What to Do When Your Dashboard Doesn’t Match the Raw Data

Why Data Discrepancies Happen—and What Analysts Should Do About Them

It’s a common moment of panic: you open your dashboard, pull the raw data from BigQuery or GA4, and realize — the numbers don’t match. Whether it’s transactions, sessions, or revenue, this misalignment can break trust and kill momentum. But don’t panic. This isn’t failure. It’s your cue to start debugging.

1. Understand What ‘Raw’ Actually Means

Many analysts assume their dashboard is pulling from raw, untouched data. In reality, what you’re seeing may be a transformed, filtered, or scoped version of that data. Raw data can include:

All events regardless of consent or errors
Traffic from test environments or internal users
Failed transactions or duplicate events

Your dashboard might be hiding those for a reason — but you have to know what was excluded to understand the difference.

2. Trace the Pipeline: From Source to Dashboard

Visualize the full path of your data:

GA4 / GTM Collection →
Export to BigQuery →
SQL transformation logic (filtering, joins, etc.) →
Data model for Looker Studio or Power BI →
Dashboard-level filters or calculated fields

At each layer, something might be excluded or renamed. Dashboards often filter by default — date ranges, traffic type, source — and these are easy to miss.

3. Use Controlled Comparisons

Don’t compare all sessions. Narrow the scope:

Choose one day of traffic
From one country or campaign
Use the same dimensions and filters across dashboard and query

Start small and confirm: “Is this number the same here and here?” That’s how you isolate drift.

4. Check These Common Culprits

Time zone misalignment (GA4 vs. BigQuery default UTC)
Bot filters or test traffic excluded in dashboard only
Event scoping: session-based vs. event-based logic
Currency mismatches in revenue reports
Consent mode logic filtering some sessions

5. Document the Differences

Once you’ve found the cause, document it. Create a clear table or tab in your dashboard that shows what is included vs. what’s not. This builds long-term trust and helps future you (or the next analyst).

Final Thoughts

When dashboards don’t match raw data, it doesn’t mean someone made a mistake. It means context is missing. Your job is to trace, validate, and communicate what’s happening between the source and the surface. That’s what makes you a strategic analyst — not just a report builder.

Written with support from AI tools and edited by Hisham Ghanayem. Based on real-world analysis workflows.