Strong recommendations turn program evaluation from a reporting exercise into a decision-making tool. In evaluation work, recommendations are the specific actions an organization should take based on evidence gathered through monitoring, interviews, surveys, administrative data, and outcome analysis. They sit inside the broader practice of reporting and communicating findings, which includes choosing the right format, tailoring messages to different audiences, and translating technical evidence into practical next steps. This matters because even rigorous evaluations fail when leaders, funders, and frontline teams cannot see what to do next, why it matters, or how quickly action is needed.
I have seen this repeatedly in program evaluation projects across nonprofits, public agencies, and grant-funded initiatives. Teams may spend months clarifying logic models, defining indicators, cleaning datasets, and testing findings, only to end with a report whose recommendations say little more than “improve outreach” or “increase collaboration.” Those statements sound reasonable, but they rarely drive action because they are vague, unprioritized, and disconnected from operational reality. Actionable recommendations are different. They name the change, identify who should act, indicate timing, reflect implementation constraints, and tie directly to evidence. They also recognize that reporting and communicating findings is not just about the final document; it is about equipping people to decide, budget, sequence, and execute.
As a hub page within reporting and communicating findings, this article explains how to write recommendations that lead to real movement. It covers what makes a recommendation credible, how to connect recommendations to findings, how to tailor them for executives and implementers, and how to present them in formats people actually use. It also addresses common mistakes, from overreaching beyond the evidence to flooding stakeholders with too many recommendations at once. If your goal is to help an evaluation influence strategy, staffing, service design, funding, or policy, the quality of your recommendations is often the difference between a report that is read and a report that changes something.
Start with findings, decisions, and audience needs
The best recommendations begin before the writing stage. They are shaped during evaluation planning, when you identify intended users, likely decisions, and the evidence those decisions will require. In utilization-focused evaluation, this principle is foundational: evaluations should be designed for use by specific people for specific purposes. In practice, that means asking early questions such as: Who will act on these results? What choices are coming next quarter or next fiscal year? What level of evidence will decision-makers expect before changing a program model, eligibility rule, staffing structure, or budget line?
When I draft recommendations, I work backward from likely decisions. If an executive team is deciding whether to expand a pilot, they need recommendations about scale conditions, staffing, cost, and fidelity. If program managers are trying to reduce participant drop-off, they need recommendations about intake, reminders, transportation support, curriculum pacing, or referral workflows. If a funder is reviewing continuation grants, they may need recommendations about measurement improvements, performance thresholds, or partnership structure. The same findings can lead to different recommendations depending on who needs to act and what authority they hold.
This is why generic reporting often fails. An evaluation may accurately describe low attendance, uneven outcomes across sites, or inconsistent data entry, yet still produce weak recommendations because the writer has not linked findings to a decision context. Good recommendations answer an implied stakeholder question: “Given this evidence, what should we do now?” For that reason, every recommendation should be traceable to a finding, but also anchored in a realistic organizational choice. Findings describe what the evidence shows. Recommendations specify what should happen next.
Use a clear structure for actionable recommendations
Action-oriented recommendations are easier to write when you use a consistent structure. The most reliable format includes five elements: the action, the rationale, the owner, the timing, and the expected benefit or risk reduction. For example, instead of saying “strengthen participant engagement,” write: “Within the next intake cycle, the program director should introduce a two-step enrollment reminder process using text and phone outreach for first-session registrants, because attendance records show that 38 percent of no-shows enrolled but never confirmed. This change is expected to improve first-session attendance and reduce early attrition.” That recommendation is specific enough to assign and implement.
Precision matters because recommendations compete for attention with budgets, compliance deadlines, staff turnover, and political constraints. Vague wording forces readers to interpret your intent, and interpretation creates delay. In contrast, concrete recommendations reduce ambiguity. They name what should change in process, policy, staffing, training, technology, or resource allocation. They also reveal whether the evaluator understands operations. In my experience, stakeholders trust recommendations more when they reflect how work actually gets done, not just what looks good on paper.
