From Hindsight to Foresight: How Evaluation Can Become Future-Informed

In their writings, Rose Thompson Coon et al. highlight the need to rethink evaluation in order to incorporate a forward-looking, even futuristic, dimension. Indeed, whilst evaluation enables lessons to be learnt, it does not always provide immediate avenues for their operational application once the intervention has ended. Without an in-depth literature review and uptake by other researchers or designers of future projects, the lessons learned from intervention evaluations tend to be forgotten once the interventions have ended. Thus, evaluation reports would benefit from incorporating more in-depth analyses, enabling this forward-looking and future-oriented vision of development to be better taken into account.

It is quite hard to comment on an article that is not fully accessible :-(

Powerful perspective. Moving from hindsight to foresight is exactly where evaluation must evolve using data not just to report, but to anticipate and improve outcomes. This is critical for effective programme design

We can definitely become better at being "forward-looking": understanding likely patterns, more intentional in interrogating likely consequences. But always escaping the temptation to make this "THE plan". Because what matters is having direction and agility, better capacity to see and feel junctures... not a pre-set future. 

The deeper question is: what is evaluation for? If it exists mainly to confirm compliance (i.e. to verify that a plan was executed as designed, that the Theory of Change held....) then adding foresight methods changes nothing. We will simply be anticipating the future in service of the same backward-looking logic and the same set of horizons. Always in "compliance mode." Before we ask how evaluation can get better at anticipating the future, we need to ask a prior question: are we willing to set evaluation free from the obligation to confirm the plan?

The paper offers  one option for evaluating new speculative ventures which are not under control of the public. Using foresight methods instead of past performance can also backfire.

 I recently evaluated blended finance funds, the performance of the funds was not as expected due to a set of factors driven from research on management and business development like know your customer. Our subject experts always tried to developed new scenarios for the future under which the facilities would work better. Foresight tools are applied within the existing evaluation system based on evaluation departments or institutions and methods which in most cases lack hard technical / subject-related  skills and resources improve the evaluation system. Alternatives to the current evaluation system should be explored like participatory process for policy formulation and implementation like stakeholder groups in specific policy areas providing  feedback on specific policies. This way future oriented  reflexive and learning systems are created using feedback loops. Here future programming is  included by stakeholders. Examples are the Brazilian policies for family agriculture or nutrition which are shaped by stakeholder commissions at all levels. 

Evaluation can no longer be content with being a mere act of measurement or a snapshot frozen in time; its true purpose lies in its ability to bring about lasting transformation. By moving beyond its traditional monitoring function to become a lever for change, it acts as a catalyst for self-reflection, compelling stakeholders to confront their practices with the reality of the results. To evaluate without transforming would be like making a diagnosis with no intention of treating the condition, rendering the exercise fruitless and purely bureaucratic. On the contrary, an evaluation focused on progress enables the identification of areas for improvement and sources of innovation, turning mistakes into learning opportunities and judgement into a tool for support.

From an ethical and strategic perspective, this transformative dimension is essential to avoid inertia and ensure the effectiveness of the actions undertaken. In a constantly changing environment, evaluation must serve as a dynamic compass: it does not merely look back to validate achievements, but propels organisations or individuals towards the future by adjusting their trajectories. In short, evaluation only achieves its full value when it becomes an “empowering” process, capable of changing behaviours and optimising systems to deliver real and tangible impact.

I really appreciate the insights shared, especially the focus on adaptive learning and forward-looking evaluation. I would also argue that one critical limitation lies in how evaluation knowledge is produced and communicated. This time, my comment is less focused on methodology and more on offering my perspective in response to your question Steven about what needs to change for evaluation to truly contribute to transformation.

Evaluation reports are often written by evaluators, using technical and methodological language that mainly speaks to other evaluators or technically trained audiences. As a result, these reports tend to be most appealing and useful to those who already understand that language.

What I find interesting is that most projects have communication plans that clearly identify who needs what information, how it should be shared, and why. However, this logic is rarely applied when writing evaluation reports. Different stakeholders have very different needs depending on how they interact with the project, and a single report cannot effectively serve all of them.

