Step. and GO! Turn your challenges into concrete actions.

Study of Trends and Best Practices in Learning Through Hands-On Impact Projects

Creation, Development, Management, Evaluation and Scaling Up

Strategic Version — 14 July 2026

Scope and Framing Assumptions

This study focuses on programmes in which people learn by effectively contributing to the creation, development or management of a project aimed at generating a positive social, environmental, educational, scientific, territorial, economic or institutional impact.

It covers in particular:

initial and continuing education;
higher education;
vocational training;
hackathons and innovation laboratories;
incubation and pre-incubation;
non-profit, scientific, entrepreneurial and citizen-led projects;
career transition programmes;
leadership and governance development programmes;
programmes led by companies, local authorities, NGOs, foundations or international organizations;
projects conducted in intercultural or multi-stakeholder contexts.

The analysis pays particular attention to possible applications within the ecosystem formed around Objectif Sciences International, Step & Go, KEMENESI, 2% for the Future, Geneva Foundation for the Future and the Global Impact Projects.

The geographical scope is international, with priority given to Europe, Switzerland, France and transnational programmes.

Executive Summary

1. Project-Based Learning Is Entering a New Phase

Project-based learning is no longer used solely as a teaching method intended to make training more active. The most advanced programmes simultaneously make it:

a means of learning;
a professional experience;
a skills assessment system;
a talent identification mechanism;
an innovation laboratory;
a tool for territorial or civic engagement;
a mechanism for creating projects;
a gateway to employment, entrepreneurship or responsibility within a non-profit organization;
an instrument for producing evidence of impact.

This convergence constitutes the most important transformation in the field.

The project is no longer merely the exercise that concludes the training. It is tending to become the infrastructure around which knowledge acquisition, cooperation, assessment, certification, the development of a professional portfolio and relationships with external partners are organized.

Strategic analysis: the sector is evolving from project-based learning, in which the project serves learning, toward a genuine impact project learning system, in which learning, project implementation and the creation of a useful contribution are managed together.

2. The Skills Being Sought Reinforce the Relevance of These Programmes

The World Economic Forum identifies analytical thinking, resilience, flexibility, agility, leadership and social influence among the most sought-after skills. It also reports that 63% of the employers surveyed consider skills gaps to be a major obstacle to their transformation and that 39% of the core skills used at work could change by 2030.

At the same time, the OECD observes the development of so-called skills-first approaches, based more on demonstrated skills than on qualifications alone. According to its recent analyses, approximately 80% of the employers surveyed expect to place significant importance on professional experience in their recruitment between 2025 and 2030, compared with just over 40% for university degrees. This development does not mean the disappearance of qualifications, but it increases the value of verifiable experiences, portfolios, simulations, real projects and authentic assessments.

Impact project-based learning responds particularly well to this development when it makes it possible to demonstrate:

the ability to understand a complex situation;
the relevant mobilization of knowledge;
cooperation across several professions;
the conduct of an investigation;
the formulation of a strategy;
planning and management;
decision-making under uncertainty;
conflict and risk management;
the creation of a usable result;
the assessment of the effects produced;
learning from mistakes.

3. Effectiveness Is Neither Automatic nor Universal

The available meta-analyses and systematic reviews generally conclude that project-based learning has positive effects on learning outcomes, motivation, problem-solving skills and certain higher-order thinking abilities. They also show a strong dependence on the quality of the design, the level of support, prior knowledge, assessment methods and the duration of the programme.

A poorly designed project can generate a great deal of activity without producing much learning. It can also generate:

cognitive overload;
an unequal division of work;
dependence on members who are already experienced;
a succession of meetings without decisions;
an impressive but unusable prototype;
a convincing presentation without in-depth understanding;
the unpaid exploitation of participants’ work;
harm to beneficiaries or partners;
immediate abandonment after the final presentation.

Good practice therefore does not consist of opposing active learning to structured teaching. The best programmes combine:

explicit fundamental knowledge;
an authentic problem;
progressive support;
genuine room for decision-making;
cycles of action and critical feedback;
individual and collective assessment;
a responsible relationship with stakeholders;
continuity after the training phase.

4. A Dual Criterion for Success Must Replace Success Based Solely on the Deliverable

An impact project-based learning programme must assess separately:

educational success: what each participant understands, knows how to do and can transfer;
project success: the relevance, quality, use and sustainability of the result produced;
the quality of the relationship: the usefulness of the process for partners and beneficiaries;
the quality of the impact: the positive and negative changes to which the project contributes.

These dimensions are not interchangeable.

A project may fail technically while producing very significant learning. Conversely, an excellent participant may enable an entire team to deliver a convincing result without the other members having genuinely progressed.

Certification must therefore not be based solely on the quality of the final product. It must incorporate the decisions made, evidence of individual contribution, intermediate versions, feedback received, the ability to justify trade-offs and reflective analysis.

5. Authenticity Has Become a Central Standard

An assessment is considered authentic when it requires skills to be applied in a situation sufficiently close to their real-world use. Recent studies associate this approach with better integration of knowledge, transferable skills and employability, while emphasizing the need for explicit criteria and appropriate support.

In an impact project, authenticity does not arise solely from the use of a real topic. As far as possible, it requires:

an identifiable commissioning organization, beneficiary or partner;
a problem that is not entirely predetermined;
genuine constraints relating to time, resources, law or field conditions;
access to real data or users;
decisions that have consequences;
feedback from people outside the teaching team;
a deliverable that is likely to be used;
clearly defined responsibility toward the actors concerned.

Authenticity must nevertheless be proportionate. Learners must not be entrusted, without safeguards, with decisions that could affect the safety, health, rights or essential resources of vulnerable populations.

6. Artificial Intelligence Is Transforming Assessment Faster Than Teaching

Generative AI facilitates exploratory research, prototyping, writing, data analysis, programming and the production of materials. It can free up time for investigation, human relationships and experimentation. It can also conceal a lack of understanding, produce fabricated sources, disclose confidential data or homogenize solutions.

Emerging research recommends explicit guidance rather than a general ban: prior learning of the fundamentals, disclosure of uses, verification of results, critical analysis of outputs, preservation of records and an oral defense.

AI makes assessment based solely on the final deliverable insufficient. Credible programmes will increasingly need to incorporate:

logbooks;
version histories;
sources consulted;
interview reports;
rejected choices and their justification;
tests conducted;
individual contributions;
disclosure of the AI tools used;
the instructions given to the systems;
human verification;
an oral defense involving critical questioning.

In the European Union, certain uses of AI in educational admission, guidance or assessment may fall within the high-risk categories established by the Artificial Intelligence Act. The CNIL also recommends clear information, the involvement of the data protection officer and particular attention to individual accounts, personal data and privacy settings.

7. The New Standard Is Multidimensional Evidence

The most credible evidence no longer lies in a single grade or a declarative certificate. It is based on a body of evidence:

a skills framework;
successive deliverables;
peer validation;
partner assessments;
observation by trainers;
an oral defense;
a portfolio;
usage results;
monitoring data;
individual progress;
analysis of mistakes;
contribution to project continuity.

Micro-credentials are developing in Europe as a means of recognizing short, specialized and stackable learning experiences. Their value nevertheless depends on the transparency of the intended learning outcomes, the volume of learning, the assessment method, the identity of the issuer, quality assurance and their potential for recognition or accumulation. A proliferation of badges without serious assessment would rapidly diminish their credibility.

8. The Main Competitive Advantage Will Lie in Continuity

Many programmes know how to organize an intensive event, a hackathon or a final presentation. Far fewer know how to:

select the projects to be continued;
transfer responsibilities;
ensure maintenance;
preserve knowledge;
support teams beyond the event;
mobilize the appropriate funders;
integrate results into an existing organization;
measure effects after six, twelve or twenty-four months.

The transition from an event to a continuum therefore constitutes a major strategic opportunity.

A relevant architecture may include:

an orientation and fundamentals-acquisition phase;
a hackathon or intensive workshop lasting three to ten days;
a six- to twelve-week project studio;
a three- to six-month pre-incubation phase;
support for governance and financing;
practitioner certification;
mentor or trainer certification;
longitudinal monitoring of projects and individuals.

This architecture closely corresponds to the assets already present within the OSI, Step & Go, KEMENESI, 2% for the Future and AGILE ecosystem.

9. The Most Robust Economic Model Is Hybrid

A model that depends exclusively on fees paid by participants risks excluding the groups that would benefit most from the programme. A model that depends solely on subsidies risks remaining unstable and encouraging a project-call logic rather than a quality-driven approach.

