Water-Scrum-Fall is the norm in many organizations today. Despite the attempts of scrum coaches and consultants, the weeds of waterfall grow back into the agile garden. What causes this? This article looks at the root cause for the water-scrum-fall phenomenon and makes a suggestion about how to address it.
The Root Cause
Water-scrum-fall’s reality is not the result of people being unwilling to adopt scrum. It’s not the result of a lack of passion for agile processes and practices. Nor is it caused by a lack of executive support. The cause?
In companies that produce software for internal use water-scrum-fall finds it’s greatest adoption. Internally used software is strictly accounted for by the regulations in SOP98-1 and this financial machinery is what guides the need to plan up front.
Capital projects are investments. To determine an investment’s return you need to know the estimated initial cost, and estimated revenue ( or savings ) the project will create. These things are estimated up front so that a decision can be made on whether or not to pursue the investment. The up-front nature of capital budgeting compliments waterfall and BDUF. It is a core financial business process guided and regulated by FASB. Think about that for a moment…and then read on.
Scrum practitioners run up against a wall with capital budgeting. It doesn’t fit their operational practice for developing software. Under scrum….we shouldn’t be designing, estimating and crafting the project up front. Instead we should approach it incrementally. The problem with this? It ignores how enterprise software development projects ( capital investments ) are funded. The result is that everyone compromises and innovates. Water-Scrum-Fall is the child of this compromise.
Scrum and other agile practices pose a challenge to enterprise software development efforts and the capital budgeting process. Indirectly they say “Why are we funding this as a capital investment? It’s not. It’s an ongoing operational cost and should be accounted for that way. If we don’t plan on funding this software development effort for the long haul…then why are we doing it?” Funding a software development effort as an operational expense, as is done within software companies, does fit the scrum operational practice better. But again…the difference between a software development effort being labeled CAPEX or OPEX is guided by FASB. It’s not up to the company.
How Do We Fix Water-Scrum-Fall?
I don’t think there’s a silver bullet here. But in my last article, Is There a Better Way to Estimate Capital Projects? , I threw out a suggestion for how to estimate a capital investment using tolerances. This bypasses the need for an up-front detailed analysis of what the the LOE ( Level of Effort ) would be for the project, but still gets the business what it needs: an initial funding point and resulting NPV that augments decision making.
So water-scrum-fall is a pragmatic reaction by agilists and IT professionals to work with the business and its financial processes. Did the originators of scrum not understand the capital budgeting process? Were they oblivious to the financial architecture of the businesses around them? Maybe…but to their credit; they weren’t trying to address this. Their focus was on how to do software in an adaptive fashion so that it more organically addressed the operational realities of manifesting a complex vision.
I think this is good insight and may be one cause of Water-scrum-fall but there are many other causes besides just this one, I think…
I agree with the analysis and this matches what I’ve seen in the market. I was initially put off using agile because of the intransigence some practitioners to this reality. The result isn’t ideal but I believe we’re making progress on this point.
I also loved your previous post on a better way to estimate capital projects. Have you actually used this method? One side effect, and a great one, is that it re-focuses the effort on thinking about the return and not just the costs. This return is derived from business value and not just software. Too much software is created with little or no thought to tangible business value. This side-lines the software and estimates the value of the business enhancements. Brilliant!
thanks for all information your site very ok
This is so interesting! I don’t think you’ve considered the human factor, but still a good post.
Thanks Marg. Check out these posts to if you liked Root Cause of Water-Scrum-Fall:
Is There a Better Way to Estimate Capital Projects
Why Agile Adoption Fails In Some Organizations