To me, evaluation reports should be seen as a starting point, not the final product. The data and findings should serve as inputs into multiple targeted knowledge products, developed by communication experts or sector specialists (health, education, economic development, etc.), that speak directly to specific audiences. These tailored outputs would translate evaluation insights into formats and messages that are relevant and actionable, supporting implementation, positioning, and donor engagement.

In addition, while I understand the rationale behind lean data approaches, they can sometimes be too restrictive. Focusing only on what is needed for indicators or donor requirements may limit opportunities to explore emerging issues or strategic areas more deeply. Sector specialists could play a role here by collecting additional data for learning, positioning, thought leadership, or future programming, as long as there is clear accountability for why that data is being collected and how it will be used. Every question has a cost, but it can also bring real value if it is intentional.

Overall, if evaluation is meant to support transformation, it is not just about improving methods or tools. It is also about making sure that the knowledge we produce is usable, relevant, and accessible to the people who need it. We need to better understand our audiences and what they need.

In African and low-resource health systems, evaluation too often serves as a post-hoc accountability exercise rather than a tool for systemic transformation. From decades of practice in health systems strengthening, the first and most critical shift must occur in mindset - how we perceive the purpose and ownership of evidence. Evaluators and decision-makers frequently operate with a compliance mindset, producing reports that satisfy external donors but fail to capture the nuanced realities on the ground. Recently, in Lagos State, Nigeria, routine monitoring in the maternal health program had focused narrowly on facility births. But, by adopting a learning-oriented approach - examining quality of care, patient experience, and referral patterns, it was uncovered that 42% of women bypassed local clinics due to perceived low-quality services. Targeted staff training and resource reallocation subsequently increased facility-based deliveries by 17% within a year.

Similarly, Community-Led Monitoring in another Nigerian district revealed a 40% barrier from hidden transport costs, despite reports showing 95% patient “satisfaction.” These insights highlight those methods and criteria, however technically sound, that follow effectively only after the mindset evolves to prioritize adaptive, locally-informed learning over extractive reporting.

Evidence from the recent Q1/2026 ‘Life & Health’ dialogues of Afrihealth Optonet Association (AHOA) shows districts using integrated digital platforms and participatory evaluation achieved a 15% rise in immunization coverage - proof that embedding evaluation in real-time problem-solving, not just retrospective reporting, produces tangible health impact.

Sustainability and long-term development hinge on this alignment. Transformative evaluation is not about better spreadsheets or fancier dashboards; it is about decolonizing intent, ensuring data serves local solutions, and fostering a culture of critical inquiry. In constrained African health systems, the mindset shift is the fulcrum upon which all methods, criteria, and institutional reforms pivot toward lasting, systemic change.

If this weeks question is "If evaluation is to contribute to transformation, what exactly must change first…our methods, our criteria, our institutions, our underlying mindset, or something else?" I will repeat my point below...there is a potentially serious misalignment between the predictive nature of a theory of change as used by evaluators and the need for preparedness in a very uncertain world, A singular view of the future versus a plural view of the futures

[By the way, the display structure for this kind of dialogue should be branching, not a single line]

And please lets not toss the word "transformational" around too lightly. Putin and Trump and other despots are all keen on transformation of one kind or another. What about incremental improvements, or perhaps even just surviving, as an objective ? :-) Inflation take many forms, including changes in our vocabulary. e.g In the past we just had "details", but now we have "granularity" Wow....things really are getting better...or are they?

Hello everyone,

I’m delighted to be taking part in this discussion. Having a background in philosophy, I will no doubt be asking questions of a different nature. Steven Lichty asked: “If evaluation is to contribute to transformation, what needs to change first… our methods, our criteria, our institutions, our underlying mindset, or something else?”

I don’t think we should frame the question in this way. Transformation is not a linear process where we move from a first stage to a second, then to a third, and so on. On the contrary, several things happen simultaneously, and where each of us starts depends on ourselves, on what we are evaluating, and on the perceived purpose of the evaluation. In other words, what matters is the feedback loop between the assessment and the people it concerns.

Our team recently assessed a building. It met the requirements for LEED certification. To our surprise, it exceeded the LEED criteria. We incorporated these additional features into our assessment (our method is quite flexible in this regard). We then used this assessment to formulate recommendations for further improvements. Thus, the assessment method included these recommendations, and these led to a conversation about feasibility, implementation, timing, importance, and so on. This conversation, in turn, feeds back into the assessment and the recommendations. The conversation also leads those involved in the assessment to ask themselves in-depth questions about themselves, who they are and what they wish to become within the context of the institution. Thus, criteria, institutions and mindsets all evolve simultaneously through our assessment and feedback process.