One practical method is to write each recommendation as a mini business case. State the action, cite the evidence behind it, identify the implementation owner, and indicate whether it is short term, medium term, or longer term. If there are prerequisites, say so. If there are risks, note them. This approach is especially useful in reporting and communicating findings because it supports multiple outputs: a full evaluation report, an executive summary, a board memo, a slide deck, or a management action plan.
| Weak recommendation | Why it fails | Stronger recommendation |
|---|---|---|
| Improve outreach. | Too broad; no owner, timeline, or evidence link. | By Q2, the outreach manager should shift 20 percent of recruitment effort from general social posts to referral partnerships with three clinics, because enrollment data shows referred participants enroll at twice the rate of web leads. |
| Increase staff training. | Does not specify skill gap or expected result. | Before the next service cycle, train case managers on standardized intake scoring, since file review found rating inconsistencies in 27 percent of records across sites. |
| Enhance data quality. | Unclear what process should change. | Assign a monthly data validation review to the evaluation lead and site coordinators, focusing on missing exit dates and duplicate participant IDs identified in the administrative dataset. |
Make recommendations evidence-based, bounded, and credible
Recommendations should be assertive, but they should never outrun the evidence. This is one of the most important disciplines in program evaluation. If your design was descriptive, do not imply causal certainty. If your sample was limited, say so. If a recommendation depends partly on stakeholder judgment, operational feasibility, or external research, make that explicit. Credibility comes from disciplined reasoning, not from sounding absolute.
Evidence-based recommendations usually draw from three sources together: the evaluation’s own findings, implementation knowledge from the organization, and relevant external standards or research. For example, if participant completion drops sharply after week three, your recommendation may be based on attendance records, staff interviews about scheduling conflicts, and established retention practices such as reminder systems or cohort check-ins. Named frameworks can help here. SMART criteria are useful for sharpening recommendations, and RACI can clarify who is responsible, accountable, consulted, and informed during implementation. For performance measurement recommendations, standards from the Centers for Disease Control and Prevention, the Urban Institute, or the American Evaluation Association often provide useful grounding.
Boundaries are equally important. Sometimes the right recommendation is not “expand the program,” but “do not scale until minimum implementation conditions are met.” I have made that recommendation when pilot sites showed promising participant outcomes but inconsistent fidelity, undocumented adaptations, and cost data that could not support budgeting. Decision-makers usually appreciate this kind of candor. It protects them from acting on enthusiasm rather than readiness.
Strong recommendations also distinguish between evidence of a problem and certainty about the best solution. If survey data shows participants feel confused about eligibility, the evidence supports recommending clearer communication, revised scripts, or simplified forms. It does not automatically support a complete redesign of the service model. Keep the recommendation proportionate to the evidence. That is how evaluators maintain trust while still being useful.
Prioritize recommendations so stakeholders can act
One of the fastest ways to weaken an evaluation report is to include too many recommendations. Long lists create false equality, as if every issue deserves the same attention. In reality, organizations need sequencing. Leaders want to know what to do first, what can wait, and what will have the greatest effect relative to effort. Prioritization is therefore part of writing recommendations, not a separate exercise.
I typically group recommendations by urgency, impact, and feasibility. Urgent items address compliance, safety, funding risk, or immediate service disruption. High-impact items are the changes most likely to improve outcomes, efficiency, equity, or participant experience. Feasible items are those the organization can implement with current authority, staffing, and budget. Recommendations that score well across all three dimensions should appear first. Lower-feasibility items may still belong in the report, but they should be labeled as longer-term or conditional.
This approach is especially important when communicating findings to mixed audiences. Boards may focus on strategic priorities, program directors on operational barriers, and evaluators on methodological caveats. Prioritized recommendations create a shared map. They also support follow-through by making it easier to convert report language into action plans, workstreams, and progress reviews. In one workforce development evaluation, we reduced thirteen draft recommendations to four priorities: standardize employer follow-up, redesign orientation, improve completion tracking, and clarify staff roles across sites. That version got adopted. The longer list had generated polite agreement but no movement.