The most robust model generally combines:

training or certification fees;
scholarships and public funding;
employer contributions;
contracts relating to real challenges;
partnerships with local authorities or international organizations;
philanthropic funding;
membership fees;
training-of-trainers services;
methodological licences;
income related to post-programme support;
potentially, a regulated share in the value created.

Success-based remuneration, equity stakes or royalties must be separated from educational and certification decisions. Otherwise, the programme could favour the most profitable projects at the expense of the best learning experiences or the least financially viable social needs.

10. Equity and Safety Must Be Designed From the Outset

Access to an intensive project often requires available time, digital equipment, mobility, language proficiency, social confidence and sometimes the ability to work without remuneration. Programmes can therefore reinforce inequalities while presenting themselves as inclusive.

The necessary mechanisms include in particular:

scholarships or allowances;
coverage of travel and accommodation;
available equipment;
formats accessible to people with disabilities;
diversified roles within teams;
language preparation;
individualized tutoring;
safety rules;
reporting mechanisms;
limits on working hours;
protection against harassment;
clarification of intellectual property;
beneficiary consent;
the right to withdraw;
data protection.

11. Human Factors Are More Decisive Than Tools

Self-determination theory emphasizes the role of autonomy, a sense of competence and belonging in lasting motivation. These factors are particularly relevant to projects in which participants must make decisions, confront uncertainty and cooperate.

Psychological safety facilitates the expression of doubts, requests for help, criticism of an idea and the reporting of mistakes. It does not mean the absence of expectations. On the contrary, it enables higher expectations, because a problem can be made visible before it becomes irreversible. Research on interdisciplinary student teams confirms its importance for cooperation and learning.

Social loafing and unequal contributions increase when responsibilities are unclear and individual contributions are difficult to observe. Repeated peer assessments, named responsibilities, checkpoints and individual evidence reduce this risk.

12. Recommended Strategic Positioning

The opportunity is not to create an additional training programme on project management.

It consists of building an integrated system for learning, creation and evidence through impact projects, capable of connecting:

training;
practice;
certification;
incubation;
governance;
financing;
research;
impact measurement;
professional integration;
a community of practitioners.

The central promise could be formulated as follows:

Learn to create, develop and govern impact projects by contributing to real projects, with graduated responsibilities, verifiable evidence and support through to implementation.

The most distinctive expressions are:

evidence of competence and evidence of impact;
learning by making a useful contribution;
from the educational project to the responsible project;
the right to experiment, the duty to provide evidence;
real projects, graduated responsibilities, verifiable results.

Methodology

1. Sources Used

The research cross-referenced six families of sources:

Conversational corpus and internal documents: programmes, projects, methods, economic models and needs described in the accessible exchanges concerning OSI, Step & Go, KEMENESI, governance certification, the Global Impact Projects, AGILE and 2% for the Future.
Institutional sources: UNESCO, OECD, European Commission, Joint Research Centre of the European Union, ISO, UNDP, CNIL and certification bodies.
Professional sources: World Economic Forum, educational institutions, innovation networks and programme operators.
Academic research: meta-analyses, systematic reviews and recent studies on project-based learning, challenge-based learning, authentic assessment, motivation, psychological safety, cognitive load, collective work and artificial intelligence.
Professional communities and social networks: LinkedIn publications, Reddit discussions and public content used to identify objections, expectations and tensions in the field.

2. Analysis Method

The information was organized according to four levels:

Facts: information directly established by the sources;
Convergences: findings identified across several categories of sources;
Analyses: strategic interpretations resulting from cross-referencing the information;
Hypotheses: plausible proposals requiring validation through investigation, experimentation or additional data.

The trends were classified according to:

their maturity;
the quantity and quality of the evidence;
the number of actors concerned;
their time horizon;
their potential for transformation;
their level of uncertainty.

The reference cases were selected not only for their reputation, but also for the transferability of their mechanisms.

Preliminary Analysis of the Internal Corpus

1. Objectif Sciences International and Step & Go Corpus

The corpus describes a long-standing legacy of learning through action, project-based education, Participatory Science and the creation of impact projects. Depending on the programme, the systems mentioned combine:

immersion;
the management of scientific or citizen-led projects;
multi-day hackathons;
cooperation between participants from different professions;
mentoring;
public presentations;
production of deliverables;
certification;
project support after the training;
the possibility of joining a professional or non-profit network.

The most distinctive elements are:

the existence of real situations;
the relationship with organizations and territories;
the combination of Participatory Science and project creation;
the diversity of ages and professions;
the ambition to create sustainable organizations;
the possibility of progressing toward governance responsibilities.

Internal tension to be resolved: the figures for former participants or sessions appear under different scopes and dates in the exchanges. They will need to be standardized by distinguishing unique participants, registrations, sessions, countries, projects created and projects still active.

Priority need: transform a rich but distributed body of experience into a common framework, a clear pathway and a comparable evidence system.

2. KEMENESI Corpus

KEMENESI is envisioned as a hybrid programme combining vocational training, case management, coaching, project-based learning, professional integration and incubation.

The structuring elements include:

limited cohort sizes;
individualized support;
the creation of a portfolio;
stackable modules;
projects aligned with territorial challenges;
partnerships with regional employment offices, institutions and incubators;
positive outcome objectives;
monitoring of project continuity.

This project addresses an area insufficiently covered by many programmes: the need to connect project-based education with social security, individual pace, professional transition and genuine access to opportunities.

Main tension: reconciling administrative requirements and expected employment outcomes with the time required for experimentation, deep learning and project maturation.

3. Governance and Strategic Leadership Certification Corpus

The proposed programme uses real governance situations as a learning environment. It distinguishes several levels of progression, ranging from discovery to practice and then to the ability to train or support other people.

This model is particularly relevant because governance cannot be learned entirely through theoretical content. It requires:

observing stakeholder dynamics;
analysing imperfect decisions;
managing conflicts;
understanding responsibilities;
experiencing deliberation;
the ability to act without having all the information;
learning from consequences.

Point requiring vigilance: participants must not be placed within real governing bodies without a mandate, confidentiality provisions, supervision and explicit limits of responsibility.

4. Geneva Foundation for the Future and AGILE Corpus

AGILE combines project analysis, due diligence, support, selection and preparation for financing. This architecture makes it possible to turn learning into an intelligence system for projects and talent.

It opens up a particularly innovative path: participants are not assessed solely through an exercise, but observed in their ability to analyse, improve and secure real projects.

Structural risk: if the same organization trains, assesses, selects, finances and receives success-based remuneration, it must establish a separation of functions, disclose conflicts of interest and provide appeal mechanisms.

5. Cross-Cutting Synthesis of the Internal Corpus

What the corpus states clearly:

learning through action lies at the heart of the identity of the programmes;
real projects provide a richer learning environment than fictional cases alone;
professional and cultural diversity is regarded as a resource;
projects must aim to achieve social, scientific, ecological or economic usefulness;
human support is decisive;
participants must be able to continue taking action after the training;
governance, financing and impact measurement are inseparable from project creation.

What the corpus suggests:

the different offerings could become stages of the same pathway;
certification could be based on a cross-cutting portfolio;
projects could simultaneously serve training, recruitment, incubation and cross-sector cooperation;
the alumni community could become a network of mentors, experts, partners and commissioning organizations;
the Global Impact Projects could constitute a portfolio of learning situations;
KEMENESI could provide an inclusive and professionally oriented version of the model;
AGILE could become the layer for selection, evidence and preparation for financing.

What has not yet been addressed sufficiently:

the common skills framework;
the distinction between programme impact and project impact;
the full cost of a participant and a project;
intellectual property rules;
protection of beneficiaries;
governance of artificial intelligence;
procedures for challenging an assessment;
continuity and maintenance mechanisms;
external recognition of certifications;
comparability of results;
assessment after six, twelve and twenty-four months;
the sustainability of the workload required of mentors;
prevention of burnout and overload;
protection of sensitive data.

Central strategic question:

How can several strong learning-through-action experiences be transformed into a coherent, recognized, measurable and economically sustainable infrastructure without losing their capacity for experimentation, their diversity and their human foundations?

Detailed Outline of the Full Study

Definition and delimitation of impact project-based learning.
State of academic research and level of evidence.
Evolution of skills, employment and certification systems.
Mapping of related educational methods.
Long-term trends, emerging trends, weak signals and fads.
Educational and operational architecture.
Technologies, artificial intelligence and data governance.
Project management, quality, risks and standards.
Measurement of learning and measurement of impact.
Psychology of motivation, autonomy and engagement.
Sociology of teams, power relations and inequalities.
Anthropology of projects, professional cultures and narratives.
Governance and multi-stakeholder cooperation.
Economic models and financing.
International benchmarking.
Best practices and poor practices.
Stakeholder mapping.
Audience segmentation and personas.
Narratives and positioning.
Risks, controversies and prevention mechanisms.
Foresight over six months, three years and ten years.
Strategic recommendations.
Operational implementation plan.
Indicators, dashboard and longitudinal assessment protocol.