When the future is looking more uncertain than ever it may be useful to think about preparedness more than prediction. To enhance preparedness we need to be thinking about mutliple versions of the future, not just one, and these versions need to be sufficiently diverse. Having generated those futures how can we then evaluate them? I would be interested to hear from others what they think might be relevant criteria to apply. To kick off, I suggest the criteria may fall into two broad categories: 1. Cognitive: Criteria relating to how we are thinking about the future, 2: Behavioral: Criteria relating to how repond to those futures before and after they are realised

Is it possible to respond to contributions and pick a conversation? I could not find the option to do so. And starting a new message breaks the flow!

FROM HINDSIGHT TO FORESIGHT: EXPERIENCE AT AFRIHEALTH OPTONET ASSOCIATION

by Dr. Uzodinma Adirieje

From hindsight to foresight, our experience at Afrihealth Optonet Association (AHOA) demonstrates that evaluation is most valuable when it moves beyond retrospective accountability to actively shaping future decisions under uncertainty. Three practical insights stand out.

Embed adaptive learning loops into programme design: 

In Afrihealth’s health systems and climate-linked interventions, periodic reviews were not treated as endline exercises but as real-time checkpoints. Evaluators facilitated rapid feedback cycles - combining routine data, beneficiary insights, and contextual signals (e.g., policy shifts, climate events like COP29 Baku) - to inform mid-course corrections. This approach ensures programmes remain relevant even as conditions change.

Integrate mixed-methods evidence for anticipatory analysis: 

Quantitative indicators alone often lag behind emerging realities. Afrihealth’s evaluations paired service delivery data with qualitative intelligence from communities and frontline workers. For example, shifts in health-seeking behaviour during economic stress were detected early through interviews and focus groups, enabling proactive adjustments in outreach and resource allocation.

Align evaluation questions with decision horizons:

Rather than asking only “what worked,” Afrihealth reframed inquiries toward “what is likely to work next, for whom, and under what conditions.” Scenario-building and contribution analysis were used to explore plausible futures, particularly in programmes intersecting with climate variability and public health risks. This made findings directly usable for strategic planning, not just reporting.

Stakeholder co-creation:

By engaging policymakers, implementers, and communities in defining evaluation priorities, Afrihealth ensured that findings addressed real decision needs. This strengthened ownership and increased the likelihood that recommendations were acted upon.

Similarly, optional readings in developmental evaluation and adaptive management further reinforce these practices, emphasising flexibility, systems thinking, and learning-oriented accountability.

This way, evaluators can enhance relevance in uncertain contexts by institutionalising real-time learning, triangulating diverse evidence, and orienting evaluations toward future-facing decisions. 

Dr. Uzodinma Adirieje is a former National President of the Nigerian Association of Evaluators (NAE). He is a seasoned evaluator, health economist, and civil society leader who was the co-consultant to drafting Nigeria’s National M&E Policy. He led SDG3 evaluation synthesis, participated in national SDG 3 and SDG 4 evaluations, and provided M&E training and mentorship, advancing evidence-based, forward-looking development practice.

Steve, thank you for this thoughtful engagement and for bringing in Michael Quinn Patton’s work, which I also find highly relevant to this discussion.

I very much agree with your reading that a foresight lens does not necessarily require a parallel methodology, but can be embedded in how we interpret and apply existing frameworks. In that sense, your point about a “more honest application” of the DAC criteria resonates strongly with my own experience particularly in contexts like One Health, climate change adaptation, and agrifood system transformation, where systems are evolving even as we evaluate them.

At the same time, Silvia’s intervention pushes this one step further in an important way. I share the concern that if evaluation remains anchored in a compliance-oriented logic, even well-integrated foresight risks being instrumentalised used to anticipate within predefined boundaries rather than to genuinely question them. The distinction she draws between evaluation as verification versus exploration is, I think, exactly right.

In my view, however, the real constraint on integrating foresight is often not at the level of tools or criteria but much earlier, at the stage of evaluation conceptualisation.

