Is it Still ‘Agile’?

I was recently leading a vendor’s engagement where the client’s executive expectations were left unchallenged and process debate trumped progressive achievements. The goals were lofty, important and of a sizable magnitude for a project team comprised of both customer and multiple vendor members with widely variable experience.

Trust and patience had worn thin. The political goals were too costly to adjust. Our project was in crisis.

The mandated methodology was a traditional waterfall approach. The belief was that the statement of work predicted outcomes and we would perform to create those per a master schedule spanning multiple years. In fact, the “What” of our scope was highly generalized and the “How” was left to be proposed and approved. The “When” was ironclad.

There’s nothing new to this scenario. One common driver is the belief that capturing a fixed relationship between scope, schedule and cost will minimize risk. As we have seen in practice, and as the research has shown repeatedly, it does not. https://learn.g2.com/project-management-statistics

Let’s go “Agile” was a rallying cry as a panacea to our circumstances. So, the discussions ensued, and it was agreed, in spirit, that we could do “timeboxed iterations”.

Timeboxing is the process of fixing the iteration date and adjusting the scope and resources to deliver the chosen tasks and demonstrate their value.

In Scrum methods the timebox is a repeated unit of time, say a month, so that what is measured is “velocity” or how successful is the team at delivering value (scope x quality/effort).

In contrast, iterative and incremental development has timeboxes of varying length to match the goal for each cycle.

In either case, these processes assume the evolution of the scope, schedule, and cost based upon adaptive inputs from a product owner or other governing bodies.

As we assembled the documentation of our new process, we quickly realized that without the cultural support or the shared vocabulary for an agile practice— ‘timeboxed iteration’ really was going to mean the continuation of a waterfall SDLC. That SDLC could be completed within a timebox or not. Regardless the sponsors needed to measure our ability to predict and perform against the fixed-priced contract.

This led to a strange yet somewhat effective hybrid approach.

Our end to end waterfall plan was decomposed into a scope that was timeboxed for inspection every month and every quarter. The monthly timeboxes could have scope move in and out of them determined by the teams doing the work. These ‘sprints’ were governed like a Kanban with sprint planning, daily stand-up, and retrospective meetings. The quarterly timeboxes were fixed demarcations of scope against contractual obligations.

Development cycles would be deployed to environments that could be inspected for completeness, but not through the entire SDLC that results in User Acceptance Testing and promotion to a production environment. That would eventually be achieved over a two-year period of quarterly timeboxes. Any adaptation would be addressed through contract amendments based upon quarterly discovery and review of timeboxed achievements.  

The goal was to fit inside mandated contract governance, allow for executive review and validation while enabling the working teams to self-manage. The Work Breakdown Structure was transposed from the end to end waterfall plan to the sprint backlog.

To date, this has begun to rebuild trust. I like to think this is from the power of the “Lean” principle of “making work visible”.

This was a bootstrapped project salvation. It remains largely waterfall in principle and governance. However, by burnishing the process with some of the lean and agile concepts does this hybrid method belong somewhere on the continuum of agile methodologies?

Often we need to pick and choose from all the tools available to tinker and jury-rig continuous improvement from wherever we may find ourselves.

What do you think? If ‘waterfall’ is timeboxed has it become slightly ‘agile’?

I’ll be watching and learning as things unfold.