Overview of the Subject

1. Operational Definition

Learning through hands-on impact projects is an approach in which people acquire, mobilize and demonstrate skills by contributing to a real or sufficiently authentic project aimed at achieving an explicit positive change.

This definition includes five requirements:

Intentional learning: the programme pursues explicit skills objectives.
Real action: participants make decisions, produce and test.
Useful contribution: the project responds to an identifiable situation or need.
Responsibility: consequences, limitations and stakeholders are taken into account.
Reflexivity: the experience is analysed in order to make the learning transferable.

Without learning objectives, it is primarily volunteering, employment or entrepreneurship.

Without real action, it is mainly a course or a case study.

Without a useful contribution, it is an educational exercise.

Without responsibility, the programme may produce a staged appearance of impact.

Without reflexivity, the experience may remain rich but difficult to transfer.

2. Related Methods

Approach	Central focus	Distinctive feature	Frequent limitation
Project-based learning	Production of an outcome	Project structuring the pathway	Project sometimes fictional or overly prescribed
Problem-based learning	Resolution of a complex question	Investigation and reasoning	Operational outcome not required
Challenge-based learning	Response to an open-ended issue	Real challenge and external partners	Risk of solutionism
Service-learning	Service provided to a community	Reciprocity between learning and service	Relationship sometimes asymmetrical
Participatory Science	Production of knowledge with non-professionals	Scientific and civic contribution	Variable recognition and quality
Design thinking	Understanding uses and prototyping	User-centred approach and iteration	Improper reduction to a few workshops
Hackathon	Intensive, time-limited production	Rapid mobilization and prototyping	Continuity often weak
Incubation	Development of a project or organization	Support toward viability	Individual learning sometimes secondary
Practice enterprise	Simulation or supervised commercial activity	Professionalization	Real impact sometimes limited
Action learning	Collective resolution of professional situations	Reflection on action	May remain internal to the organization
Action research	Production of knowledge by transforming a situation	Research-action cycle	Long time frame and methodological demands
Workplace-based training	Skill acquired through activity	Immediate proximity to the profession	Dependence on the quality of the work environment

Innovation does not lie in the exclusive choice of one of these approaches, but in their coherent combination according to the maturity of the participants and the project.

3. The Four Elements to Be Managed Simultaneously

The participant

They must understand, act, progress and be able to demonstrate their contribution.

The team

It must cooperate, make decisions, allocate responsibilities and resolve conflicts.

The project

It must become more relevant, more feasible, safer or more sustainable.

The ecosystem

It must receive real value and not be reduced to a practice ground.

Most programmes excel in one or two of these areas, but rarely in all four.

Major Trends

Table 2 — Identified Trends

Identified trends
Major transformations, emerging trends, weak signals and fads
Trend	Type	Description	Evidence or indicators	Actors concerned	Opportunities	Risks	Time horizon	Confidence
Shift from qualifications to evidence of skills	Major	Qualifications remain important, but demonstrated experience, portfolios and verifiable skills are gaining ground	Skills-first policies, micro-credentials and skills-based recruitment	Employers, training organizations, learners	Valuing real projects and portfolios	Fragmentation of badges and inequalities in readability	1 to 5 years	High
Widespread adoption of authentic projects	Major	Real situations are gradually replacing entirely fictional exercises	Development of authentic assessment and external partnerships	Schools, universities, companies, NGOs	Motivation, employability and transfer	Excessive exposure to real risks	Immediate	High
Projects as a talent-selection system	Emerging	Behaviours observed in action complement CVs and interviews	Project platforms, capstones and employer programmes	Employers, incubators, funders	More precise identification of abilities	Observation bias and unpaid work	1 to 3 years	Medium to high
Assessment of the process rather than only the outcome	Major, accelerated by AI	Records, decisions and justifications are becoming central	Research on authentic assessment and AI	Trainers, certifiers, learners	Fairer and more robust assessment	Excessive surveillance	Immediate	High
Artificial intelligence as a project copilot	Emerging	AI supports research, analysis, prototyping and documentation	Academic and professional experiments	All teams	Productivity and accessibility	Dependence, errors, confidentiality and homogenization	Immediate to 3 years	High
Stackable micro-credentials	Major	Recognition of shorter and more specialized learning units	European Recommendation and UNESCO work	Continuing education, universities, employers	Modular and lifelong pathways	Inflation of badges without value	1 to 5 years	High
Convergence of training, incubation and professional integration	Emerging	A single pathway leads from learning to project creation or employment	Team academies, studios, educational incubators	Schools, local authorities, incubators	Reduction of pathway disruptions	Confusion of missions and conflicts of interest	2 to 5 years	High
Continuity beyond the hackathon	Emerging	Actors seek to transform prototypes into monitored projects	Incubation programmes and recurring criticism of short events	Organizers, funders, project leaders	Lasting value and retention	High support costs	1 to 3 years	High
Mission-oriented projects and systemic transformations	Major	Projects are increasingly connected to ecological, social and digital transitions	UNESCO ESD, GreenComp, challenge-based programmes	Institutions, territories, companies	Strategic alignment and partnerships	Objectives that are too broad or symbolic	1 to 10 years	High
Interdisciplinary learning	Major	Complex problems require several professions and disciplines	University benchmarks and challenge-based programmes	Universities, companies, NGOs	Richer solutions and mutual understanding	Language difficulties and power relations	Immediate	High
Professionalization of mentors and facilitators	Emerging	Quality depends on skills distinct from technical expertise	Development of training for coaches and teachers	Training organizations, professional networks	Reproducible quality	Bureaucratization or excessive standardization	2 to 5 years	Medium to high
Integrated learning and impact dashboards	Weak signal	Participant, project and impact data are connected	Learning analytics, portfolio platforms and impact management	Operators, funders, researchers	Longitudinal management	Surveillance, bias and complexity	2 to 5 years	Medium
Digital marketplaces for real challenges	Emerging	Platforms connect organizations, teachers and learners	Development of Riipen and comparable programmes	Companies, schools, students	Access to a large number of projects	Variable quality and unpaid work	1 to 5 years	Medium to high
Certification through a portfolio of evidence	Emerging	The documented portfolio becomes part of certification	Micro-credentials, verifiable badges and authentic assessment	Certifiers, employers	Rich and transferable evidence	Documentation burden and fraud	2 to 5 years	High
Learning through remote international projects	Major	Distributed teams are becoming standard	Collaborative tools and university networks	International programmes	Diversity and reduced travel	Digital divide and weak cohesion	Immediate	High
Full personalization through AI	Possible fad	Promise of fully adapted and automated pathways	Communications from technology providers	EdTech, training organizations	Partial individualized support	Dehumanization and opacity	1 to 5 years	Medium
“Changing the world in 48 hours”	Fad	Spectacular narrative surrounding hackathons	Event communications	Organizers and sponsors	Rapid mobilization	Impact washing and disappointment	Immediate	High
Educational metaverse as the dominant infrastructure	Possible fad	Promise of universal immersion	Technology investment cycles	EdTech	Usefulness for certain simulations	Costs, equipment and lack of genuine need	3 to 10 years	Low to medium
Regenerative projects and consideration of negative effects	Emerging	Shift from reducing harm to restoring systems	Impact frameworks and regenerative movements	Territories, companies, NGOs	Advanced positioning	Poorly defined concepts and excessive claims	2 to 10 years	Medium

Analysis of Main Trends

1. From One-Off Projects to a Project Portfolio

The most advanced organizations no longer regard each project as an isolated experience. They build a portfolio that makes it possible to:

distribute risks;
offer several levels of difficulty;
enable participants to move between projects;
capitalize on methods;
pool partners;
ensure continuity;
produce comparable data;
create progression pathways.

For the ecosystem studied, the Global Impact Projects could provide projects of various levels and types: investigation, design, territorial experimentation, governance, partner search, communication, impact measurement or preparation for financing.

2. From the Final Presentation to an Evidence Pathway

The final pitch remains useful, but it favours people who are comfortable speaking and projects that are easy to explain. It is not sufficient to establish the depth of the analysis or the quality of execution.

The evidence pathway must begin as soon as participants enter the programme:

initial assessment;
problem formulation;
stakeholder mapping;
research and sources;
decisions and hypotheses;
tests;
negative results;
user feedback;
modifications;
individual contribution;
learning;
continuity plan.