In the FAO One Health case, the ability to incorporate a foresight perspective was enabled by an in-depth preliminary analysis and literature review conducted at the design stage. Without that early investment, it would have been significantly harder to introduce a meaningful forward-looking dimension later on. By the time questions, scope, and methods are fixed, the evaluation architecture is already path-dependent—ironically mirroring the very dynamics we are trying to assess.

So perhaps the discussion can be nuanced in three directions:

• I agree with you, Scott, that criteria can be reinterpreted.
• I agree with Silvia that culture and purpose are decisive.
• And I would add that timing is equally critical.

If foresight is to be more than an add-on, it needs to be designed in from the outset, not retrofitted.

This has practical implications for commissioners. If we are serious about developmental or formative approaches, foresight needs to be reflected in:

• the initial scoping work,
• the framing of evaluation questions, and
• the evidence gaps we choose to prioritise.

In that respect, the approach we often use at FAO, a question-driven, utilization-focused design, guided but not constrained by OECD DAC criteria does offer some flexibility. It allows us, at least in principle, to embed forward-looking dimensions early on, provided that the conceptual groundwork is strong enough.

So perhaps the challenge is not only to rethink criteria or embrace foresight tools, but also to shift attention upstream: to how evaluations are commissioned, framed, and intellectually grounded before they even begin.

Traditionally, evaluation has been viewed as a retrospective exercise aimed at analyzing the results of a project or public policy after its implementation. However, in a context marked by uncertainty and the growing complexity of public interventions, evaluation is gradually shifting toward a more forward-looking approach, focused on anticipation and continuous improvement.

First, forward-looking evaluation relies on the integration of learning mechanisms. It is no longer merely a matter of judging past performance, but also of identifying lessons learned in order to improve the design and implementation of future actions. This approach promotes adaptive management of projects and programs.

Second, forward-looking evaluation relies on the use of foresight tools such as scenario analysis, ex-ante impact assessments, and risk modeling. These tools help inform decision-making in advance and guide public policies toward sustainable outcomes.

Furthermore, future-oriented evaluation encourages real-time monitoring and continuous assessment. Through information systems and performance indicators, it becomes possible to adjust interventions as they are implemented. This dynamic enhances the responsiveness and effectiveness of projects.

Finally, the forward-looking dimension of evaluation involves greater stakeholder participation. Engaging beneficiaries, decision-makers, and experts helps identify future needs, anticipate challenges, and develop tailored solutions.

Ultimately, shifting from retrospective to prospective evaluation involves transforming evaluation into a true decision-making tool. It thus becomes a strategic lever that not only analyzes the past but, above all, prepares for the future and sustainably improves public action.

Ms. Amy MARA

Economist and procurement specialist

Ph.D. candidate in project management

In my experience, what really makes an evaluation "forward looking" is its focus on future improvement rather than just judging what happened in the past. A forward-looking evaluation digs into what we've learned, offers practical recommendations and helps us adapt to new challenges or opportunities that might come up. It’s not just about checking if we met our old goals, but about asking, “How can we do even better next time?” I find this approach especially valuable because it encourages ongoing learning and helps everyone involved plan more strategically for the future. What do others think—have you seen examples where a forward-looking evaluation really made a difference?

I find this discussion particularly relevant. In my work, I often use time series analysis and predictive modeling to estimate future trends based on historical data.

I would add that while foresight approaches that do not rely on past performance are essential, especially in contexts characterized by high uncertainty or limited data, predictive methods grounded in historical data remain among the most robust tools we have when sufficient and reliable data is available. These methods allow us to identify patterns, quantify trends, and generate evidence-based projections that can effectively complement more qualitative foresight approaches.

With the advancement of AI and machine learning, we now have the capacity to go further by integrating large volumes of data across multiple projects, regions, and even donors. This creates important opportunities to build more accurate and context-sensitive predictive models, particularly when working with similar interventions within a country or sector.

However, a major constraint remains data availability and fragmentation. Data is often siloed within individual projects or organizations, making it difficult to build sufficiently large and diverse datasets for robust modeling. In many cases, data from a single project is not sufficient to support reliable predictions.

