Why Spotify as a self-management solution failed for 45 people workforce for misunderstood autonomy?
Spotify is a structural and scaling governance solution which groups workforce as squads, tribes, chapters, guilds, trios and alliance. Each network of workforce members or each group separates from the other group in its collective goal and mission statements. Each squad network, independently or in a network is autonomous, self-organizing and self-managing. Besides having a mission a accomplish, each network can choose the agile methodology/framework that works best for them. Providing a matrix structure to the organization, the model easily captures the structural and short-term project needs; implicitly. As an example, scaling to social systems, someone may be working as a leader in a system/organization, but is also a member of communities, forums and groups that share common interests; for there is always a need to communicate, share and learn from people with common interests/shared contexts.
The model was introduced as an experiment for a 45 people remote workforce.
Starting with team introduction to a tailored version of the model, advantages and how it was expected to bridge governance and operational alignment challenges in the organization at the time of its implementation; it was relatively easy to restructure the organization into squads merging it with then current governance and management maturity comprising of two account/engagement/project managers.
Each squad was functionally equipped technically by quality assurance, devops and architectural leads for in-circle needs; needs arising from a collection of projects/accounts.
Workforce members could be members of squads and also chapters/guilds (such as quality assurance) forming common interest groups. Communication canvas for each chapter allowed to collectively look into domain specific challenges, collectively defining programs and design solutions to allow for functional capability enhancement.
The model picked up well for it aligned to organizations cultural and scaling needs, but got blocked with leadership style of one of the squad lead, fixed mindsets of the organizational leadership and reluctance to appreciate program management as a separately dedicated value stream.
The model does sometimes lead to an assumed autonomy with a squad, ignoring the connected-ness with the rest of the organization. Another symptom led to ignoring organizational priorities and systemic needs misunderstanding autonomy to comprise with the need for alignment; the squad eventually mission-ed as a silo-ed mini company within the company.
Squad leadership was misunderstood to build a castle around the teams blocking information flux (learning/challenges) across squad boundaries. There was in general a reluctance to share challenges and seek support from other squads or the organizational support team; leading to unhealthy competition between the leads themselves. Considerable resistance from the squad lead to any suggestions or ideas to work towards inter-and intra-squad engagement blocked an effective implementation of the model.
Team happiness and team engagement scores considerably reduced across the organization, with felt and reported disconnects by the workforce members. Few extreme reports also pointed out squad leadership blocking and eventually guiding and filtering on any information moving out of the squad. Workforce members felt caged.
The model was practiced for six months and was eventually called off by the organizational CEO. Besides above challenges, communication and timely processes across chapters was seen as an operational overhead for efforts there could not be mapped to a project context and hence mapped to financial numerics. Though programs and best practices was seen as important and needed; organizational leadership was in general reluctant to provide for a planned budget for program management. The efforts was expected to be weaved in the delivery fabric and within the 8 hour productivity model!!.