3. From Volunteer Mentor to the Profession of Facilitator

Mentoring is often presented as a free and abundant resource. In reality, a technical expert is not automatically capable of:

helping without imposing;
distinguishing advice from decision-making;
managing a team in difficulty;
creating psychological safety;
assessing fairly;
reporting an ethical risk;
working with vulnerable groups;
supporting reflection without doing the work in the participant’s place.

The professionalization of the mentor’s role therefore constitutes an important trend. It can take the form of short training, supervision, a code of conduct, a community of practice and progressive certification.

Technical Analysis

1. Recommended Operational Architecture

A complete system can be structured in nine stages.

Stage 1 — Challenge Qualification

origin of the problem;
legitimacy of the partner;
people affected;
available data;
risks;
level of maturity;
time required;
compatibility with the educational objectives;
possibility of using the result.

Stage 2 — Participant Assessment

existing skills;
knowledge needs;
availability;
accessibility constraints;
experience of collective work;
motivations;
exposure to risks;
professional objectives.

Stage 3 — Acquisition of the Fundamentals

Structured inputs must precede or accompany action:

project management;
understanding of the field;
documentary research;
interviewing and observation;
ethics;
data;
safety;
impact;
cooperation;
artificial intelligence.

Cognitive load must be adjusted to the level of expertise. Novices need more models, examples, short sequences and support, whereas experienced participants can benefit from more open-ended problems.

Stage 4 — Team Formation

The composition must seek genuine complementarity without creating an implicit hierarchy between professions.

Each team establishes:

a mandate;
initial roles;
decision-making rules;
availability;
a disagreement procedure;
communication rules;
a policy on the use of AI;
a confidentiality commitment;
a task-redistribution mechanism.

Stage 5 — Project Framing

The minimum deliverables are:

problem formulation;
scope;
stakeholders;
theory of change;
risks;
hypotheses;
resources;
indicators;
timetable;
decisions reserved for the partner;
decisions entrusted to the team.

Stage 6 — Investigation and Prototyping Cycles

Each cycle includes:

a hypothesis;
an action;
an observation;
feedback;
a decision;
a risk update;
a learning record.

Stage 7 — Stage-Gate Reviews

Decisions to continue, modify, suspend or stop a project must be explicit.

The criteria concern:

learning;
relevance;
feasibility;
safety;
partner commitment;
evidence of use;
sustainability;
team capacity.

Stage 8 — Transfer and Continuity

The programme specifies:

who holds the documents;
who maintains the result;
who holds the rights;
who responds to users;
which resources remain necessary;
what support is planned;
which elements must be archived.

Stage 9 — Longitudinal Monitoring

Monitoring after six, twelve and twenty-four months measures separately:

participants’ trajectories;
use of the deliverable;
project continuity;
development of partnerships;
expected and unexpected effects;
new needs.

2. Technological Architecture

A reasonable infrastructure comprises five layers.

Educational Layer

content;
activities;
skills framework;
assessments;
feedback;
progression.

Project Layer

tasks;
milestones;
decisions;
risks;
deliverables;
versions;
responsibilities.

Relational Layer

directory;
partners;
mentors;
beneficiaries;
history of interactions;
consent.

Evidence and Portfolio Layer

individual outputs;
contributions;
validations;
certificates;
results;
external assessments.

Data and Impact Layer

theory of change;
indicators;
baseline data;
monitoring;
negative effects;
contribution hypotheses.

Interoperability must be prioritized. The programme must avoid a situation in which a single provider holds the portfolios, educational data and project histories without any possibility of export.

3. Artificial Intelligence

Useful applications include:

exploratory research;
initial synthesis;
transcription;
translation;
document classification;
generation of hypotheses;
preparation of interviews;
programming assistance;
risk scenarios;
data analysis;
simulation of objections;
linguistic accessibility.

Applications requiring increased vigilance include:

automated grading;
participant selection;
psychological assessment;
career guidance;
processing data concerning minors;
interpretation of medical or social data;
generation of testimonials;
representation of absent beneficiaries;
automatic drafting of a theory of change without an investigation;
financing decisions.

Each project should maintain an artificial intelligence usage register including:

tool used;
objective;
data transmitted;
output retained;
verifications;
errors identified;
person responsible for validation.

4. Data and Indicators

Five dashboards must be distinguished.

Learning Dashboard

progression by skill;
autonomy;
quality of analysis;
capacity for cooperation;
transfer;
reflexivity.

Project Dashboard

progress;
quality;
risks;
budget;
decisions;
partner commitment.

Use Dashboard

users;
frequency;
satisfaction;
adoption;
abandonment;
requests for improvement.

Impact Dashboard

changes sought;
outcome indicators;
negative effects;
project contribution;
external factors;
uncertainty.

Ecosystem Dashboard

active partners;
mentors;
projects continued;
new collaborations;
financing mobilized;
former participants engaged.

The Impact Management Platform and the SDG Impact Standards emphasize integrating impact into strategy, governance and decision-making rather than merely producing a final report.

5. Standards and Reference Frameworks

A programme-specific framework can draw on:

EntreComp for the capacity to transform ideas into value;
GreenComp for sustainability competences;
LifeComp for personal, social and learning competences;
DigComp for digital competences;
UNESCO AI competency frameworks for teachers and learners;
ISO 21502 for project management;
ISO 37000 for the governance of organizations;
OECD-DAC criteria for evaluation;
SDG Impact Standards for impact integration;
Qualiopi in France when the relevant funding is mobilized;
eduQua in Switzerland, depending on the scope of the training programmes.

The EntreComp, GreenComp, LifeComp and DigComp frameworks make a coherent architecture possible without requiring the complete creation of an isolated framework.

Qualiopi is required in France for providers wishing to access certain public or pooled funding. eduQua is a recognized quality label in the field of continuing education in Switzerland.

Analysis Through the Humanities and Social Sciences

1. Psychology of Motivation

Projects generate engagement when they combine:

genuine but supervised autonomy;
an appropriate challenge;
a sense of progress;
visible usefulness;
belonging;
recognition;
the possibility of making decisions.

Merely rhetorical autonomy is demotivating. Asking participants to be autonomous while predetermining the problem, method, result and decisions produces autonomy of execution, not autonomy of judgment.

Conversely, premature autonomy exposes novices to disorientation. The degree of freedom must therefore increase with competence and the capacity to assess consequences.

2. Sociology of Teams

Interdisciplinary teams are shaped by power relations:

technical expertise versus lived experience;
institutional status versus field knowledge;
language proficiency versus actual competence;
ease of speaking versus less visible contributions;
voluntary availability versus family or economic constraints;
age and seniority;
geographical centre versus periphery;
funder versus beneficiary.

Diversity does not automatically produce a better solution. It initially produces more differences, misunderstandings and coordination costs. It becomes productive when the programme organizes:

a common language;
rules for speaking;
recognition of different forms of knowledge;
decision-making methods;
mediation;
clarification time;
a right to disagree.

3. Anthropology of the Project

The word project does not have the same meaning for everyone.

For a company, it may refer to a temporary operation with a budget.

For an NGO, it may refer to a programme dependent on a donor.

For a community, it may be perceived as an external intervention destined to disappear.

For an entrepreneur, it may be linked to the creation of an organization.

For a participant, it may be a learning experience or an employment opportunity.

These representations must be made explicit. A large proportion of conflicts arise less from declared objectives than from implicit time frames, obligations and expectations.

4. Political Science and Governance

An impact project distributes resources, visibility, responsibilities and the power to define. It is therefore political, even when presented as purely technical.

The essential questions are:

who defines the problem;
who selects the participants;
who allocates the resources;
who speaks on behalf of the beneficiaries;
who owns the data;
who can stop the project;
who receives recognition;
who bears the consequences;
who assesses success.

Multi-stakeholder governance must avoid two pitfalls:

symbolic participation without real power;
multiplying consultations without making decisions.

5. Behavioural Economics

Participants make trade-offs between:

future usefulness and immediate effort;
exploration and security;
cooperation and individual visibility;
learning and production;
transparency and fear of assessment;
commitment to the project and other obligations.

The most useful mechanisms are not artificial incentives, but:

intermediate deadlines;
reasonable public commitments;
visible progress;
identifiable responsibilities;
rapid recognition;
the possibility of choice;
a limited cost of withdrawal;
clear consequences for non-contribution.

6. Communication and Semiotics

The dominant vocabulary in the sector frequently uses:

innovation;
impact;
transformation;
challenge;
community;
experimentation;
collaboration;
leadership;
future skills;
regeneration;
ecosystem.

These terms lose credibility when they are not connected to observable mechanisms.

For example:

collaboration must specify who decides with whom;
impact must specify what changes for whom;
innovation must specify what differs from what already exists;
community must specify members’ rights and obligations;
regeneration must specify what is being restored;
leadership must specify the responsibilities assumed.

