Key insights about the system behavior of enterprise IT
Why enterprise IT is often criticized for not being responsive, transparent, flexible or innovative? Regardless of industry and geography, why so many IT organizations are suffering from constant delivery delays, increasing costs, growing technical debt, and demoralized workforce? What is driving shadow IT?
Our hypothesis is that most IT organizations are unable to realize their full potential not because of their leaders or people, but because of the outdated technology management practices that are optimized for the industrial-era and under-performing in the digital-age. To test our hypothesis, we conducted a comprehensive study and documented our findings and approach below.
Finding #1 - Effective IT consistently outperforms efficient IT.
The diagram on the right is inspired from the modern portfolio theory of finance and it shows several portfolio frontiers, depicting the best business outcome performance of a given bundle of technology management practices and policies for a defined level of technology investment.
Black line - Is associated with efficient IT management practices that are ubiquitously employed by most IT organizations today.
Green line - Includes emerging concepts in technology management such as backlog, dependency, inventory, batch-size, cycle-time, slack-time and cost of delay.
Blue line - Leverages same emerging concepts as the green line and augmented by several predictive management decision support algorithms.
The horizontal axis shows the amount of technology spending approved at an enterprise, and the vertical axis indicates the business outcomes realized as a percentage of all opportunities identified by the business teams.
Finding #2 - IT throughput, a component of IT effectiveness, fallows an S-shaped curve as the organization size decreases (constant demand)
When backlogs are under-sized (Point C), lower-value work creeps in, as partially utilized resources start working on ad-hoc, unplanned, or unfunded activities.
The portfolio throughput is maximized when teams are rightsized relative to demand put on backlogs (Point B).
When demand exceeds capacity, backlogs are over-sized (Point A in the diagram) and the IT organizations starve for resources. Consequently, tasks take longer to complete, delayed dependencies cause a ripple effect throughout the organization, shuffling tasks and firefighting become a norm, and ad-hoc executive interventions produce sub-optimal prioritization decisions.
The most critical implication of this curve is that overemphasis of cost savings for an extended duration progressively moves IT organizations towards point A where throughput is dangerously inhibited. Budgets and schedules may seem unaffected, but functionality and quality suffer due to relatively weaker controls applied to them. Concerned stakeholders call for additional oversight and funding restrictions, which further reduces the team size.
The above study is conducted on our Technology Management Lab (TML) where we completed hundreds of simulations and analyzed IT operations in a two-thousand-plus-year perspective. You can find more details about our approach here.
Management rules and policies unitized in our study
Efficient IT (Black line)
This bundle was designed to represent the most common form of management practices employed by IT organizations today:
Investment planning cycle runs periodically; proposals are evaluated at an program/ project level; priority decisions are based on the proposed business value, e.g., ROI; individual proposals are approved until the available portfolio funding is depleted.
Approved demand is decomposed into smaller scope, e.g., feature, story, scenario and then assigned to individual teams for execution.
Work is executed according to a predefined project or sprint schedule.
Team capacity is balanced with the planned demand volume; teams have dedicated FTEs and a portion of FTEs are reassigned between project phases / sprints; team backlogs are prioritized based on business value, e.g., ROI, schedule commitments and known dependencies.
Effective IT - Standard (Green line)
The goal of this policy bundle was to test the effectiveness of commonly known lean IT management practices, like cost of delay, weighted shortest job first, and continuous improvement.
A significant management innovation tested in this policy bundle was the introduction of a formalized causality relationship between the technology spending decisions and subsequent business outcomes.
Effective IT - Advanced (Blue line)
This group not only utilized the standard lean IT management practices, but also employed several proprietary optimization algorithms to improve the overall investment portfolio quality, supply and demand balance, flow of work, inventory control, and benefits realization.