Prioritization also helps preserve the distinction between findings, implications, and recommendations. Not every finding requires a recommendation, and not every recommendation should carry the same weight. If the central aim of reporting and communicating findings is use, then your recommendations should reflect the choices stakeholders can reasonably make next.
Tailor recommendations for different users and formats
A recommendation that works in a technical report may fail in a board presentation, and a recommendation that makes sense to executives may be too abstract for frontline managers. Tailoring is therefore essential. The underlying recommendation can remain consistent, but the phrasing, detail level, and supporting context should change based on audience and format.
For senior leaders, lead with the decision, business implication, and resource consequence. They need concise language such as: “Consolidate intake across sites to reduce duplicate effort and improve eligibility consistency; implementation will require revised workflows, one data field change in the case management system, and supervisor training.” For program managers, go deeper into process details, dependencies, and implementation risks. For frontline staff, explain exactly what practice should change and how success will be monitored. For funders, connect recommendations to outcomes, stewardship, and learning.
Format matters as much as audience. In a full report, recommendations can sit after each major finding cluster or in a dedicated section near the end. In an executive summary, they should appear as concise decisions with supporting evidence in one sentence each. In slide decks, limit text and make the recommendation headline the slide. In oral briefings, state the recommendation first, then explain the evidence. This inverted structure works because stakeholders usually listen for action, not for methodological detail.
As a hub for reporting and communicating findings, this page connects naturally to related topics such as writing executive summaries, presenting data visually, communicating uncertainty, facilitating stakeholder sensemaking, and building management response plans. Recommendations improve when all of those supporting practices are strong, because action depends on clarity, credibility, and shared understanding.
Avoid common mistakes that weaken evaluation recommendations
Most weak recommendations fail in predictable ways. The first mistake is vagueness. Words like improve, enhance, strengthen, and increase are not wrong, but they are incomplete unless they specify what should change. The second mistake is misalignment with evidence. If the findings point to referral bottlenecks, a recommendation about curriculum redesign may be interesting but unsupported. The third mistake is ignoring feasibility. Recommending a new enterprise platform, major staffing expansion, or policy overhaul without acknowledging cost, procurement, or authority constraints makes the report easier to dismiss.
A fourth mistake is treating recommendations as neutral observations. Recommendations are judgments. They should reflect evaluator analysis, not just a summary of what stakeholders said they wanted. Of course, stakeholder perspectives matter. In participatory and developmental evaluation settings, they matter a great deal. But the evaluator still has to weigh evidence, identify tradeoffs, and make a reasoned call. Otherwise, recommendations become a transcript of preferences rather than a guide to action.
Another common problem is writing recommendations that stop at the organizational level. Statements like “the organization should improve cross-team communication” are often too abstract to implement. Cross-team communication improves when a specific mechanism changes: a weekly case conference, a shared referral tracker, a revised escalation protocol, or a designated handoff role. Finally, many reports omit implementation conditions. If a recommendation depends on better baseline data, policy approval, union consultation, or fiscal-year planning, say so plainly. Clear limitations do not weaken recommendations; they make them usable.
Turn recommendations into action plans and follow-up
The most effective recommendation writing anticipates implementation. A recommendation is not complete when it sounds polished; it is complete when a stakeholder can convert it into next steps. That is why I often draft recommendations alongside a management response framework. The recommendation states what should happen. The management response records whether leadership agrees, who owns the work, what resources are needed, and when progress will be reviewed. This simple bridge dramatically increases the chance that findings lead to change.
Follow-up should also be built into reporting and communicating findings. For major evaluations, schedule a decision meeting after report delivery, not just a presentation. Use the meeting to confirm priorities, identify owners, and define evidence of implementation. Then establish a check-in point, often sixty or ninety days later. In public sector and philanthropic settings, this discipline is especially valuable because staff turnover can quickly sever the link between an evaluation report and organizational action.