7. Ethics

The guiding principle should be:

No educational benefit justifies making partners or beneficiaries bear a risk that they have not understood or accepted.

Participants must be trained in:

listening;
confidentiality;
data protection;
intellectual property;
prevention of discrimination;
safety;
scientific integrity;
disclosure of conflicts of interest;
uncertainty;
the distinction between evidence and promise.

Analysis of Social Networks and Professional Communities

The public content observed must be regarded as qualitative signals, not as representative surveys.

Recurring Positive Narratives:

learning by working on something that matters;
building a portfolio;
meeting professionals;
rediscovering a sense of purpose;
developing skills that traditional courses assess poorly;
belonging to a community;
gaining experience despite a lack of previous employment.

Recurring Objections:

“One person does all the work.”
“The instructions remain vague.”
“The project replaces the teaching of fundamentals.”
“The collective grade penalizes contributors.”
“Partners obtain unpaid work.”
“The hackathon produces nothing after the event.”
“AI can produce the entire deliverable.”
“The project is too broad for the time available.”
“Students do not have access to decision-makers or users.”

These criticisms appear regularly in communities of teachers and students. They converge around three expectations: more structure, visible individual contributions and a problem genuinely scaled to the available resources.

Effective Communication Formats:

a before-and-after account of a project;
joint testimony from the participant and the partner;
demonstration of a prototype;
progress log;
analysis of a useful failure;
presentation of a difficult decision;
annotated portfolio;
monitoring data after six or twelve months;
a behind-the-scenes visit;
explanation of roles and methods.

Less Credible Formats:

accumulation of photographs of people sitting around tables;
general statements about “impact”;
participation figures without results;
prizes or competitions without follow-up;
decontextualized satisfaction quotes;
promises of very rapid transformation;
content presenting every project as a success.

Benchmark and Inspiring Cases

The following cases illustrate different mechanisms. They should not be copied in full, but analysed as transferable building blocks.

Table 4 — Inspiring Cases

Inspiring cases
Reference programmes and organizations
Case	Country	Actor	Description	Known results or characteristics	Transferable elements	Source
Aalborg PBL	Denmark	Aalborg University	Institutional model based on real problems and collaborative work	PBL broadly structures the curricula and connects theory, investigation and projects	Institution-wide integration, supported groups, authentic problems
SCOPE	United States	Olin College of Engineering	Year-long projects conducted with companies or organizations	Students frame, plan, design and iterate on a real problem	Long-term continuity, clearly committed partner, professional deliverable
D-Lab	United States and international	MIT	Teaching, participatory design and field experiences related to development	More than twenty courses and several thousand students over the course of the programme	Design with communities, frugal prototyping, ethical reflection
EPICS	United States and international	Purdue University	Service-learning projects conducted with community organizations	Development of real systems by teams that may continue over several semesters	Continuity, community partner and maintenance
TU/e innovation Space	Netherlands	Eindhoven University of Technology	Challenge-based learning with companies and public partners	Real challenges, interdisciplinary teams and a prototyping environment	Challenge-qualification office and intermediary space between the university and the ecosystem
Product Development Project	Finland	Aalto Design Factory	Product-development projects conducted with industrial partners	From framing and investigation through to functional solutions	Prototyping as a common language and a complete pathway
Tiimiakatemia	Finland	JAMK University of Applied Sciences	Learning in team companies through real commercial projects	Degree programme based on action, dialogue and coaching	Lasting team, role of the coach, real economic activity
Kaospilot	Denmark	Kaospilot	Training in leadership, entrepreneurship and change through real projects	Experiential education and work on complex problems	Design of learning environments and personal development
Minerva Project	United States and international	Minerva University	Active seminars connected to civic and professional projects in several cities	Connection between concepts, transfer and territorial experiences	Feedback loop between conceptual teaching and real situations
Enactus	International	Enactus	Network of student teams developing entrepreneurial projects with a social purpose	Strong international reach and project competitions	Network of teams, company engagement and visibility
Hult Prize	International	Hult Prize Foundation	Competition and support for student startups linked to the Sustainable Development Goals	Reported presence in more than 130 countries and a substantial final prize	International funnel and mass mobilization
UNLEASH Innovation Lab	International	UNLEASH	Intensive laboratories bringing together young professionals around the SDGs	Rapid production of numerous solutions during international gatherings	International recruitment, intensive method and community
Social Impact Award	Europe and international	SIA	Awareness-raising, training and incubation for young social entrepreneurs	Presence in more than twenty-five countries according to the organization	Local pathway combining education, community and incubation
Mondragon Team Academy	Spain and international	Mondragon University	Entrepreneurial learning in teams and creation of activities	Development of skills through experimentation and cooperation	Team companies and an international network
Gold Standard PBL	United States and international	PBLWorks	Quality framework for project-based education	Challenging problem, inquiry, authenticity, choice, reflection, critique and public product	Simple educational quality-control framework
Riipen	Canada and international	Riipen	Platform connecting teachers, students and employers through projects	Remote, hybrid or paid projects, depending on the programme	Challenge marketplace and relationship management
Participatory Science and Step & Go	France, Switzerland and international	OSI and partners	Immersions, hackathons, training through scientific projects and impact projects	Internal corpus reporting numerous sessions, participants and projects over several decades	Diversity of participants, Participatory Science, progression toward governance and community continuity	Internal conversational corpus
KEMENESI Impact Training Center	Switzerland	Annemarie Kemenesi Foundation and proposed partners	Inclusive vocational training, coaching, real projects and incubation	Pilot project targeting small cohorts and monitoring professional integration	Combination of case management, learning, portfolio and territorial ecosystem	Internal conversational corpus
Governance and leadership certification	International	Training for Development and the 2% for the Future network	Learning in real governance situations, progressing toward peer training	Proposed eighteen-month architecture and several levels of responsibility	Governance as practice, mentoring, practitioner and trainer certification	Internal conversational corpus
AGILE	International	Geneva Foundation for the Future	Analysis and improvement of impact projects linked to due diligence and financing	Structured project-qualification and preparation cycle	Real projects as a basis for learning, selection and evidence	Internal conversational corpus

Cross-Cutting Lessons From the Benchmark

The most robust cases combine several of the following characteristics:

the programme is not based on a single event;
partners are selected and prepared;
the problem is real but appropriately scaled;
the teacher’s role becomes that of a designer, expert and facilitator;
the fundamentals do not disappear;
teams produce intermediate evidence;
reflection is integrated into action;
the result is presented externally;
projects can continue;
the institution has permanent infrastructure;
former participants return as mentors or partners.

The most highly publicized models are not necessarily those that produce the deepest learning. High-visibility competitions often excel in mobilization and selection, but may favour pitching, competition and projects that are easy to explain. More discreet models, conducted over several months with a stable partner, are generally better suited to in-depth learning and real use.

Best Practices

Table 3 — Best Practices

Best practices
Conditions for effectiveness and transferability
Best practice	Field	Example	Conditions for success	Benefits	Limitations	Transferability	Priority
Use a real but appropriately scaled problem	Design	EPICS, SCOPE, TU/e	Prior qualification, scope compatible with the duration	Authenticity and usefulness	Risk of excessive complexity	Very high	Critical
Combine fundamentals and action	Education	Aalborg, Minerva	Inputs provided when they become useful	Reduced overload and improved transfer	Requires careful preparation	Very high	Critical
Define graduated responsibilities	Education and safety	Level-based pathways	Clear progression criteria and supervision	Genuine autonomy without premature exposure	May appear restrictive	Very high	Critical
Separate individual assessment from collective results	Assessment	Portfolio, oral defence and peer assessment	Individual records and known criteria	Reduced social loafing	Assessment workload	Very high	Critical
Formalize the relationship with the partner	Governance	SCOPE, EPICS	Mandate, availability, data, ownership and continuity	Clear expectations and improved use	Negotiation time	Very high	Critical
Involve beneficiaries in defining the problem	Participation	MIT D-Lab	Consent, possible remuneration and real power	Relevance and legitimacy	Slower and more demanding	High	Critical
Organize short cycles of testing and reflection	Method	Aalto, design studios	Explicit hypotheses and access to users	Rapid learning and fewer errors	Tests may sometimes be artificial	Very high	High
Provide a continuity mechanism	Scaling up	EPICS, SIA incubation	Responsible person, resources and transfer decision	Reduces abandonment after the final presentation	Post-programme costs	Very high	Critical
Train and supervise mentors	Human resources	Team Academy, Kaospilot	Framework, code of conduct, observation and community	More consistent quality	Initial investment	Very high	High
Declare and audit the use of AI	Data and assessment	Usage register	Clear policy, approved tools and verification	Transparency and critical learning	Documentation burden	Very high	Critical
Measure learning, project, use and impact separately	Assessment	Multidimensional dashboard	Limited indicators and associated decisions	Avoids simplistic conclusions	Complexity	Very high	Critical
Remunerate or compensate for barriers to access	Inclusion	Scholarships and allowances	Dedicated budget and transparent criteria	Genuine participant diversity	Cost	High	High
Produce a verifiable portfolio	Certification	Micro-credentials and authentic assessment	Evidence, external validation and exportability	Employability and recognition	Risk of excessive documentation	Very high	High
Include a review of negative effects	Impact and ethics	SDG Impact Standards	A culture that allows harm to be reported	Credibility and prevention	May slow communication	Very high	Critical
Establish governance for conflicts of interest	Governance	Independent committee	Separation of training, certification and financing	Trust	Governance costs	Very high	Critical
Monitor results after six, twelve and twenty-four months	Assessment	Longitudinal study	Consent, stable identifiers and follow-ups	Knowledge of lasting effects	Respondent attrition	Very high	High

Poor Practices and Mistakes to Avoid

1. Replacing Fundamental Teaching With an Open-Ended Project

Novices do not always know how to identify what they do not know. A completely open-ended project may favour those who already possess the codes, knowledge or network.

