Save the Children's journey to Agile delivery teams

Over a number of years, Save the Children UK started a significant transformation process. The organisation needed to mirror externally how they wanted to operate internally – as a networked, ecosystem-based organisation rather than a traditional triangle hierarchy. The organisation has made significant progress on this very long journey, but there is a lot more to do.

Watch Gemma Sherrington, current CEO at Refuge and former Interim CEO at Save The Children explain the transformation journey Save The Children went on.

The challenge

The organisation was facing several challenges that prompted the need for change:

1. External complexity: The socio-political environment had become more complex, requiring a more adaptive organisation.

2. Shift from colonial approaches: There was recognition that imposing solutions on communities was outdated; they needed to become an empowering organisation in service of communities.

3. Need for collective impact: The realisation that impact is not delivered by one organisation alone, but through partnership with other social impact organisations, communities, governments, and corporate partners.

4. Outdated operating model: The existing organisational structure was designed for the industrial age and was not suited to the current environment where work, funding, and operating contexts had changed significantly.
   
   

The transformation process

Initial steps: top-down but disconnected

The transformation began with recognition at the executive team level that the organisation needed to become more adaptable. The CEO at the time initiated an "away day on Agile" for the executive team. Each executive was then tasked with embedding agile practices in their departments.

However, implementation was uneven. As Gemma described: "What happened is it didn't get embedded [widely initially], but I went, because I love change, .. 'I'm gonna just disrupt the whole of fundraising and marketing and reorganise it completely behind Agile.' And yeah, just leapt into the complete deep end in a really enthusiastic way."

Key mistake: off-the-shelf models

A significant early mistake was taking an off-the-shelf agile model without adapting it to their specific context:

"We took too much off the shelf. We did actually take a lot of classic agile design and process and just thought because it's agile, it'll make us more agile. Copy, paste, problem solved. Everyone's talking about this agile thing... if we do that, it'll be really successful."

This approach was described as "no regrets" as it taught them valuable lessons, but it didn't work well initially. They learned they needed to "meet your organisation where it is" rather than making too big a leap all at once.

Strategic pivot: treating the failure as a hypothesis test

Rather than viewing the initial implementation as wasted effort, the leadership team framed it as valuable learning that would inform their next strategic decision. As Gemma described: "It's no regrets, right? It wasn't that it didn't work as well as we wanted it to, so it wasn't like it was a complete fail, but it really taught us a lot about what we needed to do to go to the next place."

This "no regrets" framing is significant. The initial agile implementation effective became a hypothesis that was tested: "If we implement agile processes, we'll become more agile." When the results didn't match expectations, they reviewed what went wrong rather than abandoning the direction entirely.

The review identified a fundamental flaw in their assumptions - that process change alone would drive transformation. This learning directly drove their next strategic decision: pivot from process-first to culture-first.

Shift to success: culture first

The breakthrough came directly from reviewing why the initial approach hadn't worked as expected. The learning was clear: implementing agile processes without addressing mindset was ineffective. This insight drove a fundamental strategic pivot:

"Agile processes can be awesome, but done without proper investment in realising an agile mindset, it is just another bureaucracy in your organisation. Once I clocked that we've got to start with culture [..] that started three years ago."

This culture-first approach focused on:

1. Human-centric design and customer centricity

2. Iterative, test-and-learn methodology

3. Transparency and open collaboration

Skills and capabilities transformation

The organisation identified specific "superpowers" they needed to succeed in this new model:

Key skills and mindsets

1. Audience centricity: Putting the needs of customers/beneficiaries first, rather than imposing expertise.

2. Curious mindset: Being open to learning and questioning assumptions.

3. Systems thinking: Understanding how the organisation fits within broader ecosystems.

4. Growth mindset: Willingness to try new things and learn from mistakes.

The biggest challenge was shifting from an expertise-centered approach to a customer-centered approach:

"If your whole value and everything you've been valued for is your expertise, like you're the expert, there's a journey you have to go on that is like, 'Yes, I have expertise, but I have to place the needs of the customer first.' And we really struggled with people not just imposing expertise.”

Recruiting and developing these skills

Finding these capabilities required:

1. Clearly identifying desired skills: "Once you've identified them and tried to surface 'this is what we're looking for'... they can be found and grown."

2. Focusing on mindset in recruitment: "The biggest thing that I've seen time and time again is about people with a growth mindset that are willing to try and learn new things... every time we've had that more fixed mindset, it's very difficult."

3. Interview techniques: Asking candidates about examples of things that have or haven't gone well and what they've learned from them; asking "how do you like to learn?" to assess genuine investment in learning.

Structural changes to support these skills

The organisation created new team structures:

1. Home teams: Groups of people sharing a similar skill set or discipline.

2. Outcome objective teams: Multidisciplinary teams focused on specific outcomes.

Some of these teams were permanent and long-term, while others would "pop up and down" as needed. This created flexibility while maintaining professional development within disciplines.

Results and Outcomes

The new approach led to several benefits:

1. Better customer understanding: Teams organised around geographical spaces or audience needs rather than internal functions.

2. More adaptive responses: The ability to form multidisciplinary teams to meet specific community needs.

3. Clearer prioritisation: "It does require really robust decision making around prioritisation in the organisation. You have to be able to say, 'look, these three things are going to be the priority.'"

4. Improved collaboration: "Where it works really well is when we are really clear on a shared purpose and outcome. And really clear on who needs to be involved."

5. Better engagement: Staff surveys showed improvement, with people embracing new ways of working and feeling less frustrated by bureaucratic processes.

Why this matters

This case study illustrates two interconnected lessons for charity leaders. First, that culture change must precede process change - bolting new ways of working onto old mindsets creates "just another bureaucracy."

But second, treating strategic decisions as opportunities to learn from what doesn't work can drive more effective transformation than attempting to get everything right first time.

Save the Children's initial agile implementation wasn't a failure to be buried; it generated the insight needed to make the next decision better. The learning chain was clear: try something → review honestly → identify root cause → pivot strategy (iterate) → embed test-and-learn as ongoing practice.

Too many organisations treat unsuccessful initiatives as embarrassments to move past quickly. Save the Children's case shows the alternative: frame them as valuable data, extract the learning explicitly, and let that learning drive what comes next.

This case study was developed based on an interview by Carmen Barlow with Gemma Sherrington, current CEO of Refuge and previous interim CEO of Save The Children. Claude.ai supported in the synthesis and writing of the case study.