How Save the Children turned a failed implementation into a blueprint for bringing culture and delivery into the same room.

Your organisation has decided to change the way it works. It doesn’t matter whether it’s a new way of running projects, a new planning framework, or a restructured team model. The decision came from the board, or the CEO, and you’re expected to run the implementation. 

But, you’ve run the project reviews, updated the board on progress, and the new reporting format is in use. And yet, six months in, the questions of who gets resources, what is prioritised, and who has authority to change course (if things don’t work as expected) are decisions that are still being made the way they always were. The implementation changed the process, so things look a bit different, but decision-making looks the same as it always did; the only difference now is that it is accompanied by more frequent and louder grumblings from staff.

Save the Children UK was in the middle of a multi-year transformation when this story begins. The shift it needed to make was fundamental: from a hierarchical, expertise-led model built around delivering solutions to communities, to a networked organisation that worked in partnership with them. The external pressures were crushing: a more volatile funding environment, a sector-wide reckoning with colonial approaches to international development, and the recognition that impact at scale required working through others rather than acting alone.

Agile was the vehicle chosen for this shift. It is a way of organising work, borrowed from software development, built around short delivery cycles, cross-functional teams, and the ability to change course between reviews, rather than only at annual planning. The Save the Children CEO launched an executive-level Agile programme, asking each senior leader to embed it within their area.

One of those executives, Gemma Sherrington, then heading fundraising and marketing, leapt into the deep end. As she described it later, she reorganised her entire division around Agile principles: squads, sprints, backlogs, the full architecture, with great enthusiasm.

Save the Children UK had an annual income of over £300 million and operations across more than 50 countries. When a leadership team of that scale changes how it makes decisions, it has direct consequences for what gets delivered and to whom.

The 17th Annual State of Agile Report (Digital.ai, 2023), the longest-running survey on Agile adoption drawing on 788 respondents across sectors, found that 47% of organisations cite general resistance to change or culture clash as the primary reason Agile fails to embed in a business, and this was up seven points on the previous year. The next two barriers were closely related: 41% said insufficient leadership participation, and 38% said inadequate management support or sponsorship. Overall satisfaction with Agile dropped from 71% to 59% in a single year.

The respondents were mainly from technology and financial services, and drawing from outside the sector makes Agile’s application to charities useful. The gap between process and culture is a structural problem that appears in every type of organisation that implements a new way of working without first addressing the beliefs and authority structures it depends on. What Agile is supposed to do is help organisations make this shift more smoothly by providing a structure for implementation, much like PRINCE2® does for project management. But the structural problem exists, and is intensified in charities by complex governance, challenging funding cycles, and professional identities built around subject expertise. Changing any of these conditions requires authority - in the form of a person who can initiate cultural change - that sits above the implementation brief.

Save the Children's initial Agile implementation did what most implementations do: it changed the process. Teams had sprints. There were stand-ups. Boards were set up. All the visible apparatus of Agile was in place.

What it didn't change, and couldn't change through process design alone, was what teams thought about what they were there to deliver. Senior staff whose credibility rested on subject-matter expertise were now being asked to prioritise the beneficiaries’ needs over the application of their own expertise. Without realising it, Save the Children had thrown its staff into a cultural conflict in which values clashed with delivery.

So even though staff had been asked to adopt the new process, and they did, the belief system they always had was unchanged, which meant the process was, in effect, running on the same values and cultural logic it was designed to replace.

This is where the structural condition required for successful Agile implementation becomes apparent. In most organisations, values work and implementation work are treated as separate commissions. Sometimes implementation comes first, with instructions to proceed now, with values to follow. Sometimes values come first, with principles agreed at the senior level, then handed to operations to deliver. Either way, the two workstreams rarely share a room. And whichever one arrives second is handed to people who had no input into the first.

Implementation without values produces a process that stubbornly embeds the old logic. Values without implementation input produce principles too abstract to operationalise, aspirational on paper and unworkable in practice. The delivery experience ends up reflecting neither. 

The mismatch ends up on the ops leader’s desk, even when that person had no hand in shaping the values or authoring the implementing brief. Yet they are handed both, in sequence, and asked to make them cohere. Then a greater mismatch materialises between what the organisation says it is and what its beneficiaries experience, and the ops leader is usually left to fix it. 

Who in your organisation is currently managing the gap between what the values say and what the delivery experience is, and did anyone ask them to?

Save the Children's pivot worked because Gemma Sherrington had enough authority and enough candour to name the mismatch and redesign around it. Most ops leaders don't get that runway. All they get is the grumbling

The initial implementation hadn't delivered what the model promised. Rather than rebranding the initiative or letting it wind down quietly, the leadership team chose to admit the implementation hadn’t met expectations and to ask what the experience had revealed.

