Skip to main content

Melbourne · Evaluation and evidence practice

Know what
your evidence is
really telling you.

We help organisations understand what their evidence actually supports, and use it well in evaluation, decisions and reporting. Our work is strongest today across health, community services, government and funding.

Claim trace · Illustrative example

Not a client result

CLM-0031v3

The programme improved young people’s mental wellbeing.

Under review. Five sources are linked. One is present, two are partial, one reading is contested and one source is missing.

SRC-0142 · PROGRAMME RECORDS · PRESENT

Attendance and session logs

What it supports
Young people attended sessions across both programme years, and the logs show what was delivered and when.
Where it falls short
Attendance shows contact, not change. On its own it says nothing about wellbeing.
v3 · reviewed 12 Jul 2026 · 3 limits open · 1 reading contestedPresentPartial coverageContestedMissing
Illustrative claim trace. One claim linked to five sources, each with its coverage and limits. Not a client result.

This is how we read every claim we’re asked to stand behind.

Record IDs, dates and sources are illustrative. Lines show connection only, not evidence strength.

How people arrive

Sound familiar?

Nobody asks for an impact intelligence system. They arrive with a live pressure.

  • We have loads of data but don't know what it means.
  • The funder wants us to demonstrate outcomes.
  • We think the programme works, but we can't prove it.
  • We've got three reports saying different things.
  • We're collecting heaps of information nobody uses.
  • We need an evaluation, but we're not even sure what we should measure.
  • Our board wants to know whether we're actually making a difference.

Services

What we do

Evaluation is part of the work. So are the evidence systems, reporting decisions and organisational capability around it.

  • Evaluation design and delivery

    Finding out what worked, for whom and under what conditions.

  • Outcomes frameworks and theory of change

    Agreeing what you're trying to change and how you'd know.

  • Measurement and evidence improvement

    Fixing data you collect but can't use.

  • Impact reporting and decision support

    Saying what the evidence supports to boards and funders, no more and no less.

  • Evidence governance and capability

    Helping your team keep evidence traceable and use it without us.

See what we do

CIIS

Practice first

A connected evidence environment

Impact Intelligence Lab is the practice. CIIS is evidence infrastructure used inside some engagements. It does not decide what is true or turn weak evidence into certainty.

Evidence surfaces. People decide. Limits stay visible.

Explore CIIS

Evidence coverage · Illustrative example

PRG-007 · 4 claims · 5 evidence types

ClaimRecordsSurveyInterviewsIndependent reviewOwn accounts
CLM-0031Under reviewThe programme improved young people’s mental wellbeing.
CLM-0028ReviewedYoung people attended regularly across both programme years.
CLM-0029ReviewedThe programme was delivered as designed.
CLM-0035DraftFamilies noticed changes at home.

CLM-0031 × Survey · SRC-0157 · Partial coverageFewer people completed the exit survey than the entry survey, and there is no comparison group.

PresentPartial coverageContestedNot examinedMissing

Shapes show coverage only. They are not a score or a strength rating.

Every ID, date and source here is illustrative. This is not a client record.Lineage is not permission to publish.

How we work

Experience can be evidence.Lived Experience is expertise.Participation must carry influence.

Did people merely provide information, or could they influence what it meant and what happened next?

We don't only ask whether there is evidence. We ask whose evidence counts, how it has been interpreted, and whether the people closest to an issue had any influence over what happened next.

  • Preserve people's meaning, context and consent.
  • Distinguish what someone said from our interpretation of it.
  • Keep disagreement and missing voices visible.
  • Allow people to challenge interpretations.
  • Where possible, show how their contribution influenced what happened next.

Human Meaning Plane

One account, held in four roles: evidence, context, interpretation and governance. The words carry the meaning. Lived Experience stays in the context and influence of the account. It is not a checkbox.

  1. 01

    Voice

    Evidence

    What was said, kept distinct from what a reviewer later made of it.

  2. 02

    Context

    Context

    The setting the account came from, including what the record leaves out.

  3. 03

    Consent

    Governance

    Whether the account can be used, and for what. Consent is not assumed.

  4. 04

    Provenance

    Evidence

    Who the account is connected to, and how it entered the record.

  5. 05

    Disagreement

    Interpretation

    Different perspectives stay visible. They are not averaged into one view.

  6. 06

    Missing voices

    Context

    People closest to the issue who are not yet in the account.

  7. 07

    Who interpreted

    Interpretation

    A named human reader. Interpretation is not an automated score.

  8. 08

    Influence

    Governance

    Whether people could change what the evidence meant and what happened next.

How we treat evidence

Audiences

Who we help

  • Community organisations and NGOs
  • Councils and local government
  • Primary Health Networks and health programmes
  • Funders, philanthropy and impact investors
  • Evaluation and strategy teams

Facing the same evidence problem in another sector? Talk to us.

Who we help

Alan McCarthy, founder of Impact Intelligence Lab

Who you'd work with

Alan McCarthy

Founder

  • Close to 20 years across community mental health, youth services, primary health commissioning and local government.
  • Supported commissioning, provider performance and outcomes reporting across six headspace services within a primary health portfolio.
  • Established and evaluated a youth outreach programme that delivered more than 1,000 occasions of service in its first year.
  • Led national training and a whole-of-community youth needs assessment.

Bring the question. We work out the right response. CIIS only if it fits.

About Alan

Learning horizon

What we know, then the next question

What we know → Emerging questions → Worth examining next

  1. What we know

    An illustrative claim is on the table, with some sources, context and limits attached. It is still under review.

  2. Emerging questions

    Which voices are not in the record? What would a different reader make of the same material?

  3. Worth examining next

    A consented way to hear from people the current records do not include.

The path stays open. A new question does not raise evidence strength, reporting permission or claim authority.

Discuss the evidence in front of you.

Impact Intelligence Lab is the practice. CIIS supports some of that work. The next step is a conversation.

Discuss your evidence challenge