Response: provide a common foundation, targeted resources, worked examples and differentiated support.

2. Using a Real Partner as Mere Scenery

Inviting a partner to the opening and final presentation is not sufficient. Without regular access, teams work on assumptions.

Response: specify the partner’s minimum availability, contacts, response times and decision-making moments.

3. Assessing Only the Final Presentation

This practice strongly rewards communication skills and may conceal the limited contribution of certain members.

Response: combine a portfolio, observations, an individual oral defence, peer assessments and evidence of progress.

4. Confusing a Prototype With Impact

A prototype, application, report or campaign does not constitute impact. These are outputs that may contribute to change.

Response: distinguish activity, output, adoption, outcome and impact.

5. Promising That All Projects Will Continue

Universal continuation is unrealistic and may disperse resources.

Response: announce several legitimate outcomes from the outset: documented closure, transfer, additional experimentation, incubation or integration into an organization.

6. Depending Excessively on Volunteer Mentors

The best mentors are often over-solicited. Their availability becomes unpredictable and quality varies.

Response: define a paid or compensated core, supplement it with volunteers, limit the workload and recognize contributions.

7. Using Impact as an Unverified Marketing Argument

Multiplying references to the SDGs, regeneration or social innovation does not replace analysis.

Response: publish the hypotheses, data, limitations, negative effects and degrees of uncertainty.

8. Leaving AI Rules Implicit

The absence of rules produces significant differences between teams and makes assessment difficult.

Response: define what is authorized, what must be declared, what is prohibited and what is subject to validation.

Stakeholder Analysis

Actor	Interests and expectations	Barriers	Possible contribution	Appropriate value proposition
Participants	Skills, experience, purpose, network, employment or project	Time, cost, fear of failure, lack of confidence	Work, investigation, creativity, lived experience	Acquire skills by making a verifiable contribution
Employers	Skills, recruitment, innovation and engagement	Employees’ time, confidentiality, uncertain results	Challenges, experts, data, recruitment	Observe and develop talent in real situations
NGOs	Solutions, capacities, visibility, partnerships	Lack of time, fear of superficial work	Problems, field access, beneficiaries, expertise	Receive a supervised contribution without becoming a practice ground
Local authorities	Territorial innovation, participation and employment	Public procurement, time frames and responsibility	Challenges, data, locations, financing	Transform territorial needs into monitored learning projects
International organizations	Cooperation, innovation and skills development	Procedures, confidentiality, governance	Expertise, challenges, networks, legitimacy	Test cooperation within a responsible and documented framework
Educational institutions	Success, employability, partnerships	Educational workload, timetable, accreditation	Teachers, students, research	Access an infrastructure of qualified projects and partners
Training organizations	Distinctive offering and certification	Cost of real projects and need for mentors	Educational engineering and monitoring	Deploy a transferable, evidence-based method
Incubators	More mature projects and teams	Uneven quality of applications	Entrepreneurial expertise and network	Receive already documented projects and observed teams
Public funders	Professional integration, skills and territorial development	Difficulty measuring effects	Financing, referrals and data	Connect learning, integration and territorial outcomes
Foundations	Impact, experimentation and systemic change	Attribution and sustainability	Patient grants, network and expertise	Fund reusable infrastructure rather than isolated events
Impact investors	Better-prepared projects and identified risks	Insufficient pipeline and impact washing	Expertise, financing and due diligence	Access projects whose hypotheses and teams have been observed
Researchers	Data, experiments and publications	Consent, comparability and time frame	Assessment methods and independence	Conduct longitudinal research on a multi-site model
Mentors	Transmission, network, purpose and recognition	Time, ambiguity of role and responsibility	Expertise and support	Join a community of trained and recognized practitioners
Beneficiaries and communities	Response to needs, respect and capacity to act	Distrust, consultation fatigue and extraction	Experience, local knowledge and validation	Co-define projects and retain decision-making rights
Regulators and certifiers	Quality, transparency and protection	Innovation that is difficult to classify	Frameworks, controls and recognition	Obtain documented evidence and clear responsibilities
Former participants	Network, progression and opportunities	Loss of momentum and absence of a role	Mentoring, recruitment, projects and governance	Continue learning by becoming a contributor, mentor or commissioning partner

Analysis of Audiences and Behaviours

The following personas are strategic hypotheses to be validated through interviews. Their confidence level is medium.

Persona 1 — The Professional in Career Transition

Profile: 28 to 45 years old, professional experience but little evidence in the field of impact.

Motivations:

rediscover a sense of purpose;
transform experience into transferable skills;
build a portfolio;
meet employers or partners.

Barriers:

cost;
loss of income;
fear of starting again from zero;
doubt regarding the value of the certificate.

Trigger: a real project, a compatible timetable and evidence of professional recognition.

Persona 2 — The Young Graduate Without Demonstrable Experience

Profile: 20 to 30 years old, academic education, difficulty obtaining a first position.

Motivations:

gain experience;
stand out;
receive feedback;
join a network.

Barriers:

fear of working without pay;
lack of confidence;
difficulty choosing a field;
competition fatigue.

Trigger: a clear mission, a mentor, a portfolio and the possibility of receiving a recommendation.

Persona 3 — The Non-Profit Programme Manager

Profile: 35 to 55 years old, operational responsibilities, limited resources.

Motivations:

structure a project;
cooperate more effectively;
obtain tools;
prepare for financing;
develop the team.

Barriers:

lack of time;
jargon;
fear of an overly rigid external method;
confidentiality.

Trigger: support applied directly to their own project.

Persona 4 — The HR, CSR or Innovation Manager

Profile: medium-sized or large organization seeking skills and cross-functional projects.

Motivations:

develop employees;
identify talent;
produce concrete projects;
strengthen engagement.

Barriers:

difficulty freeing up time;
uncertain measurement;
reputational risk;
integration of results.

Trigger: a documented programme, clear governance and comparable results.

Persona 5 — The Holder of Territorial Knowledge

Profile: local association, field professional, citizen or community concerned.

Motivations:

have a need recognized;
gain access to skills;
retain decision-making capacity;
obtain continuity.

Barriers:

consultation fatigue;
fear of dispossession;
institutional language;
absence of feedback after the programme.

Trigger: co-design, compensation, explicit rights and a person responsible for follow-up.

Communication, Narratives and Keywords

Words to Prioritize

real project;
useful contribution;
demonstrated skills;
graduated responsibilities;
support;
evidence;
cooperation;
use;
continuity;
portfolio;
quality;
reflective learning;
verifiable impact;
responsible experimentation;
community of practice.

Words to Use Precisely

innovation;
transformation;
leadership;
ecosystem;
regeneration;
collective intelligence;
co-creation;
systemic;
social entrepreneurship;
impact.

Phrases to Avoid

“Change the world in a few days”;
“Guaranteed impact”;
“Revolutionary solution”;
“Everyone can become an entrepreneur”;
“AI fully personalizes learning”;
“International certification” without demonstrated recognition;
“Global community” when it is merely a contact list;
“Co-constructed project” when beneficiaries have only been consulted.

Recommended Narratives by Audience

For participants:

You do not merely learn how a project should work. You contribute to a real project, take on responsibilities suited to your level and build evidence of what you know how to accomplish.

For employers:

Observe and develop skills in real situations rather than inferring them solely from a CV or an interview.

For field partners:

Your organization is not a classroom case. The challenge, responsibilities, rights over the results and continuity are defined with you.