The answer was definitive. Process change without mindset investment produces compliance, where they were looking for transformation. 

The pivot that followed was a culture-first approach, built around three commitments: human-centred design and customer-centric approach; an iterative, test-and-learn set of practices; and transparency and open collaboration. Critically, these weren't declared as values; instead, they were operationalised through the recruitment of new staff members, the structuring of teams, and the behaviour that was recognised.

The structural response required designing around the conflict between maintaining professional discipline and development while freeing people to work across functions on specific outcomes. To address this, Save the Children arrived at two distinct types of teams with different purposes. Home teams were stable groups organised by discipline or skill set, providing professional development and continuity, the place where a fundraiser was still a fundraiser. Outcome-objective teams were time-limited and multidisciplinary, formed around a specific goal and dissolved when that goal was met (much like a task-and-finish group, but more culturally embedded). The structure gave the new process a more concrete footing in the organisation without dismantling the expertise it depended on.

Recruitment shifted, too. The capability they looked for in new staff members changed from expertise to mindset, specifically a growth mindset, evidenced through interview techniques focused on how candidates described failure and what they'd learned from it.

Based on what Save the Children's transformation required, and what the evidence suggests made it work, the core approaches were:

Before the pivot, a manager could tell you which projects were active and their current status. After the pivot, a shift from tracking activity to owning outcomes was evident: a manager could tell you which projects were most likely to create outcomes that made a difference and explain why the others were lower priority. 

What the staff survey data and Sherrington's own account suggest is that decisions escalated less, teams had a clearer line of sight between their work and the outcomes they were accountable for, and staff reported feeling less blocked by bureaucracy, not just because process disappeared, but because it was more evidently in service of a shared purpose.

Teams could form and dissolve around specific goals without the friction of permanent restructuring. The organisation became measurably faster at responding to new contexts because the constraints were better designed

Sherrington's 'no regrets' framing could be mistaken for a PR line. But it was a structural choice. By treating the failed implementation as a hypothesis test rather than a mistake everyone had to forget, the leadership team gave themselves permission to learn from it. So, what can you learn from it? If any initiative falls short, write down what you believed would happen, what did happen, and what the gap reveals. Let that answer drive the next decision. Unfortunately, organisations that can do this are rarer than they should be.

It would be easier to claim that most charities already run pilots, and that what Save the Children did is not unique. But pilots are often not by choice; more often, they are bound by funding cycles and money that comes in tranches. The more demanding discipline, that many charities will miss, is to identify, before the pilot begins, what evidence would need to be shown to justify full implementation. Without that, a pilot is just a smaller version of the same risk.

The culture-first pivot worked for Save the Children because it changed what was measured and rewarded. Values written on walls and website homepages don't change behaviour. But recruitment criteria, team structures, and decision rights do. The new way of working gave the culture somewhere to live, and the culture gave the new way of working a reason to matter.

The dual team structure solved a problem that most Agile implementations in complex organisations never address: how do you maintain professional discipline and development when staff are constantly redeployed to cross-functional work? Home teams answered that question. Without them, the outcome teams would have eroded expertise rather than deploying it.

The culture-first approach requires someone with authority over culture to own it. If the ops leader is implementing the new way of working, but a different part of the leadership team controls recruitment criteria, team structures, and what behaviour is recognised, the conditions for it to work are outside the implementation. The early warning sign that the new way of working has been adopted but not absorbed is when sprint reviews are happening, but the decisions that matter (resourcing, priorities, and scope) are still being made in separate conversations that the implementation doesn't touch. 

The dual team structure depends on clear prioritisation at the top. If the organisation can't agree on its top three priorities, outcome teams form around whatever is loudest rather than what is most important. The result is Agile theatre: the speed and flexibility of iterative working, pointed at the wrong things. Watch for teams that hit every sprint target but, when asked what is different for the beneficiaries the organisation serves, go quiet. Note also that this model is designed for organisations with enough people to sustain both structures simultaneously. In smaller charities, the principle is largely the same, but the architecture looks different. For smaller organisations, the distinction between home and outcome teams may be a conversation about how time is allocated and to what end, rather than a formal team design.

Mindset-first recruitment only works if existing staff are supported through the changes. Hiring for a growth mindset while the culture still rewards expertise-first behaviour creates a two-tier workforce, with new joiners working one way and established staff working another.

Save the Children’s story ends well. The pivot to a culture-first approach worked, the new way of working was embedded, and the structure persisted. But it took a failed implementation to generate that insight. The question worth brewing: is there a process in your organisation you’re currently pressing ahead with that isn’t working, one you’ve perhaps been hearing grumblings about, and if you treated it as a hypothesis rather than a commitment, what would the evidence tell you?

Every week, a real ops problem from a real charity:
What they did, what it reveals, and what you can take back to your desk.
Sign up to get it in your inbox.