A special episode in partnership with the World Bank Group.
An organisation that aims to reduce poverty and promote sustainable development with knowledge, financing, and expertise.
In this episode, host Debbie West brings together two leading thinkers on public sector reform and development:
Sir Michael Barber
Lisandro Martin
They discuss big topics to explore how governments and global institutions can deliver intelligent outcomes in an era defined by rapid technological change.
Debbie opens the conversation by inviting Sir Michael Barber to reflect on Deliverology, the approach to public sector reform he shaped in the early 2000s. Michael outlines the five core questions at the heart of the model, highlighting how clarity of outcomes, credible planning, real time data, problem solving and central support created a disciplined yet collaborative method for achieving large scale results.
He explains that initially viewed with suspicion inside government, the approach quickly earned trust as early wins emerged and ministers benefited from focused time with the Prime Minister. Michael stresses that the model succeeded because it combined rigour with partnership, and because delivery units remained resolutely undistracted by political crises, holding a laser focus on outcomes that mattered to citizens.
Fast forwarding to the present, Michael argues that Deliverology has only grown in importance as citizens demand better services without higher taxes, and as governments face intense pressure to improve outcomes with finite resources. He notes that technological advances, such as rapid and low cost data collection, now strengthen delivery efforts.
He shares a striking example from Punjab, Pakistan, where tablet based vaccination tracking enabled real time monitoring, targeted deployment of vaccinators and direct follow up with families. This innovation contributed to one of the fastest rises in vaccination coverage in history, illustrating how technology can transform delivery when paired with focus and intent.
Lisandro offers a complementary view, framing outcomes as a matter of dignity. For the World Bank Group, he explains, outcomes are about money translating into real improvements in people’s lives: better health, brighter futures and greater human dignity.
He argues that outcomes should no longer be seen as the end point of a project but as a lens guiding every decision. Each policy choice, budget shift and operational step links to long term results, and acknowledging this connection helps development institutions move beyond inputs towards genuine impact.
Lisandro also describes outcomes as a shared language that can unite stakeholders even when they disagree on policy pathways. While methods may differ, few argue against a world with less poverty and more stability.
The episode turns to the role of AI in shaping modern delivery systems. Michael notes that although its long term implications remain uncertain, AI is already changing how organisations operate. Faster, more granular data supports sharper accountability, as seen in the UK Department for Education’s AI enabled attendance monitoring system, which led to significant improvements in school attendance.
Lisandro expands on this, describing how AI turns rear view reflections into predictive insights. At the World Bank, AI supports forecasting progress towards ambitious global targets and identifies what must change to accelerate outcomes. It also processes vast archives of documentation to extract lessons too complex for human analysis alone.
Both speakers emphasise that AI amplifies human capability rather than replacing it. By automating information gathering, it frees people to apply judgement, empathy and creativity to complex delivery challenges.
Behavioural science also plays an important role in shaping both public perception and internal decision making. Lisandro draws on Daniel Kahneman’s concepts of system one and system two thinking, reflecting on how AI blurs the line between fast and deep thinking. He suggests that AI could help governments address biases such as present bias and over precision, enabling more informed and adaptive policy making.
Michael adds practical examples from the UK’s nudge unit, illustrating how small behavioural insights, such as peer reminders for students or community tax comparisons, can drive meaningful changes in outcomes. These approaches are strengthened by the granular data that AI can provide.
Debbie invites the guests to explore how governments can develop what Lisandro calls an outcome reflex. Lisandro explains that the first step is extreme clarity about a small number of priorities. He notes that the World Bank’s shift from 153 institutional priorities to just 15 represents a deliberate move from measurement complexity to directional clarity.
He highlights the importance of frequent reflection, enabled by real time data, and argues that traditional ex post evaluations are too slow for today’s environment. Instead, systems must support continuous learning and adaptation.
Michael reinforces this, emphasising that data informs decisions but does not replace values. He supports rethinking evaluation altogether, using AI to support near real time learning rather than retrospective assessment.
The conversation turns to urgency. Lisandro notes that urgency has shifted from speed to calibration. With information overload and fast moving data, the challenge is identifying the right moment for decisions. He stresses that policy makers are not always in a mindset to learn, so delivery systems must align evidence with the moments when actors are prepared to act.
Michael roots urgency in the lived reality of citizens who cannot wait for essential services. He points out that delays often stem from institutional habits and risk aversion rather than genuine need for more research, and that urgency must be grounded in moral responsibility.
Asked to look ahead, Michael describes outcome reflex as a natural evolution of Deliverology: keeping focus on outcomes while acting with immediacy. He believes AI will accelerate governments’ ability to monitor, learn and improve in real time, but leaders must continue to provide strategic direction and resist distractions.
Lisandro adds that in a world facing severe long term challenges, adaptability is vital. With millions entering the global job market and limited opportunities available, societies cannot afford slow course corrections. He argues that governments must work in perpetual beta, testing and adjusting policies continuously rather than committing to rigid plans.
Michael suggests his book How to Run a Government as a helpful starting point for understanding delivery principles, alongside recent World Bank reports and research from leading academic institutions.
Lisandro points listeners to extensive resources on the World Bank Group website, including dedicated content on outcomes and AI. He also shares insights regularly through his LinkedIn page.
This episode offers a rich and forward looking conversation on how intelligent outcomes can be delivered through a blend of clarity, collaboration, technology, behavioural insight and moral purpose. It highlights both the enduring relevance of disciplined delivery and the transformative potential of AI in shaping systems that improve lives worldwide.