One potential way forward would be to strengthen national ownership of project data. Governments could play a key role in consolidating data generated across projects into centralized and accessible databases. If well designed, such systems could support research, inform project design, and enable more rigorous ex-ante analysis of potential success or failure.

In that sense, I see strong complementarities between foresight methods and predictive analytics. Foresight helps us explore uncertainty and alternative futures, while predictive models help us quantify likely trends where data allows. Bringing both together could significantly strengthen evaluation practice and decision-making.

I would like to thank you for this initiative.

Retrospective evaluation is essential for assessing impact and drawing lessons. However, in contexts shaped by climate change, it presents significant limitations, particularly in integrated landscape management approaches.

For example, costly infrastructure such as dams can be destroyed or become ineffective due to extreme climate events. In such conditions, it becomes difficult to measure the real impact of a project or to attribute observed results to the intervention rather than to external factors.

This uncertainty affects the use of evaluation results, as they may be perceived as unreliable or not representative, thereby limiting their usefulness for decision-making and future planning.

Thank you for this discussion and for the initial ideas shared!

As someone who consistently puts "forward-looking evaluation" at the centre of my proposals, I want to offer a provocation.

The framing here might suggest that what evaluation needs is better foresight tools and more capacity to anticipate the future. I'd like to challenge that, not to dismiss foresight, but to locate the real problem one level up. Because the issue is not technical. It is cultural.

The deeper question is: what is evaluation for? If it exists mainly to confirm compliance (i.e. to verify that a plan was executed as designed, that the Theory of Change held....) then adding foresight methods changes nothing. We will simply be anticipating the future in service of the same backward-looking logic and the same set of horizons. Always in "compliance mode."

Before we ask how evaluation can get better at anticipating the future, we need to ask a prior question: are we willing to set evaluation free from the obligation to confirm the plan?

Can evaluation be exploration, not verification? That means evaluations that do not just answer questions, but discover better ones, that help people think through the future, not spoon-feed it to them.

Foresight tools are valuable. I have used them. And when used in a participatory way they can be liberating, revealing that people already carry vision and insights that the very plans they are working on tend to constrain.

So this is the real issue. It is not "foresight" as a technical fix. It is about the power to adapt, challenge, and explore continuously.... rather than situating evaluation in a world where our assumptions, our theories, our plans are reference points, and not starting ideas.

Thank you for initiating this important discussion. To facilitate this discussion, I would like to share some reflections.

Today’s world faces unprecedented challenges of climate change, food security, environmental sustainability, and increasing fragility due to conflicts and related crises. Agricultural development programs operate amid a backdrop of volatility, uncertainty, complexity, and ambiguity. 

Traditionally, evaluation function has focused predominantly on retrospective accountability, measuring past performance against predetermined plans, objectives and targets. While valuable, this approach, in today’s rapidly changing context, often doesn’t result in valuable insights and clear, impactful messages. Evaluations that assess relevance, effectiveness, and sustainability based on the conditions at the time of design can produce accurate reflections of past actions but offer limited guidance for future decision-making. 

When evaluation processes rely solely on historical benchmarks, they risk overlooking emerging trends and future challenges. For example, a program designed to improve crop yields based on a specific climate scenario may become less relevant if climate patterns shift unexpectedly. Similarly, a project assessed as sustainable under current conditions might prove vulnerable under future stressors. This gap underscores the need for evaluation methodologies that are forward-looking and capable of engaging with plausible futures.
 

Vairous organizations already embed foresight into their respective practices:
- The World Food Programme (WFP) has integrated foresight-informed approaches into its Anticipatory Action programs, enabling more proactive responses to food crises.
- Organizations such as GEF, CGIAR are exploring how to better assess long-term resilience and systemic impacts in their environmental and agricultural investments. 
- FAO has recently published a report that aims to inspire strategic actions to transform agrifood systems into sustainable, resilient, and inclusive ones. This report ( accessible here: https://www.fao.org/global-perspectives-studies/fofa/en/) explores three different scenarios for the future of food and agriculture, based on alternative trends for key drivers, such as income growth and distribution, population growth, technical progress in agriculture, and climate change.

