Learning Agenda

A learning agenda is a prioritized list of questions a program commits to answering through its M&E system, with each question tied to a specific decision. It turns M&E from a reporting exercise into an evidence-generation engine.

What a Learning Agenda Does

Most M&E systems are built for reporting: did you hit your targets, did you spend the money, did you reach the people you said you would. A learning agenda sits on top of that and asks a different question: what do we actually need to learn to make this program better, and what evidence will answer that question.

The shift matters. Reporting M&E tells a funder the program is on track. Learning M&E tells the program team what to do next. A good learning agenda typically has five to eight questions; a team new to this can start with three to five and add more as the first ones are answered. Fewer than three and you are probably not being specific enough. More than that and you will not have the resources to answer any of them well. Draft the agenda at the start of implementation, then update it each quarter as questions get answered and new ones emerge.

Components of a Good Learning Agenda

A learning agenda is not just a list of questions. Each question needs four things attached to it:

• A structured question. Specific enough that you know what evidence would answer it. "Is the program working" is not a learning question. "Do community health workers retain clinical skills three months after training without refresher contact" is.
• An evidence plan. What data will answer this question, how will it be collected, and when. If the question requires a midline survey you do not have budgeted, the question is aspirational (a good wish, but no one has the money or time to answer it), not operational (something the team can actually do).
• A decision point. Who will use this evidence, and what decision does it inform. If no one is waiting for the answer, the question does not belong on the agenda.
• A dissemination plan. How and when the findings reach the people who need to act on them. Internal team, partners, donor, community, peer organizations. Evidence that sits in a report nobody reads has no programmatic value.

To draft one question, copy this table and fill the right-hand column. The example is one row of a larger agenda; it can sit in a MEL plan annex or beside your indicator table, one block per question.

The So-What Test

Most programs confuse data collection with learning. A team can track 18 indicators, submit every quarterly report on time and still learn nothing that changes the program. A learning question is different: the answer changes what someone does.

Before you keep a question, ask: "If the answer is yes, what do we do differently? If the answer is no, what do we do differently?" If both answers lead to the same action, you do not need the data. Cut the question or rewrite it until each answer points somewhere different.

Example. Healthy Futures is a fictional nutrition program with 30 community health workers (CHWs) in three zones. Its team drafted five questions at Month 9, after a donor asked for them; earlier would have been better. One was "Are nutrition education sessions changing caregiver knowledge and practices?"

The program had attendance data and, since Month 7, a short quiz given to caregivers before and after sessions. Attendance was 88 percent, but average quiz scores (out of 100) rose by less than 5 points. Caregivers were attending and not learning. Zone A's lead noticed that CHWs who ran cooking demonstrations instead of lectures got scores 8 to 12 points higher.

The question passed the test: a yes meant keep the lectures, a no meant change the format. The team switched all CHWs to cooking demonstrations at the Month 10 refresher.

Learning Agenda vs MEL Plan

A learning agenda is a section of a MEL plan, not a replacement for one. The MEL plan covers the full monitoring, evaluation, and learning system: indicators, data collection methods, roles, timelines, budget. The learning agenda is the specific subset of that system dedicated to answering priority evidence questions.

A learning agenda without a MEL plan cannot operationalize itself. There is no data collection schedule, no assigned staff, no budget for analysis. A MEL plan without a learning agenda can function, but it tends to default to pure reporting and misses the opportunity to generate evidence for decisions.

Proposal Context

In a proposal, link learning questions to actual decisions, evidence sources, review moments, roles and resources. Check the applicable agreement and solicitation for any learning commitments, scoring criteria or prescribed format. A funding source alone does not establish a mandatory learning agenda or imply approval.

A proposal with a clear learning agenda signals design maturity (a well-thought-through project design, not just a list of activities). It shows the team has thought past the logframe and has a plan for generating evidence that will improve the program during implementation, not just document it after. It distinguishes you from the generic M&E narrative that reviewers read hundreds of times.

The common pitfall in proposal learning agendas is listing research questions nobody will have the authority or process to act on. If your proposed learning question is "what is the causal mechanism linking livelihood training to child nutrition," you are proposing a PhD thesis, not a learning agenda. Keep questions at the level the program team can actually investigate and use.

Common Mistakes

Too many questions. A learning agenda with 15 questions is a wish list. You cannot resource that many. Cut to the five to eight the team will actually investigate.

Questions not tied to decisions. If answering the question would not change what anyone does, it does not belong. This is the single most common failure mode.

No dissemination plan. Evidence that does not reach decision-makers in a usable format is wasted. Build the route from finding to decision into the agenda itself.