In Data and IT, there are various levels and classes of complexity which our design methods need to adapt
Choosing our methods for consulting in design, architecture, and strategic consulting need to understand the environments. For instance, Enterprise Architecture came of age during the heyday of MIS in the 1980s as standalone applications on mainframes moved to various technologies, but still for predictable mostly common corporate functions. Solution Requirements processes for Simple landscapes needed a solution to address more interconnected, yet static rugged landscapes. EA came about as a perfect solution into the mid-90s to help increase process, data, and technology integration.
But along came the concept of using IT solutions for the end-consumer via the internet. EA was still pitched as the savior, but over the first decade of the 2000s solid decade-long strategic plans ceased creating valuable direction, and EA became more of a nuisance.
- The rugged landscape melted in hot lava and became very fluid. Moore’s law kick in on a hockey stick curve as well introducing so much technological disruption, the enterprise became overwhelmed.
- The lower cost of technology has completely change the data supply options as well – data is coming from new sources that weren’t considered before.
- Private and Public Partner data sources instead of employee collected
- New interfaces and devices in pockets are helping people crowd-source data or conduct data production where considered “dirty” only a handful of years back, .
- A plethora of cheap Low to high-resolution sensors creating a data Armageddon of where to source from and soon they are being connected into the Internet of Things,
- Project Management methodologies adapted instead to responding to “epic” challenges of the period, and luckily technology caught up so the design and development could be done in shorter cycles (no more waiting)
Looking at the 3 facets of complexity theory, we can understand better how to place our finger on what level of Operating Model change is needed
1. There are various landscapes to the organization and problem space we support. Simple, Rugged, Dancing.
- Simple landscapes are like a single mountain peak. We know the challenge, its not moving,
- Rugged landscapes are a complicated set of multiple peaks like a mountain range have many peaks, but are set. What one group may see as their peak is either small or large compared to another group in another area. But, just like a mountain range, generally speaking, the peaks don’t change locally, just perspective of peaks are one moves across or interacts across the different areas
- Dancing landscapes are like rugged landscapes with multiple peaks, but throw time in there, and its more like ocean wave peaks. There are many variables that change over time, and what may be average predictable peaks one month, is completely different the next, or a tsunami comes and breaks the operating models apart.
2. In addition to the landscapes, we need to classify the level of change
- – 1 – Static, Steady State, unchanging, achieved equilibrium
- – 2 – Periodic, Seasonal, Regular Cycles, Sinusoidal
- – 3 – Chaotic, Sensitive to initial conditions, Complicated, enterprise
- – 4 – Complex, Non periodic, highly interconnected patterns, emergent structures and functionality, highly disruptive
3. There are several dials and variables to help classify and determined which landscape and class one is dealing with
This includes dials such as Inter-dependencies, Connectedness, Diversity, and Adaptive/Learning. These variables have measures to consider in Quantity, Quality, Risk, Maturity. For instance, Diversity measures variation, entropy, and is very attribute-based. An overly diverse ‘dial’ is en route for collapse, where homogeneous means its stubborn.
Given that, maybe there can be a cheat sheet after assessing the culture and business that helps select which consulting and design methods to apply?
Focusing on Data and IT Solutions for this example, lets
|Landscapes / Classes||1 (Static)||2 (Periodic)||3 (Chaotic)||4 (Complex)|
|Simple (Stand-along Systems)||Operations & Maintenance||Waterfall Design||Large Project Waterfall Design||Agile Architecture|
|Rugged (MIS, Internal)||Solution Architecture||EA||EA||+Segment Architecture|
|Dancing (B2C, B2B, G2G, etc.)||Agile Architecture||+Segment Architecture||+Operating Model Design||+Risk Portfolio|
Taking this model as an opening salvo, consider how your design services are being applied.
- Did you move from an Enterprise MIS (Rugged, 3) to the web to offer self-service solutions (Dancing, 3), did you take your MIS Operating model, and put a web front-end on? When you started delivering data as a service
- Did you move your major information product from using internal staff collection (Rugged, 2) to provided as a data generated product (Dancing, 3) and furthermore then provide as a service (Dancing, 3)?
In these scenarios, we have found our clients have circled back each time to needing to re-look at the Operation Model Design. Revisit who they serve, their challenges and drivers, how funding and revenue changes, what the new value chain is, and how their technology and workforce strategy changes? This requires extending off previous EA efforts to taking chunks of the programs a segment at a time and considering what a new operating model design may be – all the while, changing their tactical development efforts to be more agile in the mean time to keep up with the new demand and inefficiencies of the existing model they are living in.
If operating model redesign is just not in the cards, consider this one complexity theory principle: “He who has gets”. Where ever the current operating model provides leverage, even if misaligned to the new landscape needs, they/he/she will continue to get the inertia, the funding, the scope, and provide the gravitational force of where one is going. If that doesn’t align, then hopefully, the power-that-be can discern these misaligned forces, and consider maybe it is time to “change the game”.