Diagnose
See what's broken.
Know what to fix first.
The CFO is quoting one revenue number. The VP Marketing is quoting another. Nobody can defend the gap. Two weeks, fixed scope, and you walk away with a written diagnosis of what's broken across GA4, attribution, the warehouse, and the dashboards your team has stopped trusting — a prioritized fix list, a build-buy-kill call on every tool, and a defensible answer to which number is right.
What we do
Five concrete artifacts, plus an optional sixth. Each one is yours to keep, hand to a CFO, hand to a future vendor, or take to the next firm. Nothing here is a slide deck.
01
Fifteen to twenty-five pages, PDF plus editable source. Executive summary, system inventory, prioritized findings, and the build-buy-kill table. Long enough to be defensible, short enough that your CFO will actually read it.
02
A working spreadsheet. Each row: severity (P0/P1/P2), area, what's broken, recommended action, estimated effort, expected payback. Hand it to your team and they'll know what to ship first on Monday.
03
Every tool in your stack, mapped against what it's supposed to do and what it actually does. Costs, owners, overlaps, dead software. The document your next vendor renewal conversation should start from.
04
The metrics your executives are quoting, traced from source through transformation to the dashboard. Definitions that don't survive the trace are flagged, and a corrected definition is proposed in writing.
05
One page, every line item gets a verdict. Killed tools come with a migration note; bought tools, a configuration note; new layers, a scoping note. No vendor relationships, no kickbacks — an honest call your finance team can defend.
06
Optional, delivered alongside the audit, not bundled into the fee. Take it to us, take it in-house, or take it to a competitor. The audit is yours either way.
Method
Ten working days, mapped. We tell you what we're doing, when, and what you'll see at the end of each phase.
Days 1–2
A 60-minute kickoff with your executive sponsor confirms scope, names the stakeholders, and locks the communication cadence. You hand over read-only credentials for the systems in scope. We send a one-page pre-read so every interviewee knows what we'll ask and what we'll deliver.
Days 3–6
Four to six interviews of 45–60 minutes each — executive sponsor, data owner, marketing owner, finance owner, and a couple of report consumers. In parallel, we trace every executive metric from BI back to source and inventory every tool in the stack. Roughly six hours of your team's time, distributed.
Days 7–9
Every broken thing gets graded P0, P1, or P2 — revenue-at-risk, trust-eroding, or polish. The ranking is pressure-tested against your roadmap and your team's capacity, so the fix list you receive is one you can actually ship against, not a wishlist.
Days 10–12
Your data owner sees the draft first on day 10. One round of substantive edits — factual corrections, missing context, scope disagreements. We change what's wrong; we don't change defensible severity grades to soften the read for leadership.
Days 13–14
Final audit, fix list, system inventory, measurement-plan review, and build-buy-kill table delivered on day 13. A 60-minute walk-through with your sponsor and data owner on day 14. A short clarifying-questions window is included by default — defined in the SOW.
Approach
Diagnose is a methodology, not a tool. The work is interviews, lineage traces, and a defensible verdict on every line item in your stack.
Outcomes
Four anonymized patterns from prior audits. Different industries, same shape: a written diagnosis that re-grounded the conversation and made remediation surgical instead of speculative.
A multi-banner retailer
Three reporting systems were quoting three different revenue numbers to three different executives.
The audit traced each number back to source, named the defensible one in writing, and killed the other two. The fix list put a warehouse rebuild ahead of any new dashboards — sequenced so the next quarter's reporting could not contradict itself.
A digital bank
Fourteen source systems were feeding ad-hoc SQL across 60+ analysts, with no agreed metric definitions.
The audit recommended killing two overlapping BI tools and rebuilding the warehouse layer before any new reporting. The fix list sequenced warehouse work first, BI second — and gave the data leader a written case to take to the CFO.
A multi-brand CPG
A measurement plan hadn't survived the last CMO change. Three brands were reporting attributed revenue on three different models.
The audit named the gap, traced each model back to its assumptions, and scoped the rebuild so the next plan would hold up through the next two pivots. The fix list put the attribution rebuild ahead of any new media-mix work.
A healthtech marketplace
CAC and LTV were being calculated by two analysts using two different definitions of `customer`.
The audit traced both definitions to source, picked the defensible one, and put the metric layer rebuild at the top of the fix list. Cohort economics came second — once the underlying definition was settled, the cohort view became tractable.
Next
Tell us what's working, what isn't, and what you've already tried. If the shape is right, we'll scope the two weeks and book the audit.