are the same.
If you're looking for a name to wrap around the practice, I generally refer to this as ILM - Information Lifecycle Management.
Have a great weekend, all!