Teams often struggle to find new ways to reach the next level of performance. Big projects can help set north stars for what the team’s future can look like. Unfortunately, they’re hard to get buy-in for, hard to move to “done”, and don’t always show how they relate to improving the business itself in a small measurable iteration. Finding a way to improve the team and continue shipping value is challenging and not always obvious.
On the Visibility team at sennder we want to become more performant, but not simply by writing more code. Our mission is to provide our customers with reliable and in-time data for tracking and analyzing journeys to achieve end-to-end visibility of trucks. We could write ten thousand tickets at the start of the quarter and generate a million lines of code by the end of the quarter, and most of that becomes dead weight once we learn something in the plan needs to change. What’s more useful is developing an agile posture (a term we use to describe how agile a team is) for the team, something that lets us re-adapt daily to new business needs that can change with every new fact we learn. Increasing lines of code at quarter-end isn’t going to get us there. We need to measure how easily we move. How well can we chase a moving target? Our team discovered a way to merge our quarterly OKR planning with some detailed research that helps teams improve their agile posture, and this is our how-to guide to help you do it too.
Why are OKRs useful?
OKRs (Objectives and Key Results) continue to get a lot of buzz these days, and with good reason. When used correctly it is more than a fancy to-do list, it is a flexible set of initiatives that roll up into measurable results that drive a company forward. There are many reasons to start using OKRs if you do not already, but they help ensure your work is aligned with the overall company’s goals, as well as demonstrating how to measure the results of individual initiatives your team takes.
Imagine you’re working on a team at sennder, one part of a larger engineering organization that has identified a need to improve overall team efficiency in producing value for the company. Something like this below where the objective is obvious but the way to do it is not.
Isn’t gut-feeling enough?
It’s tempting but dangerous to look at vanity metrics because they’re easy to measure. Most tools offer them for free. E.g. how many lines of code do developers add?; How many tasks did they complete?; How many hours extra did they work?; etc. There’s usually a way to grab all this data, but that data just feeds hunches and biases instead of a proven methodology. Do you really measure what matters most? Are you measuring improved business outcomes or just the volume of code output by the team?
It’s equally tempting to focus on the big technical ticket that looks like a magic cure-all solution. E.g. use a new database, change tests from unit-tests to UI tests, put everything into Kubernetes, use complex design patterns everywhere, change out one async task runner for another, etc. It simply isn’t enough. The reason is, a big tech project is merely aspirational, it is a dream of doing big things. It will not change the environment, culture, skills, and ways-of-working your team has developed and accepted as a normal status-quo. Your team must develop an agile posture, big-technology is just a detail of what you worked on.
Why is organizational culture so important?
The culture at sennder is a huge factor in supporting the successful application of OKRs and measurement of DORA metrics. The Westrum study, ‘A typology of organizational cultures’, outlines how Pathological (“do as you are told”) or bureaucratic (“always follow every rule”) organizations lack the trust and information flow necessary for OKRs to succeed in delivering real-life business gains.
A generative or performance-driven culture allows sennder, with laser-sharp focus, to achieve impactful results. At sennder, we value and encourage the following, all washed down with a healthy dose of fun:
- high levels of cooperation
- trust
- ownership
- experimentation
- honest and open communication
Let’s go on to explore how the right culture and DORA metrics help the organization set realistic objectives and achieve measurable results.
What are DORA metrics?
Years of research by some truly brilliant people (including Nicole Forsgren, Jez Humble, and Gene Kim, and more) have given you the DORA-4 (DevOps Research and Assessment 4 Key Metrics, a.k.a. Accelerate Metrics). The researchers discovered that tracking these 4 metrics were a strong indicator of organizational performance.
- Deployment Frequency: How often an organization successfully releases to production
- Lead Time for Changes: The amount of time it takes a commit to get into production
- Change Failure Rate: The percentage of deployments causing a failure in production
- Time to Restore Service: How long it takes an organization to recover from a failure in production
These four metrics were benchmarked against the industry for several years, and it was discovered that “elite teams are twice as likely to meet or exceed their organizational performance goals.” Elite teams, by DORA standards, deploy multiple times per day, average commits take less than 24 hours to end up in production, teams can restore a system outage in less than an hour, and less than 15% of all changes have serious issues. You might not like all those numbers, but it’s important to remember this benchmark is based on observed performance, not hypothetical reasoning.
The DORA research program establishes a chain of cause-and-effect relationships between delivery performance and the type of culture, process, and technology an organization has. The keystone here is Transformational Leadership, which enables better Software Development Practices, Lean Product Development, and Lean Management. The stronger these foundations are, the more likely it will be that a team has a sense of identity, improved software delivery, job satisfaction, reduced burn-out, as well as a very performance-driven culture. Each topic is deserving of an article focusing on itself. If you wish to dig deeper into what each one means please check out their meanings on DORA’s capability catalog.
Identifying the relationships allows you to debug the issues holding your team back from reaching their best level. Is management not setting a clear and inspirational vision? Is work being done in large monolithic bites? Does your team lack any automated tests? Are you keeping work-in-progress limited, or is everything top priority at all times?
All teams have different shortcomings at different times, but with this tool, you can start to work with your team and evaluate what will make the biggest impact, and what is the easiest project to get implemented first. If you’re a leader, one of the most straightforward, but not always easy, things you can do is to ensure you live the Transformational Leadership concept every day or learn to improve your own status-quo in this area.
There are a few simple ways to find projects for the team to solve. First, evaluate which topics seem to be doing the worst on your team. We use a tool to regularly collect feedback with an employee engagement survey which gives us insights towards job satisfaction, culture, burnout, and identity. For software performance, we can look into metrics provided by our git host, ticket system, and/or third-party reporting tools. An informal discussion with your team, 1-on-1 or in a group, will also provide valuable feedback. We regularly talk about culture and improving job satisfaction in our team retrospectives for instance.
Once you identify the area that is most in need of improvement you can narrow it down into the root-causes and discuss which project to try first. To be successful, your team needs to be ready to constantly update the “ways-of-working” and agree to do something different than last week. Really take the time to investigate if each root-cause is well-supported. Does your manager have techniques to inspire the team? Is your team afraid of working in smaller increments? Is there something that prevents automation tests from being written? What will happen if you focus on getting one task completed instead of starting two new ones? The team’s ways-of-working will need to constantly be adapted each time you want to take one step forward.
Our team’s objective for Q1 2021 was “to build an engaged and efficient team.” The measurable key result here was to become Elite under the DORA-4 definition. As a team, we looked at what we need to improve and found concrete projects that can help us do that. These became the initiatives for our work in Q1 2021. We, therefore, track the 4-Key-Metrics as well as our progress on these initiatives. There is some lag between the two, but as they say, “what gets measured gets done.”
Key Takeaways
- Team performance is not about writing lots of code, it is about how far the business moved forward.
- Having an Agile Posture enables performance by letting your team quickly react to changes.
- DORA can guide you towards building a more performant team.
- Measuring DORA-4 can integrate easily with results into your OKRs in a measurable way, and generate concrete initiatives.
- Adopting this kind of improvement roadmap will be challenging without the culture or leadership needed to support it.
About the authors
Brian Graham, engineering manager, joined sennder in 2020 and cares about coaching his team to help each person grow into their own next level. His experience includes leading the development of transportation/logistics integrations, software design for a major e-commerce framework, and fintech and employee engagement software.
Dan Grindrod, senior agile coach & delivery lead, joined sennder in 2020 and has been part of the agile community since 2006. He’s worked with a diverse range of teams across the globe to help them continuously ship amazing and valuable products. ....