Top 10 GA4 Mistakes and How to Avoid Them

By Emily Redmond, Data Analyst at Emilytics · April 2026

TL;DR: Avoid these: ignoring custom events, mixing test and production traffic, wrong attribution models, missing User-ID, poor data quality, not excluding internal traffic, misunderstanding event limits, wrong conversion setup, bad segment logic, and ignoring data retention.

GA4 is forgiving. You can set it up wrong, and it will still collect data. But the data will be wrong, and you'll make bad decisions based on it.

Here are the top mistakes I see and how to fix them.

Mistake 1: Not Tracking Custom Events

The problem: You install GA4 and rely on automatic events (pageviews, clicks). You never set up custom events for actions that matter to your business.

Result: You know traffic, but not business metrics. "Did users sign up? Did they use the feature? Did they upgrade?"—no idea.

Fix: Map your key actions to events immediately.

• SaaS: Sign-up, feature use, upgrade, cancel
• Ecommerce: Add to cart, begin checkout, purchase
• Publishing: Article read, video play, newsletter signup

Set these up in week 1, not week 6.

Mistake 2: Mixing Test Traffic and Production Data

The problem: You're testing the site. Your team loads pages, clicks buttons, completes fake transactions. GA4 counts all of this as real traffic.

Result: Your conversion rate is inflated. Your traffic is inflated. Your metrics are garbage.

Fix:

1. Use a separate data stream for testing: Development vs. Production streams.
2. Exclude internal IP addresses: Use a data filter (Admin → Data filters) to exclude your office IP.
3. Use an internal tag: Tag test traffic with a parameter (?test=true) and filter it.

Mistake 3: Using the Wrong Attribution Model

The problem: Last-click attribution is default. You assume the last channel before conversion is the most valuable. But it's not.

Result: You over-invest in retargeting. You under-invest in awareness. Your budget allocation is wrong.

Fix: Choose an attribution model that reflects your business.

• Long sales cycle (B2B, SaaS): Time-decay or linear
• Short cycle (ecommerce): Linear or time-decay
• If you have GA4 360: Data-driven

Then budget based on fair credit, not last-click lies.

Mistake 4: Not Setting Up User-ID Tracking

The problem: Your site has logged-in users. You never send User-ID to GA4. GA4 sees the same person on mobile and desktop as two users.

Result: User counts are inflated (2-3x higher than reality). Cross-device journeys are invisible. Audience targeting is based on fake data.

Fix: Set up User-ID today.

1. Send your internal user ID to GA4 (hash it for privacy)
2. Enable User-ID in GA4 settings
3. Verify in Real-Time

Done in 15 minutes, saves months of confusion.

Mistake 5: Poor Event Data Quality

The problem: You fire events, but inconsistently.

Examples:

• Sometimes sign_up, sometimes signup (inconsistent naming)
• You fire purchase event without item details (missing parameters)
• You fire events at random times (no standardization)

Result: Data is messy. You can't reliably segment or filter. Duplicate event names create confusion.

Fix:

1. Standardize naming: Use snake_case, consistent across all events
2. Always include parameters: Every purchase has value, currency, items
3. Document your events: Keep a list of event names, when they fire, what parameters they have
4. Test before deploy: Use a test property to verify events fire correctly

Mistake 6: Not Excluding Internal Traffic

The problem: QA runs 50 test transactions/day. Engineers reload pages 100x. This all gets counted.

Result: Conversion rate is inflated by 50+ per day.

Fix: Set up IP filters.

1. Admin → Data filters
2. Create exclude filter for office IP
3. Also exclude VPN IP if team works remote

Mistake 7: Misunderstanding GA4 Event Limits

The problem: GA4 allows up to 500 custom events per property and 25 parameters per event. You try to track everything.

Result: You hit limits. You can't add new events. You're blocked.

Fix: Track the vital few events well, not everything poorly.

• Ecommerce: 5-10 events (view, cart, checkout, purchase, etc.)
• SaaS: 5-10 events (signup, feature use, upgrade, etc.)
• Plus custom events for your business logic

Quality over quantity.

Mistake 8: Marking Too Many Events as Conversions

The problem: You mark 30 events as conversions. Every click, every pageview, every page load is a "conversion."

Result: Conversion reports are noise. You're not tracking what matters.

Fix: Mark only truly important actions as conversions.

• SaaS: Sign-up, upgrade
• Ecommerce: Purchase
• B2B: Demo request, trial signup

2-3 conversions per property. Max.

Mistake 9: Wrong Funnel or Segment Logic

The problem: You build a funnel with wrong conditions.

Example: Funnel is "Add to cart" → "Purchase." But you forgot to check if the same user did both steps. GA4 counts different users.

Result: Funnel metrics are wrong.

Fix: When building funnels, ensure each step is in the right order AND from the same user.

Use the funnel exploration and verify the numbers make sense. Test with a small date range first.

Mistake 10: Ignoring Data Retention

The problem: You keep 2-month retention (default). After 2 months, user-level data is deleted.

Result: You can't do 6-month cohort analysis. You can't see seasonal patterns. Historical analysis is impossible.

Fix:

1. For short analysis: 2 months is fine.
2. For long-term: Extend to 14 months or export to BigQuery.
3. BigQuery is better: Unlimited retention, free up to 1TB/month.

Bonus Mistakes

Not Testing Before Production

Always test custom events in a test property first. Verify they fire correctly before going live.

Using Auto-Generated Default Reports Only

Standard reports are surface-level. Use Explorations for real analysis.

Assuming Correlation = Causation

"Traffic went up on the same day we changed the homepage." Correlation. You need to measure causation with A/B tests.

Not Documenting Your Setup

Document:

• Custom events and what they track
• Custom dimensions and metrics
• Conversion events and their values
• Attribution model used
• Filters and segments

Future you (and your team) will thank you.

Relying on Single-Day Metrics

Always look at trends, not single days. One day's data is noise.

Audit Checklist

Every 3 months, audit your GA4:

• Are all important events still firing?
• Are custom dimensions and metrics still accurate?
• Is internal traffic still excluded?
• Are conversion events still defined correctly?
• Is attribution model still appropriate?
• Are there any new issues or anomalies?

The Bottom Line

These mistakes are preventable. Spend an hour on setup and documentation now, and you'll have clean, reliable data for years.

Bad data is worse than no data. If you think your GA4 is broken, fix it. Don't just accept garbage metrics.

Emily Redmond is a data analyst at Emilytics — the AI analytics agent that watches your GA4, Search Console, and Bing data around the clock so you never miss what matters. 8 years of experience helping founders and growth teams turn data noise into clear decisions. Say hi →