This post will likely evolve as I study this topic a bit more. I've been fascinated with maturity models since 2000 when I was first exposed to the Capability Maturity Model. I've seen various models that were prescriptive in the types of changes required to mature. Some of those were reflective of the bias of the one propagating the model.
I created one for Accessibility that focused on change management and developing a mature program. It was less focused on improving accessibility as that was a byproduct of the healthy program.
I created one for Accessibility that focused on change management and developing a mature program. It was less focused on improving accessibility as that was a byproduct of the healthy program.
Mindtree's Model
Mindtree has an Automation Maturity Model. It is less prescriptive than models I've seen, but otherwise is similar to the Capability Maturity Model 5-level structure. There appears to not be widespread dissent on this model, so it forms a good basis.
- Ad Hoc: Minimal automation that is individually driven, script-based with no formal evaluation or tool adoption. May experience 5% productivity gains.
- Opportunistic: This is pain-based automation which is team or project driven, reactive in nature, and is tool / evaluation based within that context. May experience 5-10 percent productivity gains
- Systematic: Automation targets are objectively (metrics) defined, with experts leading the automation. It is proactive with a defined roadmap. Gains may be in the 10-20% range.
- Institutionalized: Systemic automation scaled across the organization, led by the organization. Gains may be in the 20-30% range.
- Adaptive: Automation becomes adaptive, with widespread use of Machine Learning, including self-healing and auto-optimization. Gains may exceed 30%.
This is more descriptive of a journey. Individuals automate themselves. At some point, a project automates to solve a quality or efficiency problem. If successful, management will try to inculcate automation across a smaller organization, measure it, and plan it. If it succeeds there, it will scale to cover the rest of the organization.
However, Machine Learning cannot happen until you have an active, objective plan for automation that has successfully replicated itself across the organization.
How long will this take? Twenty years ago, the adage for movement from CMM Level 1 to 2 was 18-24 months. The move to CMM level 3 was 24-36 months. Assuming these hold, then it can take four years to go from ad-hoc to Systematic. That's if we assume it is a journey.
KPMG's IA Maturity Model
KPMG has a separate model that focuses on leading a machine-learning initiative. If we ignore the ML piece, we see their 12-month target from Mindtree's ad-hoc to Systematic. I like KPMG's "Organized" terminology for Level 3, and think "Systemic" applies better to Level 4.
Test Project
Test Project shows an interesting focus on People, Process and Technology which has to be understood.
--
Ben
In tenebris solus sto