For funders:

Fund infrastructure that transforms every project into learning, every learning experience into evidence and every piece of evidence into a sustainable capacity to act.

For training institutions:

Access qualified projects, prepared partners, trained mentors and a common skills-assessment system.

Risks, Controversies and Limitations

Table 5 — Risks

Risks
Probability, severity and recommended responses
Risk	Type	Probability	Severity	Warning signs	Prevention	Recommended response
Intensive activity but superficial learning	Educational	High	High	Extensive production, weak justification	Explicit objectives, fundamentals and reflection	Review the scope and assess the reasoning
Unequal contribution or free riding	Human	High	Medium to high	Concentrated tasks, late conflicts	Responsibilities, records and peer assessments	Redistribute tasks and individualize part of the assessment
Overload and burnout	Human	Medium to high	High	Night work, absenteeism, irritability	Working-hour limits, realistic timetable and right to raise an alert	Reduce the scope and suspend non-critical obligations
Harm caused to a beneficiary	Ethical and legal	Low to medium	Very high	Absence of consent or uncontrolled testing	Ethical analysis, supervision and restrictions	Stop the activity, protect the people concerned and document the incident
Impact washing	Reputational	High	High	Confusion between deliverable and change	Theory of change and graduated evidence	Correct the claims and publish the limitations
Unpaid work benefiting a partner	Social and reputational	Medium	High	Significant value captured, limited learning	Contract, proportionality and possible remuneration	Renegotiate or terminate the assignment
Excessive dependence on AI	Educational and technical	High	High	Participants unable to explain the deliverable	Fundamentals, AI register and oral defence	Reassess through an oral or practical examination
Data leakage through an AI tool	Data and legal	Medium	Very high	Use of public tools with sensitive data	Approved tools, anonymization and training	Notification, incident analysis and corrective measures
Bias in assessment	Ethical	Medium	High	Unexplained disparities according to gender, language or profile	Criteria, multiple assessors and audit	Independent review and right of appeal
Inequalities of access	Social	High	High	Homogeneous audience and withdrawals for financial reasons	Scholarships, equipment, accessibility and adapted formats	Adjust recruitment and finance access constraints
Intellectual-property conflict	Legal	Medium	High	Absence of prior agreement	Simple clauses before the project	Mediation and suspension of exploitation
Conflict between certification and financing	Governance	Medium	High	Profitable projects being favoured	Separation of committees and disclosure of interests	Independent assessment
Abandonment after the programme	Operational	High	Medium to high	No person responsible or follow-up budget	Continuity plan before the final presentation	Transfer, archive or explicitly close
Variable mentor quality	Quality	High	High	Contradictory advice and excessive intervention	Training, supervision and assessment	Replace or support the mentor
Dilution during scaling up	Strategic	Medium	High	Deteriorating support ratio	Minimum standards and progressive pilots	Suspend expansion and restore quality
Dependence on a funder	Financial	Medium to high	High	Dominant share held by one donor	Diversification and recurring income	Reduce fixed costs and develop several channels
Political instrumentalization	Political	Low to medium	High	Pressure on results or communication	Pluralistic governance and transparency	Disclose conflicts and protect independence
Inflation of certificates without recognition	Reputational	High	Medium	Proliferation of weakly assessed badges	Framework, quality assurance and employer partners	Reduce the number of credentials and strengthen assessment

Foresight

Short-Term Horizon — Six to Twelve Months

Likely developments include:

multiplication of policies on the use of AI;
growth in oral assessments and process records;
increased interest in portfolios and real projects;
development of micro-credentials;
stronger requirements concerning data and confidentiality;
search for short formats connected to continuity;
tensions between the need for skills and employers’ capacity to support projects.

Priority decisions:

formalize a framework;
publish an AI policy;
design contracts with partners;
define the expected evidence;
select a few pilot projects;
establish longitudinal monitoring.

Medium-Term Horizon — Two to Three Years

Likely developments include:

stacking micro-credentials into longer pathways;
growing recognition of portfolios;
development of challenge marketplaces;
professionalization of mentors;
convergence between training, incubation and recruitment;
automation of part of the documentation;
development of dashboards integrating skills and projects;
emergence of consortia between educational institutions, companies, NGOs and local authorities.

Positioning opportunity:

Become an independent infrastructure capable of qualifying challenges, preparing partners, training mentors, supporting teams, certifying skills and ensuring project continuity.

Long-Term Horizon — Five to Ten Years

Several transformations are plausible:

project portfolios used as a major component of professional recognition;
training pathways composed of experiences distributed across several organizations;
AI agents capable of monitoring decisions, risks and learning;
international recognition of blocks of skills;
networked schools or universities based on territorial and international projects;
increased regulation of impact claims and automated educational decisions;
professional communities in which learning, working, undertaking and transmitting become stages of the same pathway.

Three Scenarios

Scenario 1 — Cautious: the Model Remains a Complementary Offering

Traditional training remains dominant. Impact projects are used as modules, seminars or events.

Consequences:

a limited but identifiable market;
dependence on institutions and sponsors;
weak continuity;
certification that is mainly additional.

Appropriate positioning: specialized operator of high-quality projects and hackathons.

Scenario 2 — Realistic: Creation of a Hybrid Continuum

Projects become a regular component of training, career transitions and skills-development policies.

Consequences:

multiplication of partnerships;
need to qualify challenges;
recognition of portfolios;
development of studios, incubators and mentor communities;
a multi-funder economic model.

Appropriate positioning: international network for learning and incubation through impact projects.

Scenario 3 — Ambitious: Transnational Infrastructure for Skills and Projects

Real projects become a recognized unit of learning, experience and social contribution. Several organizations share a framework, an evidence platform, a mentor network and a portfolio of challenges.

Consequences:

participant mobility between countries and organizations;
stackable certifications;
longitudinal data;
applied research;
partnerships with employers, local authorities, universities and funders;
possible recognition as a new category of educational and professional institution.

Appropriate positioning: a distributed institute or university for the practice of impact projects.

Strategic and Operational Recommendations

Table 6 — Recommendations

Recommendations
Strategic and operational priorities
Recommendation	Objective	Justification	Concrete actions	Priority	Difficulty	Time frame	Success indicator
Create a common skills framework	Make the pathways coherent and understandable	The internal offerings mobilize similar skills without common evidence	Cross-reference EntreComp, GreenComp, LifeComp, DigComp, project management, impact and governance	Critical	High	0 to 6 months	Framework tested on at least two programmes
Structure a five-level pathway	Organize progression	Current formats can become stages in a continuum	Orientation, contributor, project manager, practitioner, mentor-trainer	Critical	Medium	0 to 6 months	Pathways and criteria published
Create a qualified challenge register	Improve project quality	Not all real problems are educationally usable	Form, committee, risk analysis and difficulty level	Critical	Medium	0 to 4 months	At least twenty qualified challenges
Formalize every partnership	Protect participants and partners	Implicit expectations cause conflicts and abandonment	Mandate, availability, data, ownership, use and continuity	Critical	Medium	Immediate	100% of real projects covered by contracts
Establish an evidence passport	Certify what each person knows how to do	The collective deliverable is insufficient	Portfolio, versions, oral defence, assessments and results	Critical	High	3 to 9 months	Portfolio usable by employers
Deploy an AI policy	Secure uses and assessment	AI is transforming practices and risks	Charter, register, authorized tools, prohibited data and oral defence	Critical	Medium	0 to 3 months	100% of teams declare their uses
Professionalize mentors	Stabilize quality	Expertise does not guarantee good support	Training, observation, supervision, code of conduct and progression	High	High	3 to 12 months	Average mentor assessment and low incident rate
Measure four levels of success	Avoid confusion between activity and impact	Learning, project, use and impact are different	Separate dashboards and associated decisions	Critical	High	3 to 9 months	Complete data for the pilot projects
Create an accessibility fund	Reduce inequalities	Intensive formats exclude certain groups	Scholarships, transport, accommodation, equipment and compensation	High	Medium	3 to 12 months	Increased diversity and fewer financially driven withdrawals
Organize continuity before the final presentation	Reduce abandonment	Most events end with the pitch	Follow-up lead, budget, transfer and closure decision	Critical	Medium	Immediate	Proportion of projects with a formalized outcome
Separate education, certification and investment	Prevent conflicts of interest	The model may combine several functions	Separate committees, declared interests and appeals	Critical	High	3 to 9 months	Procedure adopted and audits conducted
Build a hybrid economic model	Ensure sustainability and access	No single funder covers all objectives	B2C, B2B, B2G, foundations, licences and support	Critical	High	6 to 18 months	No funder above a defined dependency threshold
Engage a research partner	Establish independent evidence	The sector lacks longitudinal data	Protocol, comparison group where possible and publications	High	High	6 to 18 months	First longitudinal study initiated
Prepare Qualiopi, eduQua and micro-credentials	Increase recognition	Credibility depends on quality assurance	Mapping of requirements and certification pilot	High	High	6 to 24 months	Audit or recognition obtained
Transform former participants into a community of practice	Create a lasting feedback loop	Former participants can become mentors, recruiters or partners	Statuses, roles, training and facilitation	High	Medium	3 to 12 months	Rate of former participants active in the ecosystem
Develop a library of cases and failures	Capitalize on knowledge	Knowledge often remains tacit	Anonymized records, decisions, results and errors	Medium to high	Medium	3 to 12 months	Number of cases reused in training
Establish a safety and ethics protocol	Protect all parties	Projects may affect real people	Committee, risk levels, reporting and right to stop	Critical	High	0 to 6 months	All projects classified by risk level
Test the system on two existing pathways	Learn before generalizing	A complete architecture must be tested	Step & Go pilot and KEMENESI or governance pilot	Critical	Medium	6 to 12 months	Comparable results and documented improvements

Recommended Strategic Architecture for the Ecosystem

Level 1 — Discovery and Supervised Contribution

Indicative duration: one to five days.

