How do we write a theory of change?
- TimeNot established by the research.
- CostFree — it costs nothing but time.
- Doing itYou can do this yourself
Write a theory of change backwards: start with the change you want to see, then work down to what you do. It is free, nobody owns it, and for most small charities it is the only framework you need. Written on one page, it costs nothing but time. Whether you call it a theory of change or a logic model does not matter at your size.
Do this first
England and Wales: This page sets out the position in full. Scotland: the same position, with no difference in principle — a theory of change is a method, not a legal duty, and nothing about it changes at the border. Northern Ireland: the same position, with no difference in principle, for the same reason.
What does a theory of change cost?
A theory of change costs nothing. It is not a product, and no one holds a licence to it. That is worth saying first, because most pages on this topic are written by people selling help with it. The chain describes change in one group, so this step assumes you have already identified who your beneficiaries are.
The term was popularised by Carol Weiss at the Aspen Institute Roundtable in 1995. In the UK charity sector, the most widely published method is NPC’s. NPC gives its guides away free, and aims them at charities “new to this or [with] limited budgets or capacity”. It sells consulting on top, but you do not need it.
This guide’s recommendation to trustees is plain. For most small charities, a theory of change is the only framework you need. No regulator — the Charity Commission, OSCR or the Charity Commission for Northern Ireland — requires any framework, and no rule across all funders names one.
What is a theory of change?
A theory of change is a one-page chain. It explains your work backwards, from the lasting change you want to see down to what you do. It has four links, read from the top down:
- Impact — the lasting change you want to see.
- Outcomes — what has to change for people first, for that to happen.
- Outputs — what you produce or provide that brings those changes about.
- Activities — what you do.
The “why it works” part lives in the arrows. Each arrow is a claim: this leads to that. Write one line beside each saying why you believe it. That is what makes it a theory, not a list.
Separate outputs from outcomes from impact gives NPC’s exact definitions of the four words. Use them from the start, and the chain will not need rewriting later.
Theory of change or logic model? At your size, it does not matter
For a small charity it does not matter which you use. You may have met both terms and worried you picked the wrong one; you did not.
A logic model is the classic linear chain. The W.K. Kellogg Foundation’s guide runs it as inputs, activities, outputs, outcomes, impact. NPC treats a theory of change as the richer version, because it spells out why each step leads to the next.
In practice, UK guidance uses the two terms almost interchangeably. A small charity does not need to care. If you draw the chain and write down why each link holds, you have what either name describes.
Why work backwards, not forwards?
Working backwards keeps the chain about change rather than activity. The easiest way to get this wrong is to start with what you already do. Activities are concrete and familiar. The end change feels abstract.
Start there and the chain records your activities and stops. It cannot be relabelled in Step 4 or measured in Step 5, because nothing in it says what changed.
So start at the top. Ask what would be different for your beneficiaries if the work succeeded. Then ask what has to change first. Only then write down what you do.
What you do not need yet
Do not buy a framework before you have tried this. Outcomes Star, for example, is licensed and paid, and its training is compulsory. It is built for long-term one-to-one keywork, not for general measurement. Choosing a framework compares the named frameworks and what they cost.
A funder can make a particular tool a condition of its own grant. That binds that grant. It is not a sector rule. When you write a bid, you set your outputs and outcomes from this same chain.
For the smallest charities, this guide’s own reasoning puts the proportionate minimum at a one-page theory of change and three indicators. That band is Under £25,000. Deciding what you need to measure shows the reasoning, and why it is ours rather than anyone’s rule.
How to build a theory of change on one page
These numbered steps take you from a blank page to a theory of change the trustees have agreed.
- Write the lasting change at the top, in one sentence about your beneficiary group.
- List the outcomes — the changes in people that lead there.
- List the outputs that bring those outcomes about.
- List the activities that produce the outputs.
- Write one line on each arrow saying why you believe it holds.
- Check it against your beneficiary definition. Every outcome is a change in the group you defined there; if one is not, rewrite it.
- Take it to the trustees. Agree it at a meeting and record that in the minute, in the way you take and store minutes.
Keep it to one page. If it will not fit, you have written a plan, not a theory. With the chain agreed, the next step is to relabel each box by separating outputs from outcomes from impact. Template: Theory of change canvas (planned).
Common mistake: paying for a framework before trying a theory of change
Consultancy and listicle content leads with named tools and rarely says the free option is enough. The result is an unbudgeted licence and compulsory training for something a theory of change does at no cost. This guide’s recommendation: a theory of change is free, not proprietary, and enough for most small charities. No regulator requires any framework.
Common mistake: getting stuck on “theory of change” versus “logic model”
NPC treats a theory of change as the richer, causally explicit version; the Kellogg logic model is the linear inputs-to-impact chain. Guides can read as though the difference matters. In practice UK guidance uses the terms almost interchangeably, and a small charity does not need to care.
Common mistake: building it forwards from what you already do
Activities are concrete; the end change is abstract. A chain built forwards records activities and stops, so it cannot be relabelled in Step 4 or measured in Step 5. Work backwards from the change you want to see. That is NPC’s method — good practice, not a requirement.
Worked example
Wrenfield Village Hall’s committee wants to explain to a local funder why its weekly lunch club matters.
They work backwards. The change they want: older residents are less isolated. The outcome NPC would recognise is a change in that group — residents who come to the club see more people and feel less alone.
The outputs are the lunches served and the people attending. The activity is running the club each week. Beside each arrow they write one line on why they believe it.
The method costs nothing. Whether it is called a theory of change or a logic model does not matter at Wrenfield’s size.
The committee agrees the one-page chain at its next meeting and minutes it. At the top, “impact” is the lasting change the residents achieve themselves — NPC’s definition — with a note of which definition they are using. Step 4 explains why that note matters.
What you should have at the end
A one-page theory of change, agreed by trustees.
It proves the charity can say, in one place, what it is trying to change and why it thinks its work will change it. Every indicator you choose in Step 5 hangs off a link in this chain. Template: Theory of change canvas (planned).
Common questions
This guide's own theory of change canvas is planned but not yet published. You do not need one to start. A complete theory of change is a single page with the lasting change at the top, then outcomes, outputs and activities. Each arrow carries a line saying why. NPC publishes its method free.
A theory of change is your work explained backwards. It runs from the change you want to see, through the changes in people that lead to it, down to what you do. The arrows between them are the "theory": each one is a claim that one thing leads to the next, with your reason written beside it.
No. A logic model is the linear chain — inputs, activities, outputs, outcomes, impact. A theory of change adds the reasons each link holds. UK guidance uses the two almost interchangeably, so one page that does the job is enough, whatever you call it.
A small charity's theory of change fits on one page. For a village hall lunch club, less isolation is at the top. Residents seeing more people is the outcome, lunches served are the output, and running the club is the activity. Trustees agree it at a meeting and minute it.