In the end, writing recommendations that drive action requires more than strong prose. It requires evidence discipline, operational understanding, audience awareness, and a clear view of how decisions get made. The best recommendations are specific, prioritized, feasible, and anchored in findings. They tell leaders what to do, help managers understand how to do it, and give teams a basis for tracking progress. If you want your program evaluation work to matter beyond publication, audit your next report’s recommendations line by line and rewrite anything that cannot be assigned, timed, or acted on.
Frequently Asked Questions
What makes a recommendation actionable in a program evaluation report?
An actionable recommendation is specific enough that a decision-maker can understand exactly what should happen next, why it matters, and what evidence supports it. In program evaluation, weak recommendations often sound vague, such as “improve communication” or “increase engagement.” Strong recommendations go further by identifying the desired action, the target group, the purpose, and the expected result. For example, instead of saying “strengthen participant follow-up,” a more actionable recommendation would be “implement a 30-day post-service follow-up process for first-time participants to improve retention and identify barriers earlier.” That level of precision helps leaders move from agreement to execution.
Actionable recommendations are also tightly linked to the findings. They should emerge directly from what the evaluation uncovered through interviews, surveys, monitoring data, administrative records, and outcome analysis. If the evidence shows that a program is effective overall but underperforming for a specific population, the recommendation should address that gap rather than introduce a disconnected idea. This evidence-to-action logic is what gives recommendations credibility. It reassures readers that the recommendation is not a personal opinion, but a practical response to documented patterns, needs, and opportunities.
Finally, strong recommendations consider feasibility. A recommendation may be evidence-based and well written, but if it ignores budget, staffing, timing, or organizational authority, it is less likely to be adopted. The most useful recommendations balance ambition with realism. They are clear about priority, indicate who may need to act, and often suggest an implementation path. When recommendations are specific, evidence-based, and feasible, they become far more likely to drive meaningful action rather than remain as good ideas in a final report.
How do you connect evaluation findings to recommendations without oversimplifying the evidence?
The key is to build a clear chain of reasoning from evidence to conclusion to action. Recommendations should not appear suddenly at the end of a report as standalone advice. Instead, they should reflect a logical interpretation of the data presented throughout the evaluation. That means first identifying the most important findings, then asking what those findings imply for program design, service delivery, policy, or resource allocation. If outcome data show strong results for one service model and weaker results for another, the recommendation might focus on expanding the more effective model or revising the weaker one. The recommendation is grounded in evidence, but it also depends on thoughtful interpretation.
Avoiding oversimplification means acknowledging complexity where it matters. Not every finding leads to a single obvious answer, and not every recommendation should be presented as universally certain. In many evaluations, evidence comes from multiple sources that may reinforce one another or reveal nuance. Survey results may highlight participant satisfaction, while interviews uncover implementation barriers, and administrative data reveal uneven usage patterns. A strong recommendation synthesizes these forms of evidence rather than relying on one data point in isolation. It may also note where confidence is high, where further testing is needed, or where implementation should happen in phases.
It also helps to distinguish between what the evidence proves and what it suggests. Evaluators should be careful not to overstate causality or make recommendations that go beyond the scope of the findings. At the same time, recommendations should still be useful. A good balance is to write recommendations that are decisive but appropriately qualified. Phrases such as “based on consistent evidence across interviews and service utilization data” or “pilot this change before full rollout” allow you to stay accurate while still being practical. This approach preserves the integrity of the evaluation and makes the recommendations more trustworthy to stakeholders.
Who should recommendations be written for, and how should they be tailored to different audiences?
Recommendations should always be written with a specific audience in mind, because different stakeholders have different levels of authority, different information needs, and different reasons for engaging with the evaluation. Senior leaders may want concise recommendations tied to strategic priorities, budget implications, and organizational risk. Program managers may need operational guidance, such as process improvements, staffing adjustments, or changes to implementation timelines. Frontline teams may benefit from recommendations that clarify practice changes, communication expectations, or participant support strategies. Writing for everyone at once often leads to recommendations that are too broad to be useful to anyone.
