The evolution of organisational structures
In last year’s trend report, I wrote about the emergence of Internal Developer Platforms (IDPs) that enable application teams and reduce the cognitive load for developers within organisations. In this past year, we have seen the industry embrace this approach and many more organisations continue to do so. We learned that treating the platform-as-a-product and focusing on the end user is best suited to build these modern platforms.
What we also observe is several organisations making the technological leap without having some necessary organisational structures in place. Any attempt to make an organisation-wide change should also go hand in hand with considerations about how teams are organised and interact.
Team Topologies offers a set of principles and methods that enable organisations to establish effective collaboration, autonomy, delivery focus, and product alignment while structuring and evolving their teams. Let us look at why the Team Topologies framework is an emerging trend.
Dysfunctional platform teams
Every IT organisation is a collection of teams of people. Practitioner insights (and personal anecdotal experiences) over decades of IT show us that most delivery issues ultimately stem back to one question: How are the teams organised?
Organisational structure is key to success. Technology comes later. Many organisations will require a platform team or multiple such teams to integrate commonly needed services and tools in support of their business teams.
We often find platform teams as large as entire departments, composed of several subgroups of members who are focused on different parts of the problem. Often, the dysfunction arises here when the platform team is not focused on enabling the business outcomes, but just on providing a technology-specific service, such as a database team or a tooling team.
Also, the large size of these ‘teams’ tends to dilute the focus away from business goals and decision-making is often an ordeal. Ideally, you want a small but not too small team—a “Two Pizza” team, for example—that can stay nimble and focused on delivering business value. Interactions with the rest of the organisation also suffer when you have large technology-focused teams. A healthy collaboration between teams is required to maintain a steady flow of software.
Maturing organisational models
Any organisation that designs a system will produce a design whose structure is a copy of the organisation’s communication structure.
This is Conway’s law. It observes that the structure of a software system or architecture tends to mirror the organisational structure of the team(s) responsible for its development. Modern organisational approaches like Team Topologies are inspired heavily by Conway’s law. These approaches are inspired by the now commonly accepted Agile software methods and take a step further in assisting teams in building modern platforms. Team Topologies is a tool to evolve a complex organisation into a simpler, more responsive one. Responsiveness, decoupling, and alignment to business needs is key, almost akin to breaking up the monolith in architectural terms.
Team topologies in short
Team topologies are a way to organise the enterprise in a manner that it can deliver value efficiently by teams adopting modern software and platform architectures. It is all about creating structure, reducing cognitive load, and enabling collaboration among teams.
It does this by addressing two key topics: Team types and team interactions.
Team types
Identify your current teams and map them to one of the following different team types.
Stream-aligned teams
These are long-term, business-goal-oriented teams that develop and maintain a specific business service. Stream-aligned teams make use of internal infrastructure platforms but are focused on building business services.
Platform teams
One (or more in larger organisations) team is responsible for building the platform that provides services and tooling for the stream-aligned teams. They are strongly customer-focused and follow a product-centric approach for building the internal platform.
Enabling teams
Teams of experts who help reduce cognitive load and bridge knowledge gaps of stream-aligned teams. These are often short-term teams, and do not own any software components.
Complicated subsystem teams
Groups of experts dealing with aspects of services spanning several stream-aligned teams, in order to address a specific, often complicated problem.
Team interaction models
Team Topologies suggests the following interaction models between teams.
Collaboration
Two teams working closely together for a limited amount of time for discovery purposes or to bridge knowledge gaps.
X-as-a-service
The natural end state of interaction between teams after the collaboration phase is over. Each team provides a clear service to other teams, with good customer experience as a priority.
The Team Topologies book with the same name provides a valuable introduction to applying the Team Topologies model in constructing and refining an organisational model.
Practical considerations
We know that getting rid of hand-offs and silos is key to enabling flow within the organisation. In my opinion, change starts with recognition at the management level. Well-organised teams will deliver well-architected platforms at a better pace and with reduced coordination.
Management intent and action are the first steps towards this transformation. With management setting the direction, execution can begin piecing the puzzle of team restructuring. Start with the primary focus on making stream-aligned teams. Create small, long-term teams for maximum impact, almost like an organisational mirror of a microservices architecture. Define team interaction contracts that remove friction.
Often these changes are challenging, especially within large organisations where people and teams are already struggling with constant firefighting. A practical approach here is to apply the Inverse Conway Maneuver. This approach aims to align the organisation’s communication structure with the software architecture being developed, to match the one desired by the organisation. When combined with other Agile practices such as Scrum and OKRs, it can be a potent way to build highly effective teams.
Outlook
A digital platform is not only the collection of APIs, services, and tooling, but also the knowledge and support it provides to its users. Industry analysts at Gartner recognised the emergence of modern Platform engineering as no mere hype. We will continue to see the growing adoption of modern organisational modelling techniques like Team Topologies across the industry.
Technology might be an enabler, but it is the people and teams that are going to drive success. What matters now is how quickly the organisation can adapt to change. We could possibly be at the cusp of rapid technological changes, fuelled by the advent of modern AI. How resilient is the organisation to such disruption? This is where practices like Team Topologies may be best put to use, not as a prescriptive framework, but one of gradual evolution. Read more on this topic from this success story.