Objectives:

discover impact projects;
understand a problem;
contribute to a clearly defined task;
learn the fundamental rules of cooperation and ethics.

Evidence:

short output;
individual reflection;
observed participation.

Level 2 — Hackathon or Creation Workshop

Indicative duration: three to ten days.

Objectives:

investigate;
design;
prototype;
cooperate;
present a proposal.

Evidence:

framing;
prototype;
tests;
decision log;
oral defence.

Level 3 — Development Studio

Indicative duration: six to twelve weeks.

Objectives:

deepen the investigation;
test uses;
build an operational model;
manage risks;
organize a transfer.

Evidence:

project file;
test results;
partner assessment;
individual portfolio.

Level 4 — Incubation or Professional Practice

Indicative duration: three to eighteen months.

Objectives:

manage the project;
develop partnerships;
govern;
finance;
measure results;
adjust the strategy.

Evidence:

real decisions;
usage results;
financial data;
governance;
impact analysis.

Level 5 — Mentor, Assessor or Trainer

Objectives:

provide support without taking over;
assess;
prevent risks;
facilitate cooperation;
transmit the method;
contribute to improving the system.

Evidence:

observation;
supervision;
team feedback;
case analysis;
contribution to the community of practice.

Recommended Action Plan

From Zero to Ninety Days

appoint a small design team;
inventory existing programmes and projects;
create the first version of the framework;
define levels of responsibility;
produce the partner contract;
produce the AI charter;
define challenge-qualification criteria;
select two pilots;
choose a maximum of fifteen to twenty indicators;
establish an ethics and quality committee.

From Three to Twelve Months

launch the two pilots;
train the first mentors;
test the evidence passport;
organize the oral defences;
measure the actual workload;
collect feedback from partners and beneficiaries;
monitor participants after six months;
adjust the economic model;
document at least five failures or difficulties;
prepare the quality-assurance file.

From Twelve to Thirty-Six Months

expand the challenge portfolio;
deploy micro-credentials;
integrate the programmes into a common pathway;
open the alumni community;
develop university and employer partnerships;
publish a first impact study;
train trainers;
test deployment in several territories;
consolidate the economic model;
prepare for transnational recognition.

Summary Table

Issue	Finding	Opportunity	Major risk	Recommended decision
Positioning	The market contains numerous training programmes and hackathons	Become an integrated infrastructure for learning, evidence and continuity	Offering perceived as an additional training programme	Position the system, not only the event
Education	Projects are effective under certain conditions	Combine fundamentals, action and reflection	Activity without learning	Formalize a common educational model
Certification	The final deliverable is insufficient	Portfolio and multidimensional evidence	Badge without credibility	Create an evidence passport
Artificial intelligence	Use is already established	Increase analytical capacity and accessibility	Dependence and data leakage	Policy, register and oral defence
Partners	Real challenges are attractive	International network of qualified problems	Unpaid work and unrealistic expectations	Contract and prior qualification
Mentoring	Quality depends heavily on individuals	Create a community of practitioners	Variability and excessive intervention	Train, supervise and recognize
Impact	Projects quickly claim impact	Connect learning and impact management	Impact washing	Distinguish output, use, outcome and impact
Inclusion	Intensive formats favour people who are available	Differentiate through accessibility	Reproduction of inequalities	Funds, compensation and support
Continuity	Prototypes are often abandoned	Studio and incubation after the hackathon	Costs and dispersion	Decide explicitly on the future of each project
Governance	Several functions can be combined	Complete value chain	Conflicts of interest	Separate training, certification and financing
Economic model	Subsidies alone are unstable	Hybrid and recurring income	Dependence or exclusion through pricing	Diversify and finance accessibility
Scaling up	Know-how exists across several programmes	Shared framework and training of trainers	Dilution of quality	Pilot before licensing or replicating
Research	Long-term effects are poorly documented	Become a scientific reference in the field	Internal measurements lacking credibility	Independent and longitudinal assessment

Strategic Conclusion

Learning through hands-on impact projects has considerable potential because it simultaneously responds to several transformations:

the need for transferable skills;
the search for professional experience;
changes in recruitment methods;
the search for meaning;
the complexity of transitions;
the convergence of training and innovation;
the development of micro-credentials;
the need for cooperation between sectors;
the need to demonstrate impact;
the questioning of traditional assessments by artificial intelligence.

Its credibility nevertheless depends on demanding discipline.

A real project is not automatically a good learning environment. An active participant is not automatically competent. A prototype is not an impact. A diverse team is not automatically collaborative. A certificate is not automatically recognized. A successful event is not automatically a lasting transformation.

The most promising model is based on a combination that is simple in principle but demanding in practice:

Solid knowledge, real projects, graduated responsibilities, human support, evidence-based assessment, ethical governance and organized continuity.

The OSI, Step & Go, KEMENESI, 2% for the Future, GFF and Global Impact Projects ecosystem already possesses a large proportion of the necessary building blocks. Its main challenge is not to invent an additional activity. It is to make these building blocks coherent, comparable, recognized and accessible.

The strongest differentiation would be to offer not only project-based learning, but to build a continuum in which:

every participant can progress;
every skill can be demonstrated;
every project can be qualified;
every partner knows their rights and responsibilities;
every impact can be discussed honestly;
every useful result can find continuity;
every former participant can become a contributor, practitioner, mentor or partner.

Selective Bibliography and Sources

Internal Documents and Corpora

Guide to SPIP typographical shortcuts.
OSI, Step & Go and Participatory Science conversational corpus, 2025-2026.
KEMENESI Impact Training Center conversational corpus, 2025-2026.
Governance and strategic leadership certification conversational corpus, 2025-2026.
Geneva Foundation for the Future and AGILE conversational corpus, 2025-2026.
Global Impact Projects and 2% for the Future conversational corpus, 2025-2026.

International Institutions and Frameworks

UNESCO, Education for Sustainable Development for 2030.
OECD, Learning Compass 2030.
OECD, work on skills-first approaches.
World Economic Forum, Future of Jobs Report 2025.
European Commission and JRC, EntreComp, GreenComp, DigComp and LifeComp.
European Union, European approach to micro-credentials.
UNESCO, artificial intelligence competency frameworks.
ISO, standards 21500 and 21502 relating to project management.
UNDP, SDG Impact Standards.
Impact Management Platform.
OECD-DAC, evaluation criteria.
CNIL, recommendations concerning artificial intelligence in education.

Academic Research

Meta-analysis of the effects of project-based learning, 2023.
Meta-analysis of problem-, project- and case-based pedagogies and motivation, 2024.
Systematic review of challenge-based learning, 2024.
Systematic review of authentic assessment, 2024.
Scoping review of authentic assessment, 2025.
Research on self-determination in education.
Research on psychological safety in project teams.
Research on social loafing and peer assessment.
Research on generative AI and project-based learning.

Benchmarks

Aalborg University, institutional PBL model.
Olin College, SCOPE.
MIT D-Lab.
Purdue University, EPICS.
TU/e innovation Space.
Aalto Design Factory.
Tiimiakatemia.
Kaospilot.
Social Impact Award.
PBLWorks, Gold Standard PBL.
Riipen.

Social Networks and Professional Communities

Public discussions concerning the limitations of group work, individual contribution and the lack of fundamentals.
Professional publications on authenticity and challenge-based learning.

Any message or comments?

Forum registration required

You must be registered before participating in this forum. Please enter your personal identifier . If you have not yet registered, you must register.

Connection | Sign up | password forgotten?