Ten years, two spreadsheets: how Echo built Perth Biodesign's decade of impact

Perth Biodesign's 10-year impact review began with incomplete historical data, including years where only annual reports survived and no raw records existed. Echo Impact, a social impact technology platform based in Perth, Western Australia, cleaned, imported and enriched the source spreadsheets, classified thousands of alumni career records, and ran a structured validation loop with the client before producing two outputs: a static anniversary report and a live, filterable dashboard.

This case study walks through the method, because the method is what makes the result repeatable for any organisation sitting on years of uneven programme data.

The brief: a 10-year review with 10 years of uneven data

Perth Biodesign, based at the University of Western Australia, marked ten years of advancing WA's health innovation ecosystem in 2026. The team wanted an impact report worthy of the milestone, covering their core programmes and the iPREP industry placement programme that came under Perth Biodesign's administration in 2020.

The data told a familiar story. Records from 2020 onwards were reasonably complete. Before that, the picture thinned quickly. For iPREP, the years prior to 2020 existed mainly as handed-over annual reports with no underlying raw data, and the programme's origins traced back to a single pilot round in 2015.

That created a genuine editorial and governance question before any analysis began: which decade were we actually measuring? Reporting iPREP from its 2015 inception meant including inherited figures from years Perth Biodesign never administered. Reporting from 2020 was more accurate to Perth Biodesign's stewardship, and it changed several charts in the report. The team, led by Director Oliver Bazzani and Chair Professor Kevin Pfleger, worked through this with us openly, and the final position was principled: the anniversary report would present iPREP metrics from 2020, while the internal dashboard would carry the full 2015 onwards dataset with filters to select any timeframe.

The hardest part of a decade-long impact report is deciding which decade you are actually measuring.

How do you build an impact report from messy historical data?

Cleaning and importing inconsistent spreadsheets

The core alumni data arrived as spreadsheets built by different people, in different years, to different standards. Names were formatted inconsistently, programmes were labelled differently across cohorts, and some fields had drifted in meaning over time. The first stage of the Echo pipeline standardises this: records are cleaned, de-duplicated, mapped to a consistent schema and imported into the platform as structured entities rather than rows.

Enriching alumni records and classifying career outcomes at scale

With clean records in place, Echo's AI-assisted enrichment layer went to work on the question the anniversary report most needed to answer: where did a decade of participants end up? This involved classifying 4,381 alumni job history rows into consistent career outcome categories, turning a sprawl of job titles and employer names into defensible metrics about sector destinations, career progression and pathways into the health innovation ecosystem.

Enrichment at this scale is where AI earns its place. Classifying four thousand rows by hand would take weeks and produce inconsistent judgements. The platform applies the same classification logic to every record, and every classification remains traceable back to its source row.

Choosing the baseline as a governance decision

The iPREP timeframe question deserves its own heading because it recurs in almost every long-running programme we work with. Baselines feel like a design detail. They are actually a governance decision that shapes the entire narrative. Inherited data, pilot rounds, administrative handovers and programme mergers all complicate the question of where a story starts. Our approach is to surface the options early, show the client what each timeframe looks like in the actual charts, and let the organisation make an informed call. In this case, we exported the dashboard under both scenarios so the team could compare them side by side before deciding.

Validating the numbers with the client

The first dashboard export went to the Perth Biodesign team for review, and the review did exactly what reviews should do: it caught mismatches. Programme manager Sandra Medic identified figures that did not align with her records, including university placement counts that were out of date in the source spreadsheets we had been given. We traced the discrepancies, she supplied the current data, and the affected figures were re-run.

We treat this reconciliation loop as a feature of the method. Historical data always contains drift between what was recorded and what actually happened, and the people who ran the programme are the only ones who can catch it. Building a formal validation pass into the timeline, with a working session to walk through analytics and data sources together, is faster and produces a more trusted report than aiming for a perfect first draft.

AI-enriched data still needs the eyes of the people who lived the programme.

One dataset, two outputs: the anniversary report and the live dashboard

A milestone report and an operational tool have different jobs, so we built both from the same underlying dataset.

The 10-year impact report is a polished, static document designed for the anniversary moment: board members, funders, university leadership and sector partners. It launched in May 2026 alongside Perth Biodesign's milestone celebration.

The dashboard is the living counterpart. It holds the complete dataset, including the full iPREP history from 2015, with filters for programme, cohort, timeframe and outcome. The team uses it internally for ongoing reporting, and components can be embedded on the public website for outward-facing transparency. When the eleventh year of data arrives, it joins the same structure rather than triggering another archaeology project.

What the data showed

The report drew on industry feedback data including Net Promoter Scores, project summaries, job offer and placement outcomes, and the enriched alumni career trajectories described above. Final published figures are the client's to share, and the report itself is available through Perth Biodesign.

The repeatable method

Stripped of the specifics, the Perth Biodesign engagement followed a five-stage pipeline that applies to any organisation with years of accumulated programme data:

1. Gather everything, including the imperfect material, and audit what exists for each year.
2. Clean, standardise and import records into a structured platform.
3. Enrich records with AI-assisted classification, keeping every judgement traceable.
4. Validate with the people who ran the programme, and expect corrections.
5. Publish twice: a report for the moment, a dashboard for the long run.

The starting conditions for this project were two spreadsheets and a set of annual reports. The finishing conditions were a validated decade of impact evidence and infrastructure that makes year eleven easy.

Frequently asked questions

What data do you need to produce an impact report? Less than most organisations assume. Participant lists, programme records and any historical surveys or feedback provide the foundation. Gaps are normal, and enrichment can recover a substantial amount of outcome data, particularly career and organisational outcomes.

How far back can historical data go? As far as records exist in any form. The practical limit is usually traceability rather than age. Where raw data is missing, summary figures from old annual reports can still be incorporated, provided the report is transparent about the difference.

What is the difference between an impact report and an impact dashboard? An impact report is a fixed document telling the story of a defined period for a defined audience. A dashboard is a live tool over the same data, filterable and continuously updated. Most organisations benefit from both, built once from a shared dataset.

How long does a project like this take? The Perth Biodesign engagement ran over roughly two months from data handover to launch, including the validation cycle with the client team.

---

Echo Impact helps foundations, universities and programme operators turn their data into evidence. Read more on longitudinal impact measurement and building a measurable theory of change, or explore our impact reporting platform.