Skip to content

Services

Data stack audit.

One week. A written plan you can execute with or without me.

Who it's for

The numbers disagree and nobody can say where it starts. GA4 has one count of sales, the ad platforms have another, and the dashboard shows a third. The warehouse was built by someone who left, and the reports get checked by hand before every meeting.

You want to know where the problem is before you pay for a fix. The audit follows the data from the tag on the site, through the warehouse, to the report, and tells you which step is costing you.

What you get

  • →A tag check. What fires on the site, what reaches GA4 and the ad platforms, and which events are missing or counted twice.
  • →A review of the warehouse and the dbt or SQL layer. Where each number is defined, what is tested, and what only one person understands.
  • →A review of the reports. Whether each number traces back to one definition, and which dashboards people actually open.
  • →A punch list ordered by what each problem is costing you. What is dangerous, what is annoying, what is fine.
  • →A walkthrough call, and recommendations you can hand to your team or hire me to carry out.

How it runs

One week, start to finish. I need view access to Google Tag Manager, GA4 and the ad accounts, read-only access to the warehouse (BigQuery, Snowflake or Databricks) and the code that builds it, and viewer access to the reports. A kickoff call first, so I know which numbers matter most to you.

If you only have part of that stack, the audit covers the part you have. Then I go dark for a few days. When I come back, you get the writeup and the walkthrough. Whatever we decide to do next is a separate conversation.

Related

Already know which step is broken? Go straight to website tracking, data cleaning and warehouse build, or reporting. Or read a recent case study.

Next step

Send me an email.

Tell me which numbers you don't trust. I'll come back in a day or two with a scope.

Email me