Hoverview

Project practice · Donor

Reading a donor template as a set of rules, not a spreadsheet

Most of what a donor template requires is not in its cells. What to extract before typing, and why to write it down once.

A donor template usually arrives as an attachment called something like Annex 2a. Someone opens it, sees a grid, and starts typing. That is the right instinct for a form and the wrong one here, because the grid is the smallest part of what the donor is asking for.

Two things are true at once. The template really is a spreadsheet, and filling it in is most of the visible work. It is also a compressed rule set, and the rules that get a submission sent back are mostly invisible in the cells. Nothing in a column header tells you that assumptions may be recorded at only two of the four levels. An empty cell does not announce that it is forbidden to stay empty. A box with room for a paragraph does not mention that the field is capped at 250 characters.

So the question worth answering before anyone types is not what goes in each box. It is which rules this donor has encoded, where each one is written down, and which of them the template is structurally incapable of showing you.

What the grid cannot tell you

Take the case that produces the most rework. Defra's Biodiversity Challenge Funds head their indicator column "SMART Indicators (including disaggregated targets)", which is a composition instruction hiding in a header: the target goes inside the indicator sentence. The UK guidance from 2011 takes the opposite view and says outright that there is no such thing as a SMART indicator, because an indicator is what you measure and a target is the value you expect. Both of those are rules about the same cell, and only one of them is written in a place a person filling in the grid would look.

The pattern repeats. The Biodiversity Challenge Funds require a floor of two indicators and set no ceiling; the UK frame caps you at three per output. Neither number appears in the spreadsheet, because a spreadsheet has no way to express "at least two". The guidance note holds it, and the guidance note is a separate file that often does not travel with the template when it is forwarded internally.

This is the practical claim of this post. If you have received the template and not the guidance, you have the columns and none of the constraints, and the constraints are what determines whether the thing you submit comes back.

Six kinds of rule

A donor frame encodes six distinct things. They are worth separating, because they fail differently and they live in different places.

  • Level names and cardinality. What the top row is called, how many of each level are permitted, and whether activities appear in the matrix at all. DG ECHO's Single Form Guidelines (2017) add a Principal Objective level above the goal, which most other frames do not have, and cap indicators at ten per result.
  • The column set. Which columns exist, in what order, and which levels each one applies to. The UK frame permits assumptions at outcome and output level only, which is a column that stops partway down the table. Norad's guidance (2008) splits means of verification into four sub-columns: Means of Verification, Collection Methods, Frequency, Responsibility.
  • Composition. What must be inside a cell rather than beside it. The SMART indicator disagreement above is a composition rule, and it is the one most likely to be discovered late.
  • Numbering and identity. Dotted decimal, prefixes, whether a code stays stable when a row is inserted, and whether the donor's numbering is the same one your team uses internally. The GEF Results Measurement Framework (2022) forces every project onto numbered core indicators one through eleven, so numbering is not the team's to choose.
  • Governance. Who approves a change at which level, and when changes are not accepted at all. The Global Fund's performance framework (2026) states that the framework is part of the Grant Agreement, making every indicator a term of a contract rather than a planning choice.
  • Publication. What may not appear in a published document, which is a rule about people rather than data.

That last one produces the sharpest conflict in the corpus. WWF's project and programme management standards put a named individual in the responsibility cell, because a delivery team needs to know whose job a thing is. The UK guidance forbids naming individuals, because logframes are published. Neither is wrong. They are answering different questions about who reads the document, and no amount of careful cell mapping reconciles them.

Cardinality first, because it changes the design

Of the six, cardinality is the one to extract before anything else, because it is the only one that can force a redesign rather than a reformat.

The permitted number of outcomes runs from one under the UK frame, to two or three at the Green Climate Fund, which hedges it as "generally (but not always)", to eighteen in GFDRR's banded framework, which is not a matrix at all and carries no indicator columns. The Asian Development Bank's DMF guidelines (2024) allow one outcome statement with several dimensions of performance, and at the impact level require no indicators or targets at all, because the ADB does not consider project-level impact measurable through a project's own monitoring. A programme designed around three outcomes and submitted to a funder that permits one is not a formatting problem. Somebody has to decide which two outcomes become outputs, and that decision changes what the programme claims to be accountable for.

