How to write a logframe
Last updated
Logical framework
read from the bottom up
The short answer
A logframe is a short table with four rows (goal, outcome, outputs, activities) and four columns (summary, indicators, means of verification, assumptions). Write it from the top down, then test it from the bottom up: if we do these activities and the assumptions hold, we deliver these outputs; if we deliver the outputs and the assumptions hold, the outcome follows; if the outcome is achieved, the project contributes to the goal.
What a logframe is, and what funders use it for
A logframe (logical framework, or logical framework matrix) sets out a whole project in a short table, usually one to three pages. An assessor uses it to check three things quickly: that the activities really lead to the results you promise, that the results can be measured, and that you have thought about what is outside your control.
After the award it becomes the document you report against. So only promise in a logframe what you can deliver and count. It is not fixed for ever: funders expect it to be updated, with their agreement, as the project learns.
The four levels
| Level | What it is | Test |
|---|---|---|
| Goal | The wider change the project contributes to. One only. You do not achieve it alone. | Would other organisations and the government also be working towards this? |
| Outcome | The change in people's lives, behaviour or services by the end of the grant. One is best: FCDO allows only one, and the EU asks for one specific objective and lets you list “other outcomes” under it. | Is this a change for people, not something we do? |
| Outputs | What the project itself delivers and controls. Number them so each is tied to its outcome. More than six is a sign to look again. | Can we deliver this by doing the work and spending the budget? |
| Activities | The work that produces each output, numbered to match. Many templates ask for the inputs beside them: the staff, equipment and money needed. | Does every activity belong to an output? |
The most common error is writing an activity as an outcome. “Train 40 community health workers” is an activity. “40 community health workers trained” is an output. “Young people are seen by a counsellor within a week of referral” is an outcome.
The four columns
- Summary. One sentence for each row, in plain words.
- Indicators. What you will measure: “Share of referrals seen within 7 days”. Then the baseline (48% in 2025), the target (75%) and the date (December 2028). The EU and FCDO templates have a separate column for each; in a simple four-column table you can write them together. Where an indicator counts people, report it by sex and age, and by disability where you can.
- Means of verification. The record that proves the indicator, who keeps it and how often: “Referral register, kept by the clinic nurse, checked quarterly”.
- Assumptions. Conditions outside your control that must hold for this level to lead to the one above: “County facilities keep counselling rooms available”. If you can manage it yourself, it is not an assumption: add an activity to deal with it, and list it in the risk table.
Write it in six steps
- Start from the problem. Write the change you want for people. That is your outcome. Then name the wider goal it contributes to.
- List what you must deliver for that change to happen. These are the outputs.
- List the activities under each output. Leave out anything that does not lead to an output.
- Write the assumptions at each level: what must be true, beyond your control, for this level to lead to the one above? For each one ask whether it matters and whether it is likely to hold. If it matters a lot and is unlikely to hold, it is a “killer assumption”: change the design, or do not submit the project. Leave out assumptions that are almost certain to hold.
- Choose indicators and evidence. One or two per row, using numbers you already collect where you can. Take the baseline from your own records. Where you have none, say so and plan a baseline study in the first three to six months.
- Test it from the bottom up. Read aloud: “If we do these activities, and these assumptions hold, then we deliver these outputs.” Then the next level. If a step needs a leap of faith, something is missing.
A worked example: community mental health
This example is for an invented organisation, Kijani Afya Trust, which runs counselling and peer support for young people in Western Kenya. Every figure is invented.
| Level | Summary | Indicator | Baseline and target | Means of verification | Assumptions |
|---|---|---|---|---|---|
| Goal | Better mental health among young people aged 15 to 24 in Siaya and Homa Bay counties. | Taken from the county health plan: name the indicator. | As set in the plan. | County health department annual report: name the document and year. | Not needed at this level. |
| Outcome | Young people in the two counties get counselling within a week of referral. | Share of referrals seen within 7 days, by sex and age group. | 48% (2025) to 75% by Dec 2028. | Referral register, kept by each clinic nurse; checked quarterly by the programme officer. | County hospitals continue to accept referrals for severe cases. |
| Other outcome | Schools identify and refer students in distress. | Schools making at least one referral a term. | 5 (2025) to 14 by Dec 2028. | School referral forms, collected each term by the schools officer. | Parents and guardians allow students to attend counselling. |
| Output 1.1 | Counselling available in 10 health centres. | Health centres with a weekly counselling session. | 6 (2025) to 10 by Jun 2028. | Session timetable and attendance register, kept by each counsellor. | County facilities keep counselling rooms available. Trained counsellors remain in post. |
| Output 1.2 | 40 community health workers trained to recognise and refer. | Health workers who complete the training and pass the test, by sex. | 0 to 40 by Mar 2028. | Training register and test results, kept by the trainer. | Trained health workers stay in their areas. |
| Output 2.1 | Peer counsellors active in 14 schools. | Schools with two or more active peer counsellors. | 5 (2025) to 14 by Sep 2028. | Termly school visit report, written by the schools officer. | Schools keep time in the timetable for peer support. |
| Output 2.2 | Two teachers trained in each of 14 schools. | Teachers who complete the training, by sex. | 0 to 28 by Jun 2028. | Training register, kept by the trainer. | Trained teachers are not transferred. |
| Activities | 1.1 Recruit and train 4 counsellors. 1.2 Train 40 health workers; set up referral registers. 2.1 Train and supervise peer counsellors. 2.2 Train 28 teachers. All: quarterly review with county health teams. | Inputs: staff, trainers, transport, registers. | Cost: KES 9,600,000, as in the budget. | Work plan and training registers. | Supervisors release health workers. Head teachers release teachers. No long health-worker strike in the two counties. |
Notice four things. The outcome and output indicators say what is measured, the starting point, the target and the date. Each means of verification names a record and who keeps it. Each assumption sits on the row it must hold for, so the conditions for training to happen are beside the activities. And the goal uses an indicator that the county already measures: do not promise to measure a goal-level change yourself.
