In IT, “shift-left” means integrating testing, quality checks, and feedback loops earlier in the software development lifecycle (SDLC). Instead of finding problems during final testing, when fixes are slow and expensive, teams identify and resolve them at the earliest possible stage.
Interestingly, the same idea can be transposed to management.
The management lifecycle consists of defining the mission, values, processes, and structures, and then conducting the work accordingly. It is a lifecycle because management must periodically inspect how the system performs and iterate the whole system to adjust to new realities.
Instead of fixing problems late in the lifecycle, we should try to address them in earlier stages. To prevent operational problems, improve processes and structures. To prevent processes issues, improve the mission and values. 1
Shift-left aims at fixing root causes, not symptoms. In IT, scaling issues often lead teams to add servers or optimize queries. But if the software design wasn’t built for scale, these are temporary fixes. The real solution is to fix the architecture. In management, cross-team performance issues often lead to more meetings, stricter coordination, or additional oversight. But if the team structure creates silos, no amount of coordination will help. The real solution is to reorganize teams to match workflows. In both cases, fighting operational limits treats the symptom. Fixing the underlying design addresses the cause. And the earlier you do this, the less it costs.
In IT, you can test changes on test systems before rolling them out. In organizations, this is harder. You have only one organization, and you cannot simply “roll back” a failed culture change. However, the patterns still apply. You can run A/B tests or “canary releases” by piloting new processes in a single team before a company-wide rollout. You can also improve quality through design reviews and retrospectives.
There are many workable parallels, because IT is about engineering systems, management is about engineering organizations. Both require designing for reliability, iterating based on feedback, and fixing root causes before they propagate.
- For instance, to prevent bugs in production, improve the testing process. To improve engagement in testing, anchor quality and testing as a value. ↩︎
://ewernli 