Strategic foresight should be based on a suite of accessible tools and approaches to address this challenge. While various tools and methods have been developed, practical guidance on their applicability remains limited. Many evaluators lack training in foresight methods. To utilize the full potential of foresight in evaluation, several steps are essential:
 

• Capacity Building: Training evaluators in foresight methods and fostering collaboration between foresight and evaluation practitioners.
• Practical Frameworks: Developing accessible guidance, case studies, and toolkits tailored to the realities of food security, environmental, and agricultural programs.
• Institutional Change: Encouraging organizations to incorporate foresight into their evaluation mandates and to allocate resources for ongoing learning and adaptation.
   

In an era of unprecedented change, evaluation must evolve from a retrospective mirror to a forward-looking compass. Integrating foresight methods into evaluation processes can enhance relevance, sustainability, and systemic impact assessments, ultimately supporting programs that are resilient and adaptable in the face of uncertainty.

Looking forward to further discussions and shared learning on this important topic.

Best regards,

Serdar Bayryyev, Senior Evaluation Officer

Food and Agriculture Organization 
 

Good morning colleagues, and thank you for launching this very timely discussion.

I’d like to share a recent experience from the FAO Office of Evaluation where we explicitly drew on foresight principles in the design and conduct of an evaluation.

In evaluating FAO’s work on One Health, we began with a familiar retrospective lens: how did the approach evolve, and what did FAO contribute? This analysis showed a strong trajectory—leadership over 20 years, particularly in animal health, zoonotic disease control, biosecurity, and more recently antimicrobial resistance (AMR) and pandemic preparedness.

But we quickly ran into a temporal mismatch.

One Health is not a stable field. It is being reshaped by climate change, biodiversity loss, land-use pressures, AMR, and broader food system transformation. Evaluating performance against past conditions risks producing findings that are valid—but less useful for navigating what comes next.

So the question shifted: not just did FAO perform well? but is its approach fit for the futures now emerging?

That’s where a foresight lens—informally, thinking in terms of emerging risks, system shifts, and plausible futures—added value.

It helped us reinterpret a central tension. FAO’s strengths—deep expertise in animal health, strong country platforms, and operational experience—are also its path dependencies. While FAO has adopted a broader, more holistic definition of One Health, implementation still often appears animal health-centred, with ecosystem and systems dimensions less consistently integrated.

From a forward-looking perspective, this matters. Future One Health challenges are likely to be more interconnected, not less. They will require deeper integration across sectors (animals, plants, environment, food systems) and stronger cross-sectoral coordination at country level.

One takeaway for me is that foresight can enter evaluation through existing criteria:

• Relevance becomes future fitness
• Sustainability becomes resilience under change
• Coherence becomes the ability to work across systems

Retrospective evaluation tells us how we got here. A future-informed lens helps us ask whether we’re ready for what’s next.

I’d be very interested to hear how others have approached this—have you found practical ways to bring even light-touch foresight into evaluation design or interpretation?

Role of changing perceptions on sustainability over time

Good morning / hello to everyone and thank you for starting this interesting discussion round, Steven!

I'm not an evaluator, but worked in M&E / MEL / MEAL / MERL in different contexts over the past 10+ years and saw the challenges of mostly backward looking evaluation way to many times. I've not many experiences with foresight evaluations so here just some general thoughts:

What makes an evaluation foreward looking? For me, in a way it's creating awareness of the L in MEL / MEAL / MERL. Why do we do an evaluation? Way to often I saw that for project teams it's just a troublesome check box to tick off to please donors / project requirements. The evaluation is done in whatever way and then put aside. It was often hard work to raise awareness to the importance and potential of evaluation - to uncover what worked well, what lead to actual positive change, what didn't work and why and what even might have brough negative change to then use these Learnings for better and improved projects / policies / strategies / measures etc.

I'm wondering, if outcome harvesting and mapping is one way to increase the awareness of the potential of evaluations, as many stakeholders involved get more actively involved in several reflection rounds around outcome mapping and outcome harvesting - which then ideally triggers an important reflection process on what might work why, what has worked, what didn't, what was unexpectedly positive or negative and can then in turn be used for an improved follow-up process.

So in a way aren't good tools, arguments and practices to really focus on the learning aspect of evaluations and engage with all involved key stakeholders in reflection processes as much as possible greatly support the evaluation foresight? I hope I'm not completely off track here and am very much looking forward to the discussion threads around this topic and learnig from everyone here.

Cheers, 

Conny