Common mistakes
- Activities written as outcomes. “Conduct awareness sessions” is not a change for anyone.
- A goal you promise to achieve alone. The goal is a contribution.
- Several outcomes of equal weight. One main outcome keeps the project honest about what it is for.
- Too many indicators. You will have to collect and report every one.
- Indicators with no baseline or date. “Increased awareness” cannot be measured.
- Counting people without saying who. Funders expect figures by sex and age.
- Invented baselines. If you do not know, say when you will find out.
- Assumptions that are really your job. “Staff will be recruited on time” is yours to manage.
- A logframe that disagrees with the budget or the narrative. Assessors compare them, and the inputs must add up to the budget.
The words different funders use
The logic is the same everywhere, but the labels change. Always use the funder's own template and wording when the call gives one.
| This guide | EU template | FCDO | Older USAID logframes |
|---|---|---|---|
| Goal | Impact (overall objective) | Impact | Goal |
| Outcome | Outcome (specific objective), with “other outcomes” below it | Outcome (one only) | Purpose, and sub-purposes |
| Output | Outputs (older EU forms call this row “results”) | Outputs | Outputs |
| Activities | Activity matrix, with means and costs | Activities, kept in a separate log | Inputs |
| Indicator | Indicator, baseline, target and current value, in separate columns | Indicator, baseline, milestones and target | Indicators |
| Means of verification | Source and means of verification | Source | Means of verification |
| Assumptions | Assumptions | Assumptions, with a risk rating | Assumptions |
Be careful with the word “results”. The EU uses it for all three levels: impact, outcomes and outputs. FCDO also asks for milestones between the baseline and the target.
If you are also asked for a theory of change, write it first. It explains why the steps follow from each other; the logframe then summarises it. See also how to write a concept note.
Drafting a logframe in Afrant
Afrant drafts a logframe from your own fact sheet: indicators use the numbers in your reports, each linked to its source, and a missing baseline is left as a visible gap for you to fill. You still decide what the project promises. How drafting works.
Questions people ask
What is a logframe in simple terms?
A logframe, or logical framework, is a short table that summarises a project, usually on one to three pages. Its rows are the goal, the outcome, the outputs and the activities. Its columns say what each one is, how it will be measured, where the evidence will come from, and what must hold true for it to work.
What is the difference between an output and an outcome?
An output is something the project itself delivers and controls, such as 40 health workers trained. An outcome is the change that follows in people's lives or in a service, such as young people being seen by a counsellor within a week of referral. If you can deliver it by doing the work and spending the budget, it is an output.
How many outcomes and indicators should a logframe have?
One outcome is best: FCDO allows only one, and the EU template asks for one specific objective with any other outcomes listed under it. Use one or two indicators for each row, and no more than three for an output. You will have to collect and report every one.
What are means of verification?
The actual records that prove an indicator: an attendance register, clinic records, a survey report. Name the record, who keeps it and how often it is collected. “Project reports” is not a means of verification.
Is a logframe the same as a theory of change?
No. A theory of change explains in words why the activities should lead to the change, including the steps in between. A logframe is the summary table used to manage and measure it. Many funders ask for both, and they should agree with each other.
What if we do not have a baseline?
Say so, and plan and budget a baseline study in the first three to six months. Never invent one. Some funders, such as FCDO, expect a baseline at outcome level before they approve a project.
What does SMART mean for indicators and targets?
Specific, measurable, achievable, relevant and time-bound. In practice: say exactly what is counted, the starting figure, the target and the date, and choose something your organisation can really measure.
Read next
- How to write a concept noteWhat funders expect in a few pages, the usual sections, the fixed EU form, and how a concept note differs from a full proposal.
- Using AI in grant proposalsWhether funders accept it, what to disclose, what goes wrong, and how to use it without inventing facts or exposing people's details.
- Proposal writingDrafts written from your own documents. Every figure links to the page it came from, and a claim with no source is flagged.
Draft your logframe from your own figures.
Afrant is in pilot and opens one organisation at a time. It is free during the pilot, and there is a sample workspace to explore before you upload anything.