Responsive strategy, being and doing Agile are keys to effective adaptive leadership within an Agile organization. Jim Highsmith spoke recently at the AgileAustin hosted event about adaptive leadership and how it can accelerate and sustain the organizational agility. He said that agile leaders should focus on aspects like quality and speed to value as executive levers to achieve business goals.
The leaders should create an Agile business strategy with focus on responsiveness over efficiency and develop an IT Agility roadmap. The strategy should also include a continuous delivery of the software products. IT teams shouldn't just limit their process to Continuous Development or Continuous Integration but enhance it to include the Continuous Delivery (which includes all three phases, Continuous Development, Integration and Deployment). Jim talked about different levels of agility in organizations like operational, portfolio, and strategic level agility. Everyone should start looking at the areas within their companies where responsiveness and agility could be strategic as well as tactical/operational to the business.
The adaptive leadership is two dimensional which includes doing Agile and being Agile. Doing Agile requires leaders to understand strategic agility from a business perspective as well as specific principles and practices to help build Agile organizations that can weather business turbulence. Agile leaders should use the following execution levers to achieve business goals of responsiveness (agility), profitability, market share and customer satisfaction.
- Quality: One of the quality aspects is managing the technical debt which, if not addressed correctly, would lead to high cost and high risk. The leaders should focus on qualitative as well as quantitative metrics and an incremental investment in addressing the technical debt situation.
- Doing Less: The project teams should do the simplest thing possible that delights the customer. Jim noted that 64% of features are never or rarely used in the applications so it’s important to deliver the right features. The delivery value can be achieved by reducing work-in-process (WIP) and multi-tasking (duration/effort). Agile value curve can be used to prioritize features between different projects in an organization to realize agility at the organizational level, not just at project level.
- Engage/Inspire: He said Agile leadership should encourage and promote the concept of self organizing teams that have autonomy, mastery, and purpose.
- Speed-to-Value: The three components of Agile triangle - Value (releasable product), Quality (reliable, adaptable product) and Constraints (cost, schedule, scope) - need to be managed properly to realize the value. We should also measure the value down to the feature level. He mentioned about the Value Engineering Model used at organizations like GAP to measure the business value.
The second dimension of adaptive leadership is “Being Agile”, which is based on values and principles, requires leaders to realize how practices like continuous delivery and a mind set of sustainable agility combine to create highly responsive IT organizations.
Adaptive leadership behaviors are different from the traditional leadership ones. A traditional manager focuses on following the plan with minimal changes, whereas an agile leader focuses on adapting successfully to inevitable changes. Jim also mentioned other agile leadership characteristics like being adaptable versus predictable. Agile leaders also follow an Envision-Explore approach rather than a Plan-Do approach. They are facilitative in nature and encourage a collaborative engagement among the agile team members. He said that self-organizing doesn't mean anarchy and adaptive leadership doesn't mean no leadership. As a leader, it's ok to be wrong but it's not ok to be uncertain.
Jim concluded the discussion by saying that all models (Waterfall, PMI, CMMI, Deming, Scrum, XP, Kanban, and Lean) have some limitations but all of them are potentially useful and the best agile implementations are usually "hybrid" implementations. We should improve the effectiveness of each model based on the situation.