< Insights

Unified vision and the paradox of autonomy

  • Uncategorized

Studies show that three pursues are necessary for a high-performance team: autonomy, mastery, and purpose. And, when we have a small number of teams, optimizing an individual team’s performance is enough to optimize the overall result.

However, when we have many teams, it’s no longer enough to optimize individual teams. Now, managers will need to balance a large number of current demands with the capacity of their teams. But, if each team has the autonomy to define its process and monitoring metrics, how will the manager compare them to optimize the overall result? And how does one manage interdependencies between teams?

When scaling teams, the core values for achieving high individual performance conflict with a high overall performance achievement. We call this situation “the autonomy paradox.”

Defining capacity and demand

Suppose the manager’s job is to balance capacity with the demand of teams with different processes and metrics. In that case, the question naturally arises: do we get metrics for capacity and demand regardless of the differences between teams?

Our proposal to standardize measuring the budget and time of demand has a low impact on the autonomy of teams and has great value for global optimization.

Whether using story points, function points, estimating in hours, or T-shirt sizing, if all teams use the same criteria to assess the size of demand, we solve this part of the equation.

Typically, the demands are very large. Breaking it down into functionality to be delivered is a natural process to increase the accuracy of demand size estimates. It’s something we commonly do.

With demands having a unified size metric, each team’s capacity can be directly measured by looking at how much time and cost it takes to complete a feature, regardless of the process they use.

Unified view of the portfolio

With a unified capacity and demand, you can build dashboards of all your teams and demands to compare them and manage your portfolio. How did the team’s ability change over time? How did the demands vary over time?

Is there a relationship between the capacity and composition of team members? Is there a relationship between customer satisfaction, profitability, capacity, or demand? And so on.

You can build predictability models to estimate when demands will be completed and include this information in your dashboards to create a unified portfolio view from these same metrics.

And with a unified predictability model for all current demands and teams, you can simulate decision scenarios to optimize your portfolio or solve specific initiative issues by reallocating your resources. What happens if I increase the capacity of this team? What if I decrease the number of people? What is the impact if I grow? What if I vary the demand?

Creating and applying scenarios gives the unified view the ability to tackle the problem of the paradox of autonomy, giving more power to the manager so he can optimize the portfolio.

Unified view of activities              

Another tool that we have found to be of great value to a portfolio manager is looking at all their portfolio activities in a unified kanban framework.

But, how does one do this if each team has the autonomy to define its flow? Well, generally speaking, all streams manage to “fit” into a simple framework. Regardless of the specific process of each team, there are the mandatory stages: “To Do,” “Doing,” and “Done.”

Here is a picture with several different workflows, and indicating the 3 phases: To-do, Doing and Done of all of them

In practice, what ends up happening is that teams use autonomy to define their processes. Still, organizational culture also influences, and many end up having standard stages or checkpoints. For example, at Objective, in addition to automated testing, we always have a Quality Assurance stage to catch issues that are difficult to detect during the development time. And it’s always after the development stage.

So finding the stages common to all teams, and mapping them, gives the manager a unified picture of all activities. And it allows for them to detect and correct bottlenecks that transcend a single project. In our case, we could see bottlenecks in QA, which were hard to see looking at just one project, but it became apparent when we unified them all!

And the unified view of projects can be used as a step towards unified flow management, a disruptive way of scaling development teams that by definition eliminates the autonomy paradox.

Attacking the autonomy paradox with Sinccera

Here we show how the above concepts were applied to Sinccera, generating unified views that effectively attack the autonomy paradox.

Demand and capacity at Sinccera

At Sinccera, we use T-Shirt Sizing to measure the budget and deadline of the demands. In short, we classify the demands in terms of XS, S, M, L, and XL. From the perspective of multiple teams, two features strengthened the use of T-Shirt Sizing in Sinccera:

1.     Easy to learn – we had no difficulty in training teams to use T-Shirt sizing.

2.     Uniformity – different teams rank demand with low deviation.

 Two reasons for this: 

  1. It’s not a measure directly linked to effort or deadlines.
  2. It’s a relative measure focused on comparing items evaluated (this one is larger than that but is smaller than this one) and not in absolute estimates (this story costs an x amount of points).

Other reasons reinforced the use of this methodology, which we explore in more depth in our predictability article.

In summary, on the one hand, we have the total size of the scope to be executed (backlog + doing) calculated with the t-shirt size of the demands and, on the other hand, the capacity defined by the team’s delivery history using the same metrics. And it is on these two metrics of capacity and demand that Sinccera’s predictability models are built.

Combining all of the above information, Sinccera provides a series of unified views of demands.

Unified view of the portfolio at Sinccera

At Sinccera, you group demands into strategic objectives, and we define our budget and deadline expectations. And, from Sinccera’s predictive models, you can see which goals are more likely to meet or break expectations.

The strategic dashboard organizes this information, summarizing each objective and highlighting the most significant planning deviation. This vision is the first step for you to balance capacity with demand, optimizing portfolio delivery. It increases your effectiveness by focusing on strategic objectives with the most significant risk of breaking assumed expectations.

Diving into the details of a strategic goal, the visions are uniform because they use a standard definition of demand and capacity. This standardization increases your efficiency by analyzing each of the goals. It also ensures greater alignment with the individual managers of each project, now expanding the team’s efficiency.

And, from the predictability models, you can simulate different scenarios to solve specific objectives problems. You can design hypotheses such as renegotiating scope and deadlines, reallocating resources, and realigning customer expectations in a scientific, data-driven way.

You define the simulation scenario as the new plan by realigning with customers, ensuring alignment with everyone involved in the strategic objectives.

Unified view of activities at Sinccera

You can also view all your activities on a unified kanban board on Sinccera from our mapping of Jira issue statuses with Sinccera stages. Here is the vision:

A particularity of Sinccera: in addition to grouping demands into strategic objectives, we break deliveries into two other levels: features and activities. Simply put, demand delivers value to the business, functionality provides value to the software, and activity is what we are doing on a day-to-day basis.

Now, the developer can see how his work contributes to the software, the value generated for the company, and at what strategic objective he is acting. As mentioned before, high-performance teams need autonomy, mastery, and purpose, and this break helps to align the purpose.

Conclusion

By unifying the way to measure the size of the demand, you sacrifice a small part of the autonomy of the teams, and in exchange, gain a unified view of the status of all demands, opening the possibility of creating predictive models and simulation of scenarios.

The team still has autonomy over its process to enable the individual optimization of its results while creating metrics and visions that allow acting in the global optimization of the portfolio.

And we can see on Sinccera these concepts translated into functionalities, based on an opinion on how to measure the budget and deadline of demand with T-Shirt Sizing.

Where to go now? Well, the Predictability article explains in more detail the choice of T-Shirt Sizing. Also, it shows you how to create a model that increases its accuracy when working with historical data. And the Unified Flow article discusses how this methodology significantly increases your teams’ total capacity and adaptability to change, tackling the problems of the autonomy paradox from another perspective.

Insights from our team

Get insights from our team of experts on software development methodologies, languages, technology and more to support your team in business operation and strategy.