How do we know if anything changed?

England and Wales: This page covers setting a baseline in full. It is practice, not law. Scotland: the same position, with no difference in principle. Northern Ireland: the same position, with no difference in principle.

Is there a standard method for baseline data?

No. No official or published source sets out how a charity your size should collect baseline data. That is true in England and Wales, Scotland and Northern Ireland alike, and no one publishes a template for it. We looked. This is a craft page.

So what follows is practice, and we say where it comes from. One point is sourced and it is the important one: why a baseline matters at all. This step assumes you have already chosen your indicators: the baseline is the first reading of each one.

Why does a baseline matter?

Without a starting point you can only describe people; you cannot show that anything changed. At the end of a course you can ask people how confident they feel. The answers describe them now. They say nothing about change, because you do not know where anyone began.

A baseline is the first reading of your outcome indicators: the same question, asked at the start. With both readings, you can say what moved.

That matters more than it sounds. It is what makes an honest claim about your work possible. NPC’s Code of Good Impact Practice puts the goal plainly, verbatim. “It’s more realistic to talk about the contribution you’ve made rather than to try to attribute impact to your work alone.”

A contribution claim says: things moved, and we were part of why. You cannot make even that claim without a starting point. Without one, you can describe what you did. You cannot say what changed.

What a baseline does not settle

A before-and-after difference is not proof that you caused it. Some of the change would have happened anyway. People get older, find work, make friends, and move on for reasons that have nothing to do with you.

The SROI method has names for what a baseline leaves open:

• deadweight — the change that would have happened anyway
• displacement — change moved from somewhere else, not newly made
• attribution — the share that other people and organisations caused
• drop-off — how much of a lasting change fades in later years

You do not have to calculate any of these. But it helps to know they exist, so you do not claim more than your numbers show. A baseline tells you what moved. It does not tell you why.

What you can honestly claim from what moved is a later step: work out what you can honestly claim.

When should you take the baseline?

Take it at the first session, because the baseline moment passes quickly. Delivery usually starts before anyone thinks about measurement, and by the second session the starting point has gone. For a grant-funded project, the baseline point also belongs in the bid’s monitoring and evaluation section.

So build it into the first contact:

• Use the same measure you will use at the end. The same questions, the same scale, the same wording. A different question at each end gives you two facts, not a change.
• Record it per person, or per group. If you work with groups where individuals cannot be tracked, a group baseline is still a starting point.
• Keep it where the end measure will go. A baseline in one notebook and an end score in another is how the two never meet.
• Keep it short. A few questions people can answer in a minute will be asked every time. A long form will not.

How systematic to make it depends on your size. This is our view, not a published standard.

In the income band £25,000 to £100,000, one consistent before-and-after measure for your main activity is a sensible level. In the band £100,000 to £500,000, we would collect baseline data systematically across the work. Choosing your indicators sets out the reasoning behind the bands.

With the baseline built into first contact, the next step is to design how you will collect the data for the rest of the work.

Common mistake: describing what you did instead of what changed, because there was no starting point

Delivery starts before anyone thinks about measurement, and the first session — the baseline moment — passes. A baseline is what makes a contribution claim possible at all.

Without one you cannot even begin to address deadweight, the change that would have happened anyway. With one you can at least say what moved. This is good practice, drawn from NPC and the SROI vocabulary, not a legal requirement, and no official source sets a baseline method.

Common mistake: treating a before-and-after difference as proof that you caused it

A movement between two measurements looks like a result. A baseline settles what moved, not why. SROI’s four terms — deadweight, displacement, attribution, drop-off — name what a baseline alone leaves open. The honest claim is contribution, which work out what you can honestly claim covers.

Worked example

Tollgate Youth CIO wants to show a funder that its confidence course works.

Last year it ran an end-of-course survey. The survey described the young people at the end. It could not show change, because nobody recorded where they started.

This year the session lead asks the same three questions at the first session and the last. Each young person’s first answers go on one line of the course record, and the last answers go on the same line.

Now Tollgate can say what moved, and only that. The SROI vocabulary names what the baseline does not settle — what would have happened anyway, and what others contributed. No published source sets a baseline method or sample size for a charity Tollgate’s size, and this example uses none.

The outcome: a one-line baseline record per participant, taken at the first session and stored where the end-point measure is stored. Good practice, not a legal requirement.