Data Platform Audit
In one week: exactly what to fix, what to cut, and what your data platform should cost
A fixed-price teardown of your data stack: 12–20 hours of senior work over one week — a ranked fix-list, a cost model comparing your spend to a lean rebuild, and a small proof-of-concept on your own stack. $3,000, and if day one shows no credible opportunity, you pay nothing.
The problem
Your data stack grew by accretion, the warehouse bill keeps climbing, and no one can say cleanly what to fix first.
What you get
- Ranked fix-list — every issue, with effort and cost estimates
- Cost model — your current spend vs a lean rebuild
- Small proof-of-concept, built on your own stack
- Written roadmap, yours to keep either way
- $3,000 credited in full toward any follow-on build
Guarantee
I only recommend the audit after seeing your cost export — and if day one of the audit shows no credible opportunity, you pay nothing.
Built for B2B SaaS and product companies, 30–200 people, running a cloud warehouse with zero to three data people and no senior platform owner — or one AI pilot that never reached production. Over one week I go through your ingestion, transformation, warehouse and dashboards, find where data is late, wrong or silently missing, and map where the money is actually going. Your side of the effort: read-only access and about two hours of your team's time — I do the rest. I take two audits a month.
What I look at
Four areas, worked against your real systems rather than answered from memory. They are the same four sections as the free teardown; the difference is that I run them for a week against your stack, with your cost data in front of me.
- Where the money goes — your warehouse cost export and query history, read for spend that buys nothing: warehouses that never suspend, dashboards refreshing on a timer for a team that stopped opening them, dev and CI running on the production meter.
- Whether the numbers are right — where data arrives late, lands wrong or goes silently missing, and whether anything would tell you before a customer or a board meeting does.
- Whether it runs without you — how long a failed pipeline takes to reach a human, and how much of that depends on someone remembering to look.
- Whether you can change it safely — how much of the platform exists only as habit: untested transformations, hand-run fixes, changes that go straight to production.
How the week runs
- Scoping call — free, 30 minutes. You send a warehouse cost export, and I tell you whether an audit is worth $3,000 to you. If it is not, I say so and we stop there.
- Day one — read-only access and a first pass over the stack. If there is nothing worth finding, the audit ends there, unpaid.
- The rest of the week — ingestion, transformation, the warehouse and the dashboards in detail; the cost model; and a small proof-of-concept of the highest-value fix, built on your own stack.
- The deliverables — the ranked fix-list, the cost model, the proof-of-concept and a written roadmap, all in writing and yours whether or not we work together again.
What a fix-list entry looks like
Every entry names the problem, what it costs, the fix and the effort, and the list is ranked so the first line is the best return on an hour. An illustrative entry, not from a client:
- Problem — the transformation warehouse auto-suspends after 10 minutes, but a scheduled job queries it every 9, so it never suspends.
- What it costs — the warehouse is billed as if it were always on, every hour of the month, and the settings page shows nothing wrong.
- Fix — batch the job to run hourly, or cut the suspend window to 60 seconds.
- Effort — under an hour.
Who it is not for
- A team with a senior platform owner already in the seat. Give them the free teardown; they can run it themselves.
- A company with 40 hours a week of data work ahead of it and the budget and months to hire. Hire.
- Anyone who wants the audit to confirm a decision already made. The roadmap says what the numbers say.
Want to try it yourself first?
The questions I work through are published in full — all 18 of them, with what a bad answer sounds like. Run the teardown free →
How the engagement runs — hours, payment, contracts — is written down on the how-I-work page.
Common questions
- What does a data platform audit include?
- Four areas, worked against your real systems: where the money goes, whether the numbers are right, whether the platform runs without you, and whether you can change it safely. In practice that means your ingestion, transformation, warehouse and dashboards, plus the warehouse cost export. You get a ranked fix-list with effort and cost estimates, a cost model comparing your spend to a lean rebuild, a small proof-of-concept on your own stack, and a written roadmap.
- How much does a data platform audit cost?
- $3,000, fixed, for one week of work. The scoping call before it is free, and day one carries a guarantee: if the first pass finds nothing worth fixing, you pay nothing. If you go on to a build, the $3,000 is credited in full.
- How long does a data platform audit take?
- One week, and 12–20 hours of my time. Your side is read-only access and about two hours of your team's time.
- What access do you need for the audit?
- Read-only access to the systems in scope — the warehouse, the pipelines and the dashboards — and your warehouse cost export. Nothing in the audit needs write access.
- What if the audit finds nothing worth fixing?
- That is what the scoping call and day one are for. I only recommend the audit after seeing your cost export, and if day one shows no credible opportunity, the audit ends there and you pay nothing.
- What happens after the audit?
- Whatever the roadmap says is worth doing. You can run it yourselves — it is written to stand on its own. I can build the highest-value fixes as a Data Platform Build ($15–40k, with the audit fee credited in full). Or I can stay on in a fractional advisory seat at $3–4k a month, with a three-month minimum.