The Impact Data Audit
An independent assessment of whether your data can prove what your programs actually do.
The question
When a funder asks “what outcomes did our grant produce?” — can you answer from your systems, or do you assemble the answer by hand?
Most organizations between $5M and $50M don’t know which one they are until they look. You’re large enough to carry serious reporting obligations and small enough that no one owns the data end to end. The honest answer is usually somewhere in between, and the interesting part is where.
That’s what the assessment establishes. Not a diagnosis I bring with me — a finding I reach after looking.
What I assess
Over two weeks I trace your data from program delivery through to funder report, and test it against five criteria:
- Provenance — for each number you report to funders, can it be traced back to a source record? Which numbers can’t?
- Reproducibility — if two staff members produced the same report independently, would they agree? Where do the discrepancies come from?
- Cost to produce — actual staff-hours consumed assembling grant reports, board packets and compliance filings.
- Coverage — what your programs do versus what your data records. The gap between them is what you currently cannot prove, regardless of how good the programs are.
- Continuity — what happens to your reporting if the person who knows the spreadsheet leaves.
What you receive
A written assessment, findings first: each criterion rated, each material weakness stated plainly, with the evidence behind it. Where I find nothing wrong, I say so — a clean finding is a real result, and worth having in writing before your next funder conversation.
The report separates what’s broken from what’s merely inelegant, and flags which weaknesses would actually survive scrutiny if a major funder examined them.
Your cost: two or three conversations with the people who touch the data, and read-only access. No system changes, no software, no commitment.
Independence and disclosure
You should know my interest before you start, not after.
I do build data systems, and some assessments surface problems I’d be a reasonable person to fix. Others don’t. I write the findings before I know which kind it is, and I’ll hand you a report that recommends doing nothing if that’s what I find — that outcome is common enough that I’ve stopped treating it as a failed engagement.
If you’d rather remove the question entirely, take the report to someone else to implement. It’s written to be usable that way.
Who I am
Ravikanth Andhavarapu — independent director and audit committee chair of a BSE-listed company; 15+ years building data platforms and engineering teams. Currently acting CTO at Wraffle; formerly Head of Strategy at Biliti Electric and founding engineering lead at several ventures. IIT Madras.
I work with a small number of organizations at a time.
Next step
A 30-minute conversation. Bring your most recent funder report and I’ll tell you on the call whether an assessment would find anything worth the two weeks.