At AgilityHealth®, our team has always believed there’s a relationship between qualitative metrics (defined by maturity) and quantitative metrics (defined by performance or flow) for teams. To prove this, a few years ago, we adjusted our Agile team assessments to gather both qualitative and quantitative data. Once we felt we had
sufficient data to explore through our AgilityHealth platform, we partnered with the University of Nebraska’s Center for Applied Psychological Services to undertake a review. The main question we wanted to answer was: What are the top competencies driving teams to higher performance?
Below are the five quantitative metrics that form the Performance Dimension within the TeamHealth® radar and
the metrics we focused on during our analysis.
To find out what the top drivers for team performance are, we analyzed both quantitative and qualitative data from teams surveyed between November 2018 and April 2021. Our analysis included 146 companies representing a total of 4,616 teams (some who took the assessment more than once) which equates to more than 46,000 individual survey responses.
We used stepwise regression to explore and identify the top drivers for each of the performance metrics. Stepwise regression is one approach in building a model that explains the most predictive set of competencies for the desired outcome. The results of our analysis identified the top drivers for each of the five performance metrics in the TeamHealth assessment, along with the corresponding “weight” of each driver.
Based on our insights, teams that focus on increasing the top drivers should see the highest gain on their performance metrics.
After analyzing the top drivers for each of the performance metrics (Predictability, Value Delivered, Time to Market, Quality, and Response to Change), we noticed that five qualitative practices kept showing up across these metrics. Those are short iterations, generalizing specialists, planning and estimating, creativity and innovation, and self-organization.
But what does this mean for you? Let’s start by focusing on each of the five qualitative drivers and how they help drive quantitative performance metrics.
What defines a “short iteration”? Most guidance defines “short” as a time-box long enough to deliver increments of value, but short enough to pivot and make changes when needed—usually about 2-4 weeks. Some types of teams, like support teams, may use a weekly iteration based on the kind of work they do. But it is more than just the length of the iteration that matters—it is the quality of the planning that goes into an iteration that counts.
When teams have a good understanding of their backlog with complete stories—and by complete, we mean stories that include a Definition of Ready, Acceptance Criteria, and Definition of Done—the teams know what is expected of them. It is much easier to be productive and high-performing when you know what the customer is expecting.
It is also important to size the stories correctly. We all know the pressure we feel when there is too much work and not enough time to get it done, right? This is what happens to teams when stories aren’t sized the right way, or when the backlog is filled to capacity. Being realistic about how long it takes to get work done, and about how much work can get done, in an iteration, is crucial.
Leaving capacity for unplanned work is another hallmark of high performance. We all know that there is some level of “unknown” in our work—someone on the team calls out sick (reducing capacity), something takes longer than expected (increasing the size of that piece of work), or something unexpected happens (requiring more work than expected). A good rule of thumb is to leave 20% of expected capacity open for unplanned work to allow for these inevitables.
Utilizing Short Iterations has a high impact on the team's ability to be predictable. When a team understands how to work well in shorter increments they can better predict how much work they can get done within the timebox they have, and therefore are more likely to get the work done they have committed to completing.
What exactly is a Generalizing Specialist? These are team members who are deeply specialized or skilled in a specific area but acquire skills in another area through cross-training, allowing them to help other team members when needed. This enables teams to become cross-disciplined and cross-functional so they can more easily produce end-to-end value and reduce bottlenecks. When teams are T-shaped or full of Generalizing Specialists, team members help each other as needed to meet team timelines and objectives. If this is a goal for your organization, what steps can be taken to enable your teams?
First, make sure team members are open to learning about other roles on the team and new skill sets. Experiment with how they can expand on each other’s knowledge. For example, if there is someone interested in learning how to be a Scrum Master, ask them to lead the Daily Stand Up meetings for a week and help with planning for the next iteration. If a team member is interested in learning about testing, ask them to sit in on the next test planning session to learn what is involved in developing these activities. Another suggestion is to have Developers hold paired coding sessions for those interested in understanding how a system is constructed—the list of ideas is endless.
The superpower behind T-shaped teams is their ability to “swarm” when something needs immediate attention. In a cross-trained team if something urgent happens and there is an overabundance of testing to get code out the door, those that don’t have testing as a specialty can step in and help to meet the urgent deadline. If a team’s Scrum Master is out sick or on vacation, key ceremonies can still effectively take place with cross-trained team members.
Intentionally creating Generalizing Specialists on teams not only enhances the skills of the individual, but it makes teams stronger by removing single points of failure and sharing the work across multiple people. Quality is measured by Mean Time to Recover, Change Failure Rate, and Defect Ratio, and when a T-shaped team can swarm to handle unexpected circumstances, you won’t see these metrics negatively affected.
We touched on Planning and Estimating a little when we talked about Short Iterations, but let’s dive a little deeper into this topic. In an ideal state, all of the team’s work is tied back to the strategic outcomes of the organization, and this alignment is visible by creating and sharing Outcomes with fully defined Objectives and Key Results. In reality, most of us are not here yet, and that’s okay. Some things can be done in the interim to mature Planning and Estimating for teams while working toward this ideal state for your organization.
