We are delighted to share this thought-provoking guest post from Louisa Mitchell MBE, Chief Executive of AllChild.
AllChild works with communities in underserved neighbourhoods to help children and young people develop the skills they need to thrive. Their place-based approach is rooted in deep partnership, early identification of need, and a commitment to personalised support through dedicated Link Workers.
In this blog, Louisa reflects on how AllChild’s values-led model has shaped their evolving approach to evaluation – blending traditional methods with innovative practices to capture both ‘hard’ outcomes and the vital relational changes that underpin lasting impact. She also outlines how AllChild’s learning journey is informing their ambition to build a national movement for investing differently in children’s futures.
We are grateful to Louisa for sharing these insights and for her leadership in championing early, preventative action for children and families.
Values and evaluation
AllChild is guided by a CLEAR set of values – Collaborative, Local, Evidence-led, Accountable and Relational. These values define not only how we work, but also how we evaluate our impact, which can prove a challenge. We hold ourselves to high standards of evidence and accountability – to the children and communities we work with, every day.
Our work is rooted in a whole-system, place-based, and personalised approach. We believe that a shared vision and a shared outcomes framework can create value such that the sum of the parts is greater than the individual – focusing on contribution as well as attribution. Our model – delivery, impact and funding – is intentionally dynamic, continually adjusting and evolving in response to the unique needs and contexts of each community we work with. We meet each child where they are at, so while we follow core principles, each child has their own unique journey to achieve the social, emotional and/or academic outcomes they aim for.
Trust is at the heart of our engagement model. And we take a long-term, preventative approach, recognising each child’s great potential and unique strengths, and working with them to address barriers that are preventing them from fulfilling this potential, rather than waiting to retrofit them into a standard list of services or interventions once challenges escalate.
Evaluating trust and proving something that hasn’t happened yet, or which is much less likely to happen due to action now, presents an evaluation challenge, but one we embrace as essential to evidencing lasting, positive change.
Our approach to local transformation
We take a local transformation approach by:
- Listening to and co-designing with the communities where we work
- Identifying children early who are not well supported and face multiple risk factors that may lead to poor outcomes later in life
- Delivering two-year Impact Programmes that provide each child with tailored support and opportunities, grounded in trusted relationships. Every school has a dedicated AllChild Link Worker leading the programme, supported by a connected local ecosystem including schools, councils’ Early Help systems, voluntary sector organisations and local businesses.
Through the deep trust and connections established via these Impact Programmes, we aim to:
- Mobilise communities to lead the long-term change they want to see
- Embed connected systems of support and opportunity within each place
- Establish an AllChild Network that builds a national movement for a new way of investing in children and young people.
Combining tradition and innovation in evaluation
We would traditionally be categorised as a ‘complex intervention’ – a term I find limiting. We prefer ‘Social Outcomes Partnership’[1] which perhaps better reflects our child-centred system of support and opportunity in which thousands of children over the last 8 years have achieved positive social, emotional and academic outcomes.
While complex interventions can be evaluated using traditional evaluation methodologies[2], we have recently found it challenging to compare outcomes for children and young people on our Impact Programmes against any progress they might have made without our support. In partnership with UCL[3], we took the two-year Impact Programme – one component of AllChild’s local transformation model – and tried to compare social, emotional and academic outcomes between participating children and comparable peers. UCL employed Propensity Score Matching using wide and narrow comparison groups but determined that ‘…young people selected into AllChild are extremely different to their peer group and, as a result, it is extremely difficult to identify a suitable number of truly comparable individuals as a comparison group”.
This confirmed that our targeted, detailed methodology of identifying children who will most benefit from AllChild’s Impact Programme is reaching the right children. While the evaluation results are somewhat promising – and insights informed by an accompanying Process Evaluation enabled us to make some improvements – this work has continued to reveal the significant challenges faced in trying to develop a robust evaluation design for our model of support using comparison groups.
From our early days as West London Zone, we focused heavily on numbers, driven by our values and our strict commitment to the five Collective Impact principles[4] one of which was ‘shared measurement’ about ‘collecting data’. But numbers alone cannot tell the full story, especially for an organisation that puts ‘soft’ trust and relationships, as well as ‘hard’ outcomes, at its heart.
We have always used case studies incorporating context and stakeholder testimony to learn from and to communicate our contextual and hard-to-measure impact. Methods such as pairs analysis of children who have progressed well with AllChild and those who haven’t, interviews, focus groups and annual deep dives highlight impact that cannot be easily represented numerically but is transformational – like a child who couldn’t make eye contact now confidently speaking to adults.
Going forward, we need to design a robust evaluation approach that values both the ‘hard’ and ‘soft’ outcomes, especially relational outcomes, or even relationships as outcomes.[5] Because at AllChild, while we have the depth and range of hard data demonstrating the social, emotional and academic outcomes the children have achieved, we know that often the most meaningful progress for a child and their family cannot be reduced to a number.
Our evolving evaluation approach
We want to learn, improve and refine, in real time – especially as we look ahead to our next 10-year ambition to:
- Work in more places
- Build our evidence base for local transformation
- Drive a national movement for a new way of investing in children.[6]
We are committed to better understanding and communicating the impact of our place-based model, not only with children and young people, but also with families, schools, and local Early Help systems. We want to support every community we work in to build a connected, early action ecosystem of support and opportunity.[7]
Learning by doing
To inform our future evaluation strategy, we have partnered with Coram Impact and Evaluation, to help us evaluate our process and impact in our newest place of work, Leigh (Wigan Council). This extensive project strategically combines a range of evaluation methodologies, including Contribution Analysis[8] to test our theory of change. The approach includes deep interviews led by young researchers, aimed at providing rapid learning for real-time adaptation in Leigh. The insights will guide our work in new places.
In parallel, we are working with partners to explore new ways of benchmarking outcomes achieved by young people on our 2-year Impact Programmes and to improve targeted data sharing with our Local Authority partners. This will help estimate the long-term savings resulting from AllChild’s Impact Programmes and strengthen the case for shifting investment towards early and preventative action – supporting young people when it matters most.
Our future evaluation strategy
In 2026/27, as we learn from the Coram learning partnership, we will establish our future evaluation approach that will bring together a range of methodologies, recognising that no single method alone will be perfect. Traditional evaluation approaches may still play a role, but these will need to be tailored for our Social Outcomes Partnerships and implemented alongside a range of tools.[9]
We now understand that to be truly aligned with our values, our evaluation approach needs to be as innovative, curious and wide-ranging as our delivery, impact and funding model itself. And we will be taking the time needed in the coming months to build an evaluation strategy that is rigorous, robust and focused.
[1] chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://golab.bsg.ox.ac.uk/documents/The_Evolution_of_Social_Outcomes_Partnerships_in_the_UK.pdf
[2] https://taso.org.uk/news-blog/evaluating-complex-interventions-using-randomised-controlled-trials/
[3] UCL Centre for Education Policy and Equalising Opportunities (CEPEO) and the Helen Hamlyn Centre for Pedagogy (0 – 11 years)
[4] https://ssir.org/articles/entry/social_progress_through_collective_impact
[5] chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://groundswell.org.uk/wp-content/uploads/2021/12/Re-thinking-Outcomes-Guide-for-Services.pdf
[6] Our 10-year ambition and the learnings behind it are set out in our latest report ‘From Local Learning to National Ambition’ https://www.allchild.org/our-impact/reports/from-local-learning-to-national-ambition
[7] https://www.thinknpc.org/resource-hub/systems-change-a-guide-to-what-it-is-and-how-to-do-it/
[8] https://www.betterevaluation.org/methods-approaches/approaches/contribution-analysis
