In brief
- Engagement with IT vendors in a managed delivery model enhances delivery ownership and risk management, offering benefits like freeing up in-house expertise, efficiency gains, predictability, and cost optimization.
- There are three main managed delivery engagement models: autonomous squads, application ownership, and portfolio management, each suited to different needs and risks.
- Clients can choose the most suitable model based on their objectives and risk tolerance.
An organization’s engagement with their IT vendors in a managed delivery model increases delivery ownership and risk levels. The client benefits from things like the freeing-up of in-house expertise for strategic initiatives, as well as increased efficiency, predictability and cost optimization.
There are three main types (models) of managed delivery engagement:
1. Autonomous squads
- Limited features and functionality within a single or small group of applications
- Delineated roles, possibly including specific onshore, nearshore or offshore locations
- Development or support (potentially) in scope
2. Application ownership
- Logical group of applications with corresponding business functionality and all-encompassing roles
- Includes all the functional, technical, process and project management skills needed to manage the application group
- Development or support (potentially) in scope
3. Portfolio management
- Complete portfolio of applications for end-to-end business functionality, products or services
- Includes all the relevant functional, technical, process, project, program and portfolio management skills
- Development and/or support in scope
Together, we’re going to tackle these key models in detail, examining the different levels of benefit, characteristics and selection criteria, as well as weighing their pros and cons.
Main building blocks of the different models of a managed delivery engagement
Team composition
Ways of working
Delivery performance model
Pricing models
Knowledge management framework
Innovation and transformation model
Choosing the right model
Autonomous squads managed delivery
For clients:
- Experimenting with a managed delivery model prior to sharing a significant scope of work with vendors
- Looking to form centers of excellence for specific skills and roles (provides a reliable supply of niche skills)
- Aiming to introduce a managed delivery approach in small scale applications or systems
Application ownership managed delivery
For clients
- Transferring responsibility and risks for a subset of business functionality to vendor, and refocusing corresponding in-house expertise on more strategic business functions
- Driving efficiencies in a specific group of applications representing a logical business scope
- Iteratively developing a specific set of solutions from a larger portfolio of features
- Setting up an independent maintenance and support team for a logical group of independent applications
Portfolio management managed delivery
For clients:
- Transferring responsibility and risks for an entire portfolio of applications, products or services and refocusing corresponding in-house expertise on other strategic platforms
- Driving portfolio-wide transformation and efficiencies for multiple business clusters within a vertical, or shared services across verticals
- Iteratively developing solutions and products for a business portfolio
- Embarking on vendor-consolidation initiatives
- Setting up an independent maintenance and support team for a complete business portfolio
Other managed delivery models
You might come across other models too:
- A hybrid of the three models involving customized roles and responsibilities, segregation between client and vendor, unique benefits models and a combination of services
- Cross-portfolio managed-services models for a specific catalogue of services like: Level 1 support, Quality Assurance as-a-Service, Program Management as-a-Service, DevOps as-a-Service and so on
In summary
Benefits vary across the models, so the client simply chooses the best fit.
Pros and cons
Autonomous squads
Good:
- Excellent model for experimenting with organizational change during the adoption of small-scale managed delivery
- Minimal organizational changes and knowledge risks
- Complete control over ways of working
Less than ideal:
- Limited cost and efficiency benefits
- Greater share of roles and responsibilities requires more governance effort
- Increased complexity from aligning in-house and cross-vendor teams
- Not much scope for innovative pricing models
- No transformation
Application ownership
Good:
- Flexible pricing provides greater benefits
- Enhanced benefits from efficiencies
- Less governance effort, more vendor accountability
- Increased control over ways of working
- Tighter management of key business, technology and process knowledge retained in-house at portfolio level
Less than ideal:
- Greater (but manageable) client-change effort
- Vendor’s managed-delivery-ownership benefits might be reduced by architecture plus key SME and portfolio management responsibilities being retained in-house
- More complex because of the need to align in-house and cross-vendor teams
- Limited scope for innovation and transformation because the vendor owns delivery for only a subset of the portfolio
Portfolio management
Good:
- Flexible pricing provides significant benefits
- Benefits from efficiencies
- Minimal governance effort with extensive vendor accountability
- Greatly reduced complexity from aligned portfolio-level ownership and scope with vendor
- Portfolio-level innovation and transformation
Less than ideal:
- Minimal control on ways of working
- Largely depends on the vendor for knowledge
- Major client-change effort which most likely needs third-party management
In summary, the client needs to choose the managed delivery model that’s the best fit, or a hybrid of models based on their risk tolerance, past experience, benefits and transformation requirements, level of comfort with vendors, client readiness or similar. Luxoft has partnered extensively with many clients, advising on the selection of engagement models and, subsequently, implementing those models.
