Building high-performing teams through continuous improvement

At Intercom, we’ve always focused on building excellent products which our customers value and love. Continuing to meet this goal as we scale means constantly re-evaluating how our teams work to keep performing at our very best.

There is no “one size fits all” approach to building high-performing teams. Once you find a winning formula, it doesn’t stay relevant for long. We’ve developed a way to measure and address the challenges that prevent our R&D teams from performing at their best; we call it the High-Performing Teams (HPT) framework. We’re aiming for excellence at scale, and this framework will help us to continuously improve the way we work as our company grows.

Why should we focus on how teams work?

A high-performing company is made up of many high-performing teams. If one team’s performance starts to slip and isn’t quickly corrected, other teams’ performance will be affected – with company-wide implications. And it isn’t just companies that want high-performing teams, it’s individual team members.

But how do you identify when a team’s performance is slipping? What does a tail-off in performance look like? What is the difference between the natural ebbs and flows of performance and a more serious decline? And once a performance problem is identified, how do you go about fixing it?

“Teams need to continuously evolve, update strategies, and integrate new team members and leaders to maintain high levels of performance”

A team that isn’t growing or adapting is stagnating. Teams need to continuously evolve, update strategies, and integrate new team members and leaders to maintain high levels of performance. We wanted to create a framework that could evolve with our teams; adapting to the way they work now, and how they might work five years from now.

We’re also facing a future of remote or hybrid work that plays into every decision we make. To continue to strategically grow in the way we want, we need to consistently strive for high performance across teams.

Setting our goals

At Intercom, our product principles are: “start with the problem”, “think big, start small, learn fast”, and “deliver outcomes”. Any framework we chose or built had to incorporate this approach to problems and solutions. Our priorities were:

• Alignment: We’re in a phase of major growth, and we have big ambitions for the future. To achieve our goals, our teams need to agree on what high performance means to them, and pull in the same direction to achieve it. That will help us focus on aligning team processes and improving collaboration.
• Continuous improvement: The framework had to aim for ongoing reflection and revision rather than a one-off fix. Continuous improvement is embedded in Intercom’s culture, and essential to maintaining high performance within a team as it grows.
• Shared learning: We want to make it easier for teams to share their problems, approaches, and solutions. For that to work, we need universal metrics and parameters so that information can be clearly understood by other teams.

Implementing the HPT framework

We first implemented the High-Performing Teams (HPT) framework in early 2020 to focus on team alignment, continuous improvement of processes, and collaboration. HPT facilitates in-depth quarterly reflections, using an agreed set of metrics to represent a shared understanding of high team performance, as well as supporting structure and process.

How does it work?

HPT revolves around four steps, each one mapped to a period within the quarter. We assigned specific timeframes to each stage to keep us to the desired timeline, and understand where we might need more time or resources. The four steps of the HPT framework are:

• Measure: Take your measurements to identify the strengths and challenges that exist for the team. Intercom’s R&D teams use seven metrics: purpose, impact, composition, focus, collaboration, trust and growth, and decision making. Using the same metrics each time deepens understanding and focuses discussions within the team.
• Act: Get the team together and review the measurements. Decide on actions your team will take to improve the problem areas, and work on the most urgent first.
• Review: Towards the end of the specified period, close out any open actions. Review the closed actions to see what worked and what didn’t.
• Share: Pick out the actions that worked – these are your reusable plays for next quarter. Add them to the shared team playbook to create a global resource and share your learnings with other teams. In subsequent quarters, teams will be able to find relevant plays for the metrics they want to focus on.

“HPT facilitates in-depth quarterly reflections, using an agreed set of metrics to represent a shared understanding of high team performance, as well as supporting structure and process”

From version 0.5 to version 1

When we first implemented the framework in early 2020, the sudden move to remote working meant we were dealing with a fresh set of circumstances. The first iteration of HPT wasn’t implemented as effectively as we’d hoped, and highlighted some problems we’ve tried to address in the subsequent version. However, there were enough positive indicators to encourage us to keep trying to get it right.