Tailoring starts with understanding who can act on the recommendation. A recommendation is much more effective when it is directed toward the people or groups with the power to implement it. In some cases, that means naming the likely owner of the action, such as leadership, a program unit, a data team, or a cross-functional committee. It also means using language that fits the audience. Technical readers may want methodological confidence and supporting detail, while non-technical readers may need a plain-language explanation of what the evidence means in practice. The broader reporting and communication strategy matters here: the same recommendation may appear in a full report, an executive summary, a slide deck, or a briefing memo, each adjusted for the audience and format.
Tailoring does not mean changing the evidence. It means changing how the recommendation is framed so it is easier to understand, prioritize, and act on. For example, a board-facing version may emphasize impact, accountability, and long-term outcomes, while an implementation-facing version highlights workflow, training, and measurement steps. The strongest evaluation communication plans recognize that recommendations are not just written once and delivered passively. They are translated for different audiences so that the right people can see both the urgency of the issue and the practical steps needed to respond.
What is the best structure for writing strong recommendations that decision-makers will use?
A reliable structure is to include five elements: the action, the rationale, the evidence base, the intended outcome, and implementation considerations. Start with a direct statement of what should be done. This should be written in clear, active language, such as “standardize intake procedures across sites” or “expand outreach to participants who disengage within the first two weeks.” After that, briefly explain why the recommendation is needed by tying it to the central finding or problem identified in the evaluation. This helps readers see the immediate relevance of the recommendation rather than treating it as generic advice.
Next, anchor the recommendation in evidence. You do not need to restate the full analysis, but decision-makers should be able to understand the factual basis behind the recommendation. Referencing patterns from monitoring data, interviews, survey responses, administrative records, or outcome trends strengthens the recommendation and reinforces its legitimacy. Then describe the intended result. This is especially important because it shifts the recommendation from activity to impact. A recommendation should not only say what to do, but also what improvement it is expected to produce, whether that is better participant retention, more equitable access, stronger implementation consistency, or improved outcomes.
Finally, include practical implementation considerations when appropriate. This can involve sequencing, resource implications, responsible parties, or whether the recommendation should be piloted before wider adoption. In some cases, it is helpful to categorize recommendations by time horizon or priority level, such as immediate, medium-term, and longer-term actions. That structure helps organizations decide where to start. Decision-makers are more likely to use recommendations that feel organized, realistic, and clearly connected to operational decisions. A good structure turns recommendations into a roadmap rather than a list of observations.
What are the most common mistakes to avoid when writing recommendations in evaluation reports?
One of the most common mistakes is being too vague. Recommendations like “improve services,” “increase collaboration,” or “do more outreach” may sound reasonable, but they do not tell an organization what action to take. Without specificity, even a well-supported recommendation can be difficult to implement. Another frequent issue is offering recommendations that are disconnected from the findings. If readers cannot see how the recommendation follows from the evidence, they may question its credibility or dismiss it as subjective. Recommendations should always be traceable to the evaluation’s documented results and analysis.
Another mistake is ignoring organizational context. Recommendations that fail to account for staffing capacity, budget constraints, governance structures, or political realities may be technically sound but practically unusable. Similarly, recommendations can lose impact when they are too numerous or presented without prioritization. If every issue is framed as urgent, decision-makers may struggle to identify what matters most. Strong evaluators help organizations focus by distinguishing between high-priority actions, promising longer-term changes, and ideas that require additional evidence before implementation.
A final mistake is writing recommendations in language that is too technical, too cautious, or too abstract for the intended audience. Evaluation reports often contain complex evidence, but recommendations should translate that complexity into practical meaning. They should be clear without being simplistic, confident without overstating certainty, and detailed without becoming inaccessible. When recommendations are grounded in evidence, written for the right audience, prioritized appropriately, and framed in plain but authoritative language, they are much more likely to influence real decisions and improve program performance.
