Saving £millions by transitioning to a mature live service

The challenge

A 15% budget cut forced the leadership team to consider an alternative operating model.  

 

A series of long-standing challenges plaguing the digital service deemed it financially and operationally unsustainable.  

​

• Heavy reliance on expensive, day rate contractors to support and maintain the service (80% of teams, compared to 20% permanent staff) 

• Lack of senior management visibility and grip on what teams were doing and why 

• Excessive time to onboard new funding lines and make changes to existing ones  

• Internal DfE users employing ‘offline workarounds’ due to unmet needs 

• Endlessly adding and improving product features with no agreed ‘end game’ 

​​

Pebble co-founder, Stephen Blackmore, was commissioned to design a leaner, cheaper digital service, without compromising the user experience. 

​

Specifically, a ‘run state’ operating model was to be explored. Also known as a ‘steady state’, the run state is equitable to GDS’s ‘live service’ – no longer in a never-ending beta, just a mature digital service meeting user needs through repeatable, predictable business as usual operations.  

​

The response

The design work was structured around five main themes: 

1. What could a run state look like, structurally and operationally? 

2. What are the benefits and indicative cost savings? 

3. How far away is the existing operating model away from a run state?  

4. What are the core constraints or blockers from achieving a run state?  

5. What needs to be done to move towards a run state? 

 

To address each one, it was crucial to get a solid understanding of what existing product teams were doing, why they were working that way, and the supporting tooling and technical architecture.  

​

At its simplest, the assignment was structured around a standard research-analyse-options cycle. However, the complex nature of the digital service and the wider funding landscape demanded a deep and comprehensive examination to ensure design options were both credible and rooted in reality. 

​​

Key supporting activities included: 

• A maturity assessment of each product against 18 run state characteristics. 

• An objective analysis of existing operational constraints limiting or restricting the service from operating in a cost efficient, lean manner. Constraints were either: 

  • Service variations – differences in how grants and funding lines are designed, delivered, changed and assured 

  • Working practices – including delivery methodologies, processes, governance, team structures and communication 

  • Technical architecture – including tools, products, software, networking, security components, data flows and integration points 

• Creating a BAU (business as usual) framework to identify essential activity to support a run state 

• Mapping core capabilities needed support BAU activity to help sketch out future roles and resource requirements 

​​

The outcome

A two stage approach provided a clear path to achieving a financial and operational sustainable state without compromising user experience or introducing additional delivery risk. The primary differentiator between the two stages being implementation timescales and the scale of ambition. 

​

Target run state: 12-18 months.  

• 38% reduction in total team size (73 to 45) 

• Decreasing reliance on managed services (70-50%) 

• Removing all testing staff by switching to test driven development 

• Creating a single, shared platform team to replace individual product-team capabilities (DevOps, tech architects) 

• Stopping all non-essential development work 

• Moving to more sustainable working practices 

• Designing out unnecessary service variations 

• Shifting to data driven architecture

​​

Stretch run state: 3-5 years 

• 64% reduction in total team size (73 to 26) 

• Goal to move to 80-100% permanent staff 

• The existing product landscape may change significantly, with some products radically redesigned or even removed 

• Move from a product-based to a capability-based model  

• Focus shifts from maintaining individual products to sustaining the core capabilities of the Funding Service - data collection, calculation, contracting, payment and communication 

• A small, multi-disciplinary team maintains and improves the underlying platform and capabilities. 

• Secured pilot reference customer  

​

Estimated savings from the target run state are expected to be in excess of £3 million per year. This far exceeds the initial 15% budget reduction. The additional savings will be invested into developing the proposed stretch run state.

.