It's important to understand there are three levels of planning: Release Planning, Iteration Planning, and Daily Planning. In Release Planning, the focus is on the overall time it takes to complete a feature (from a Scrum technology team perspective) or, if you want to timebox it, a 13-week increment. In Release Planning, the roadmap for the quarter is created: what pieces of the feature can be created and released in what increments to provide value for the customer? An example from Human Resources, a team responsible for the benefits of an organization that uses 2-week iterations, may look something like this:
Week 1: Contact the broker to view benefits options for the coming plan year
Week 3: Receive benefits options from the broker
Week 5: Create comparison document for benefits options to present to stakeholders; get budget information from Finance
Week 7: Review benefits options with stakeholders and include budget information
Week 9: Finalize benefits options with the broker and make selections
Week 11: Distribute benefits information to employees and schedule four informational meetings
Week 13: Finalize benefits sign-ups for all employees
Iteration Planning takes Release Planning to the next level. Now that it is known what value the team wants to deliver in the next two weeks, what are the necessary steps or what stories need to be completed to deliver on this value? Continuing with the example above, we will look at Week 5, where the team wants to create a comparison for stakeholders so a decision can be made on which benefit options are offered to employees.
Day 1: Review information from the broker; ensure all requested information has been received. Request any missing information
Day 2: Request finalized benefits budget from Finance—ask for it to be received by day 5
Day 3: Receive any missing information from the broker
Day 4: Begin building comparison of benefits options
Day 5: Receive benefits budget from Finance; ensure the format will insert into the presentation. If not, add to the backlog for Day 6/7 to reformat
Day 6: Get peer review on comparison document; incorporate feedback into the format
Day 7: Insert budgetary information into the presentation
Day 8: Get peer review on presentation, incorporate feedback and any additional changes
Day 9: Final review and edit of presentation—both for content and grammar
Day 10: Presentation to stakeholders
The final level of planning is Daily Planning, which is done by the team in the Daily Stand Up—what did each person complete yesterday, what do they plan to do today, and what impediments do they have? By doing this level of planning together each day, teams can ensure they are staying on track to achieve their goals, learn where they need to help each other, and team members have the opportunity to ask for help.
Strong skills in the areas of Planning and Estimating have a direct impact on the value delivered by an organization. Value delivered is typically assessed by looking at the throughput of work items or story points, feature throughput, and how well teams are delivering on Business Outcomes. When teams can effectively plan the work being done, you can ensure that the highest value work is being completed by the organization.
For years, the idea was that the “boss” would decide how the work needed to be done and the “workers” would do what they were told. One of the changes that came about with the emergence of Agile was the idea that team members were better equipped to determine the “how” as long as they had a clear understanding of the problem they were trying to solve. We were so excited to see this theory supported by the data!
But what does it mean to allow for Creativity and Innovation within your teams? It can be as simple as allowing the team to decide how they will solve a problem rather than telling them the solution. Many organizations today are incorporating Creativity and Innovation exercises into their work cadences because they have learned that team members come up with great ideas when allowed to play. Some common options your teams might find worth trying are:
Hackathons - these are timeboxed events where teams come together to collaborate, design, build and present solutions to problems or new ideas. This is not restricted to software builds; this has also been done to build courseware for new learning and in supply chain management.
Innovation Games - developed by Luke Hohmann, these games are designed to generate feedback from customers about a product or service and allow the team to learn how their work is being used, or not used, by the customer and what can be done to improve it. These are guided activities that help a team learn how to innovate with the customer in mind.
Mobius Loop - this discovery method as taught by Gabrielle Benefield, helps teams to walk through what is possible by connecting strategy and design thinking. By going through the process associated with these concepts, team members are taught to think through the possibilities from multiple angles as they determine how to achieve outcomes.
The ways for teams to be creative and innovative are numerous, but it is important to share some guidelines, such as timeboxes, which problems to focus on, and if there are potential budgets the team needs to stay within for any proposed solutions. This way, when the team comes up with a solution, it can be implemented or experimented on and everyone finds value in the exercise.
The ability to be creative and innovative can highly impact the team’s performance around the time to market metrics, which focus on deployment frequency and lead time to deployment. This is most likely because allowing them to be part of the solution and have upfront buy-in to how the work will be accomplished will enable them to get work done faster.
Last, but by far not least, is the concept of self-organizing teams. This concept has ties to Creativity and Innovation—allowing teams to define how they will work together within a set of guidelines. Truly self-organizing teams are empowered (and encouraged) to decide how they will deliver on the vision of the Product Owner. If they have an impediment they need help solving or an obstacle they need help removing, the Product Owner is there to help, but only after the team raises their hand and asks for assistance.
One of the hallmarks of Self-Organization is when the team actively manages the work themselves—they are estimating the size of the work, defining their capacity, and keeping the work visible. Oftentimes team members will volunteer innovative ideas as a part of their normal process. The team has an understanding of what’s most important because they have clear acceptance criteria and are not afraid to ask clarifying questions.
When teams are self-organized, the impact is most evident in their ability to respond to change. Because they are adept at designing solutions, managing their work, and understanding what is being asked of them, it is easier for team members to pivot when the world around them changes and they need to do something different than what had been previously planned.
In summary, it is clear that if teams want to impact their performance metrics positively, the focus should be on following Short Iterations, establishing Generalized Specialists, utilizing effective Planning and Estimating, enabling Creativity and Innovation, and mastering Self-Organization. Our research showed us that teams that invested in these drivers saw a 37 percent performance improvement over teams that did not.
It may feel overwhelming to consider bringing these five ideas to your organization but try starting with one change at a time. Allow that change to become part of your process and then make the next change. Finding the best way to improve performance metrics will look different for every team and every organization—it truly only matters that you are implementing change for the better.