Why are agile organizations different from traditional organizations? Key differentiators are rapid feedback, short learning cycles and accelerated adaptation. Feedback enables agile centric cultures to calibrate practices that enhance team’s ability to generate customer value better than the competitors.
Team is the fundamental unit that produces value, not individuals. A scrum team enabled by agile principles, values, behaviors delivers value employing subject matter expertise (domain knowledge), technical practices(XP, reviews) and artifacts (architecture, design practices). As teams scale connected complexity increases. Practices within a team need scaling too. Collaboration,Capability and Culture form the foundation to scale agility across enterprise. Together, they foster a customer centric organization generating value.
Note: Terminology from Scaleable Agile framework (SAFe(R)) is used in this article. The 3Cs are foundations for any scale agile model. Reference URLs are provided where such terminologies are used.
Scaling agile to Accelerate Value
Organizations that are adopting SAFe framework employ coordinated teams called an “agile release train(ART)”. Even prior to these guidelines, organizations had teams that coordinate with each other. But when they adopt agile practices to deliver value, the challenges are not sum of challenges each team faces. Non -linearity adds complexity (and vice versa). With just one team, the flow is easy to manage. Right from start team members build a local context, and culture; they transition along the Tuckman stages from forming-to-performing in a shorter time. Collaboration is an easy outcome.
Not just collaboration, teams now have take effort to fit multiple local cultures. They also need to ensure that collectively, the capabilities complement than conflict with, each other.
Collectively 3C, collaboration, culture and capability form the foundation which rests the scaling advantage. What is stated here encompasses several aspects of agile manifesto and core values of SAFe(Scaleable agile framework). For example, transparency is a dimension of culture and collaboration, while design thinking can be considered a capability. Interplay of these three factors help drive outcome.
Scaling agile to deliver value requires high bandwidth collaboration between teams. But this is not an abstract idea. As size of the team increases (typically 50 to 125 in a release train), fragility can come from any communication failure. “For want of an email, the train can derail”. In a single scrum team, collaboration can be on demand. But that is a risk when working as a train. Agile teams power the train.
During a program increment, in one of the scrum of scrum meeting a dependency was high lighted; even a cursory analysis revealed impact on architecture. The architect and his manager were at technology workshop. The VP of that department was at a leadership offsite. The release train engineer (RTE) sought advice from other teams. Luck intervened. One of the team members provided guidance and problem was resolved. While the approach sounds inefficient, this was an example of culture, capability and collaboration coming into play.
In an alternate world, team would have escalated to the management or waited till the architect arrived. The guiding principle is — collaboration is tool to make system “anti-fragile”. But the culture has allow for initiative and risk taking — i.e, ensure that when luck decides to help, the team is prepared.
Leaders always like to promise and showcase their team to potential clients. In many cases, the leaders promise features earlier than planned. That is, in a shorter cycle than a program increment timebox. Faced with such opportunity and risk, the teams in the train need to collaborate, explore opportunities, share insights — in a very short time. This is “high bandwidth collaboration. The ART is always prepared to handle the unexpected.
Collaboration is a force multiplier that enables such a preparation
What is culture? ‘ It is the way things work here, we just do what we are told to do, it is just a job, we are told at the last minute, just do it, we work in an escalation culture’ — exemplify culture. For ART to work effectively, culture need to support or evolve to support collaboration and capability build up. Culture welds collaboration and capability. Say, transparency (as cited as core value in SAFe) is vital, and happens when openness is supported at all levels.
Evidence based problem solving can occur only when evidence gathering is valued over speculative approaches. Here we see culture of openness drives capabilities. Built in quality is achieved through capabilities supported by tools. A culture that prefers “defect detection” over “defect prevention, will appear reactive than proactive. But capabilities need to be engineered to prevent defects. Logging tools for example, may be needed. Team need to be trained on new tools.
Studies have shown, organizations embracing agility have not reduced bureaucracy. Such cultures prevent teams to handle customer expectations swiftly. Organization structures need to be fluid to adopt a shape to meet client needs.
Learning culture, enhances capabilities of the ART and the organization. Innovation culture enriches individuals, teams and organization. Innovation is single most important driver for standard of living in developed economies. Scaling agile offers immense opportunities to innovate at product and process areas.
One core example is as follows. Agility, as we all know, emphasizes empiricism over statistics. That is, agility values data from experience more than data from sampling. When such data is collected, teams can explore areas of correlation say
a) between estimated and actual time for a feature,
b) between lines of code and defect count,
c) between number of dependencies and build failures.
Such efforts cannot happen without support from leadership. Leadership builds culture. ART can then embrace a predictive culture.
Peter Drucker once said “Culture eats strategy for breakfast”. Experienced agile transformation leads would agree.
Why do organizations seek to scale agility? — Why not teams operate on their own instead of being part of ART? First, scaling agility enables organizations to deliver larger products/services. Long lived teams as in ART build long lasting systems and solutions. When teams collaborate together, such solutions are faster, well integrated and deployed efficiently.
ART needs to be organized to release value on demand through the Continuous Delivery Pipeline. They need to have the technical and domain capability to explore, integrate, deploy and release on demand. Capabilities need to coordinate the flow of intermediate outputs, to achieve eventual outcome. For example, Hypothesis validation during exploration requires SME engagement; Current work in progress also needs SME involvement. The release train engineer, being the steward of overall capacity need to strike balance between “needs” and “outcomes”. (Mathematically, borrowing from queuing theory, items being served need to have minimal service time to delivery but items waiting in the queue need to be explored to reduce wait time too.)
ART is the enabler of DevOps culture through technical and domain capabilities. These capabilities subsume agile competencies and add others to the portfolio of skills needed. But as stated before, these need to be supported by leadership culture, and management capabilities.
Dr. Nicole Forsgren, in her presentation at DevOps summit in 2017 cataloged the list of capabilities at different levels, that are needed for sustained evolution of DevOps (a key goal for ART). The key takeaway from the research data is speed and stability can be achieved together. The data replaces the myth that “ stability comes at cost of speed and vice versa”. Employing constructs model and validated by four year research, the list of capabilities are explained. The RTE needs to ensure that the capabilities are employed to drive profitable outcome.
Further, capabilities, technical and managerial work well only when culture and collaboration provides the support. This seems obvious in theory, but forgotten in practice. Dr Forsgrens’ work substantiates this with data and hard evidence collected over four years.
In fact, the study establishes capability centric models deliver better than maturity models. The latter is built of subjectivity, while the former provides direct line of sigh with DevOps outcomes (deployment frequency, mean time to recover, for example).
Final word — RTE and 3Cs
Synergy between the 3Cs drive customer critical outcomes. As a servant leader the RTE drives that synergy and hence is the heart of any agile transformation. The soul of servant leadership is to identify the change needed (in team capabilities) for the team to do better and makes the journey towards change painless. Transformation initiatives often include creating culture that promotes collaboration. As a change catalyst, RTE engages diverse stakeholders proactively to improve quality of decisions, speed and [adapted from here].
RTE needs to facilitate the change taking a systemic view. If a local change needs cutlure calibration at leadership level, RTE needs to craft the path towards that calibration. This requires a systemic view (to understand organizational impact and economic thinking (everything has cost, benefit and risk)
The RTE needs to be the compass directing the team always to the north star, channel team bandwidth and capability towards right challenges. Program increments being time boxed, can cause team stress as time chips away. By developing a culture of “mindfullness” about the 3Cs, RTE can create a shared mental model amonsgt the stakeholders.