Doing that under deadline, in the week of submission, is how a results logic ends up shaped by a template rather than by the theory. Doing it in week one, knowing the number, is a design conversation.

The rules that arrive somewhere else

Governance and timing almost never appear in the template, and they are the rules most likely to make a legitimate correction into a breach.

UK Aid Match will not accept a logframe change within three months of an annual review assessment, and requires baseline data on impact, outcome and output indicators within three months of project start. The UK guidance from January 2011 made a logframe mandatory for every newly approved project regardless of value, then applied a fuller profile above a one million pound line and a shortened one below it, so two programmes in the same office under the same funder are governed by different rules. Several frames name a different approver for a change to a statement than for a change to a target.

None of that is in the grid. Most of it is in the grant agreement, the guidance note, or an email from a programme officer, which means it is discoverable only if somebody goes looking before the first review rather than during it.

What to write down once per donor

When a new template lands, the work is to answer these once and keep the answers where the next person can find them:

  1. What is each level called in this frame, and how many of each are permitted?
  2. Do activities appear in the matrix, or somewhere else entirely?
  3. Which columns exist, in what order, and does any column stop partway down the levels?
  4. What must be composed inside a cell rather than held beside it, and does anything have a character limit?
  5. Is disaggregation required, and against which named dimensions?
  6. Who approves a change at each level, and is there a period when changes are refused?
  7. What may not appear in the document because it will be published?
  8. Which of these answers came from the template, and which came from a guidance note or a conversation?

Question eight matters more than it looks. An answer that came from a phone call is an answer that will be lost when that person changes role, and it is the category most likely to be wrong a year later.

Why once is the right number of times

The economics here are unusual and they favour doing this properly. Funders are few, and their rules are stable for years at a time. The UK guidance being quoted in this post dates from 2011. The WWF standards date from 2012. A rulebook read carefully once serves every project submitted to that funder afterwards, and every colleague who inherits one.

Against that, the current arrangement is that each organisation decodes the same PDF independently, stores the result in one experienced person's head, and loses it when that person moves on. The second project under the same funder starts the decoding again, usually from a copy of the first project's spreadsheet, which carries the answers without the reasoning and so cannot be checked.

The honest cost is that reading a frame properly takes most of a day, and the payoff is invisible until the second project. That is a hard case to make to a team that is already late on the first one.

Where the method stops

Two limits, both real.

Reading each frame precisely does not reconcile them. The rules genuinely conflict, and a team reporting to several funders still ends up maintaining several renderings of one logic, which is a structural problem rather than a comprehension one. Precision makes the conflicts visible and legible. It does not remove them.

The second limit is harder. Some rules are not written anywhere. They exist as a reviewer's preference, applied consistently enough to feel like policy and never documented, and the only way to find them is to submit something and be told. Every experienced programme manager can name two or three of these for the funders they work with most.

Which raises the question this post cannot answer. Of the rules that actually govern a submission, how many are in the guidance, and how many are in somebody's head? Nobody has counted, and the answer would tell you whether this is a documentation problem or something else.

Sources

The frames quoted

  • Guidance on using the revised Logical Framework, DFID (January 2011). Source for the mandate on all newly approved projects, the fuller and shortened profiles either side of £1 million, assumptions at outcome and output level only, the three-indicator cap, the 250 character limit, and the statement that there is no such thing as a SMART indicator
  • Supplementary Logframe Guidance, Biodiversity Challenge Funds, Defra approved. Source for the column headed "SMART Indicators (including disaggregated targets)" and the floor of two indicators with no ceiling
  • Simplified Approval Process Funding Proposal Template, Green Climate Fund (2022). The project-level logframe sits inside the funding proposal template as annex 2a. Source for two to three outcomes stated as "generally (but not always)"
  • GFDRR's New Logical Framework, Global Facility for Disaster Reduction and Recovery (2017). Source for the banded causal diagram, the absence of indicator columns and the eighteen outcomes
  • A guide to logframes, UK Aid Match (2020). Source for baseline data within three months of start and for changes being refused within three months of an annual review assessment
  • WWF Standards of Conservation Project and Programme Management, WWF (2012), hosted by the Conservation Measures Partnership. Source for the named individual in the responsibility cell

Wider evidence

On frames rather than templates