Quick answer: A Google Analytics audit is a structured review of your Universal Analytics property to confirm that data is collected correctly, filtered sanely, and tied to meaningful goals. Work through property access, tracking code coverage, view filters, referral exclusions, goals and ecommerce, site search, UTM hygiene, privacy settings and dashboards. Fix foundational issues before you debate attribution stories. Clean measurement is the base for every marketing argument that follows.
Why audits pay for themselves
Teams argue about channel performance while:
- The thank-you page goal is double-firing
- Office traffic inflates sessions
- Payment gateways show up as referring sites
- Mobile apps or subdomains are missing the tag
- Goals still track an old URL from two redesigns ago
An audit stops those arguments by fixing the instrumentation. If you are still installing the basics, read the Google Analytics setup guide first, then return here.
1. Account and access
- List users and roles; remove people who left
- Confirm the right company owns the account (not a former agency alone)
- Check that billing and support emails are current
- Note linked products: Search Console, Google Ads, Data Studio
Linking Search Console helps query analysis; see linking Search Console if that connection is missing.
2. Properties and views
For each major site:
- One property per distinct product/site unless you have a deliberate roll-up plan
- Keep a raw unfiltered view
- Keep a working filtered view for reporting
- Optional test view for experiments
Never filter the only view you have. Filters permanently alter data in that view.
3. Tracking code coverage
- Verify the Universal Analytics tag (analytics.js or gtag.js) on all templates: home, content, checkout, docs, landing pages
- Check tag managers for duplicate UA tags firing together
- Confirm single-page apps push virtual pageviews or use proper history tracking (SPA tracking)
- Spot-check thank-you pages and payment return URLs
View-source and Tag Assistant beat assumptions.
4. Filters and IP exclusions
- Exclude office and agency IPs where practical
- Lowercase campaign parameters if your team is inconsistent with casing
- Exclude known referral spam hostnames if it still appears
- Document every filter with an owner and date
5. Referral exclusion list
Add your own domains and payment processors so returning visitors from PayPal or similar do not start new sessions as referrals. Missing exclusions break conversion paths and attribution.
6. Goals and events
- List business outcomes: leads, sales, signups, key downloads
- Match goal URLs to the live thank-you pages
- Use events for clicks that do not load a new page (event tracking)
- Mark non-interaction events correctly so bounce rate stays meaningful
- Destination goals should not conflict with ecommerce purchase tracking on the same hit without a plan
Review goals setup if goal types are unclear.
7. Ecommerce (if you sell)
- Enable ecommerce (and enhanced ecommerce if you use it) on the right view
- Verify purchase data on a test order in a staging property or carefully in production
- Check currency settings
- Confirm refunds handling expectations with finance (UA has limits here)
8. Site search
- Turn on site search tracking if your site has a query parameter
- Review top search terms monthly for content gaps and UX issues
9. Campaign and UTM hygiene
- Publish a UTM naming standard (source, medium, campaign)
- Spot-check that email tools and ads use it
- Watch for (not set) and self-referrals from bad tags
- Keep a shared spreadsheet or naming doc
10. Privacy and retention
- Review data retention settings
- Confirm IP anonymization if that is part of your policy (IP anonymization)
- Align consent banners with how tags load (cookie consent and GDPR considerations)
- Avoid sending unnecessary personal data in custom dimensions
11. Content and structural sanity
- Hostnames: only expected domains
- Soft 404 pages producing titles that look like errors
- Parameter explosions creating duplicate page entries (trim query parameters in view settings when safe)
12. Reporting layer
- Rebuild one executive dashboard that uses the cleaned view
- Connect Data Studio only after the property is trustworthy
- Annotate major site launches in Analytics so charts have context
- Schedule a quarterly mini-audit so entropy does not return
Suggested audit order for a one-day pass
Morning: access, tags, views, filters, referrals.
Afternoon: goals, ecommerce, UTMs, privacy, dashboard smoke test.
End of day: written issue list with severity (blocks decisions / mild / tidy-up) and owners.
Red flags that mean “stop reporting until fixed”
Pause executive reporting debates when you see:
- Goals firing on every pageview
- Huge spikes of traffic from a single hostname you do not own
- Ecommerce revenue off by an order of magnitude vs the store
- Zero mobile traffic on a site that clearly has mobile users
- Multiple UA properties collecting overlapping subsets of pages
Those are instrumentation failures. Continuing to optimize campaigns against them wastes budget.
Sample findings write-up (short form)
Good audit notes look like this:
- Critical: Duplicate gtag on checkout thanks to theme and GTM both firing UA-XXXX. Fix by removing theme hardcode. Owner: dev. Due: this week.
- High: Thank-you goal still points to
/thanks-old/. Update destination. Owner: marketing. - Medium: No site search tracking; enable for
?s=. Owner: analytics. - Low: Dashboard still uses the unfiltered view. Point to filtered view.
Severity plus owner plus due date turns an audit into motion.
Cross-checks with Google Ads
If you run ads, verify that Google Ads conversion tracking and Analytics goals are not telling incompatible stories without explanation. Linked accounts help, but someone still has to reconcile definitions. See AdWords conversion tracking and linking AdWords when those pieces are part of the stack.
Annotation habit
After the audit, add Analytics annotations for the fix dates (tag changes, filter updates, goal edits). Six months later those notes explain weird chart shapes better than memory. Encourage marketers to annotate campaign launches the same way.
FAQ
How often should we audit?
A full pass after major redesigns or tag manager rebuilds, plus a lighter quarterly check.
Can a plugin audit replace this?
Plugins catch some tag issues. They will not replace human review of goals and business definitions.
Is this the same as a marketing performance review?
No. This checks whether the data is fit for use. Performance analysis comes after.
We use Google Tag Manager. Does that change the checklist?
The themes stay the same; you audit triggers and tags inside GTM as well as the Analytics views.
Trust the numbers, then debate the strategy
A Google Analytics audit is maintenance, not a luxury. Work the checklist, fix the high-severity gaps and only then argue about channels and creative. If you want help validating tracking and building reports your team will actually use, see Know what’s working.