Support Content’s HPT experience

Intercom’s Support Content team implemented the HPT framework to pinpoint the most important team improvements they could make. One of the top problems initially reported was feedback; both frequency of feedback received by team members, and lack of space to provide feedback to teammates.

This prompted the team to introduce always-on feedback through our feedback tool. Team members were given the option to request and receive feedback every six weeks. Each team member customized feedback requests to reflect a current project or task they were working on to make feedback as relevant as possible.

After six months, feedback had gone from being the team’s biggest issue to an essential part of how they worked. The statement “I’ve been given useful feedback from my teammates in the past” was rated third-highest by the end of Q4. The team score for the trust metric also dramatically improved.

Refreshing the HPT framework

There were still a lot of areas we needed to work on. For example, HPT was initially engineering-biased, and we needed a more cross-disciplinary approach to better incorporate the perspectives of Product and Design.

In late 2020, we revived and refreshed the framework to produce v1. We worked to make the framework more inclusive, usable, and useful for every team member, across all our teams. We added some useful resources and filled glaring gaps – building out an accessible HPT information hub, R&D playbook, and clearer definitions of our metrics. We used the following principles to build HPT into a more valuable framework for our teams.

1. A team-first approach

The framework should work for teams, not the other way around. If teams don’t engage and find value in the framework, the output won’t be useful. We’ve made the framework easier to use by further defining metrics, reducing the steps involved in the process, and establishing clearer timelines.

2. Shared learning and collaboration

We want to make cross-team collaboration easy and obvious, so that anyone on any team can access a tried-and-tested solution to a problem they’re facing. To improve these elements for version 1, we focused on:

• Building out a resource hub for codified challenges and actions
• Connecting individuals at all levels to collaborate on challenges through the framework, even if they weren’t experiencing those challenges at the same time.

3. Better support through transparency

Many challenges can be tackled in-team, especially by leveraging learnings from others. But some require external help – support from group or org leaders. We’re working to make it easier for teams to bubble up common external challenges so that appropriate support can be given.

“We’re aligned on what high performance means to us across R&D, we’ve defined and measured our blockers, and we’ve taken action to eliminate them”

What are some of the ‘plays’ we’ve implemented?

We’re positive about what we’ve achieved so far, and we’re ready to keep improving. We’re aligned on what high performance means to us across R&D, we’ve defined and measured our blockers, and we’ve taken action to eliminate them.

We saw high participation in the “Measure” and “Act” steps, with 52 actions captured and many put into practice in Q1 of this year. Some of those actions have worked, some haven’t. We know we won’t always achieve high performance, and some factors, often external, will continue to affect the way teams work. HPT aims to make it easier for teams to prolong phases of high performance and to escape the lows faster. Here are a few things we’ve implemented so far:

• We’ve committed to share and discuss rationales behind decisions to de-prioritize or drop goals during a cycle.
• We’ve created a more structured and efficient on-call rotation process.
• We’ve introduced an “X in 15 minutes” session every week, where a member of the team presents or explains something to the rest of the team. Topics are mostly, but not exclusively, work-related.
• We’ve organized workshops for sharing personal goals within teams to facilitate greater support of individual growth.

Next steps

We’re keen to maintain the high adoption and engagement with HPT that we’ve seen so far and we’re turning our attention to the impact our actions can have. We’re adding new reusable ‘plays’ weekly based on team suggestions – building out our playbook to accelerate learning, collaboration, and performance across the org.

Applying a HPT approach to the HPT framework

The tenets of HPT also apply to the framework itself. We need to stay focused on the how and why of using the HPT framework within the team. If the team loses sight of why they’re doing this – or becomes frustrated with the process – HPT itself becomes a blocker to high performance. Just as we do with our products, we’ll continue to iterate and make improvements.

As Intercom grows, ensuring consistent quality and impact will increasingly depend on frameworks like this – thinking deeply about how we build is as important as thinking deeply about what we build.