Onboarding Analytics: Where Are New Users Getting Stuck?

By Emily Redmond, Data Analyst at Emilytics · April 2026

TL;DR: Track onboarding step completion rates. Find the cliff (where 50%+ drop). Fix that step. Move activation aha moment earlier.

Your onboarding funnel is leaking at every step.

50% of new users never complete step 1. Of those who complete step 1, 40% don't complete step 2. By step 3, you've lost 80%.

Most founders don't see this because they don't measure it. They assume "if people finish account creation, they're onboarded." Wrong. Onboarding is a funnel. Every step has drop-off.

Here's how to find where users are getting stuck and fix it.

The Onboarding Funnel

Most SaaS onboarding has 4–6 steps:

1. Account creation - User enters email and creates password
2. Profile setup - User fills in name, company, industry
3. Configuration - User configures their workspace/environment
4. Integration (optional) - User connects third-party service (Slack, Zapier, etc.)
5. First action - User creates their first "thing" (project, report, etc.)
6. Activation - User experiences "aha moment" and sees value

Your job is to track completion at each step and find where the cliff is.

Set Up Onboarding Step Tracking in GA4

Create a custom event for each step:

// Step 1: Account created
gtag('event', 'onboarding_step_1_completed', {
  'step_name': 'account_creation',
  'step_number': 1
});

// Step 2: Profile setup complete
gtag('event', 'onboarding_step_2_completed', {
  'step_name': 'profile_setup',
  'step_number': 2
});

// Step 3: Configuration complete
gtag('event', 'onboarding_step_3_completed', {
  'step_name': 'workspace_configuration',
  'step_number': 3
});

// Step 4: First action (create first project, etc.)
gtag('event', 'onboarding_step_4_completed', {
  'step_name': 'first_action',
  'step_number': 4
});

// Step 5: Activation (experienced aha moment)
gtag('event', 'user_activated', {
  'activation_method': 'onboarding'
});

Build the Onboarding Funnel Report

In GA4:

1. Go to Explore → Funnel exploration

2. Add steps in order:

   • Step 1: onboarding_step_1_completed
   • Step 2: onboarding_step_2_completed
   • Step 3: onboarding_step_3_completed
   • Step 4: onboarding_step_4_completed
   • Step 5: user_activated

3. Set time window: 14 days (give users 2 weeks to complete onboarding)

You'll get a funnel like this:

Where's the cliff? Step 3 (27% drop) and step 4 (50% drop). Those are your problems.

Diagnose Each Step

Step 1: Account Creation

Healthy: >90% completion

If <90%, your signup form is too long or confusing.

Fix:

• Remove unnecessary fields (ask for name later)
• Single-field email entry (no "repeat email")
• No CAPTCHA unless you have spam problems
• Clear CTA button ("Create account" not "Next")

Step 2: Profile Setup

Healthy: >80% completion (some drop-off is expected)

If <70%, users don't see why profile matters.

Fix:

• Don't force it (make it optional)
• Show progress: "Step 1 of 5"
• Auto-fill what you can (if they signed up with Google, grab their name)
• Explain why (e.g., "We'll use this to personalize your dashboard")

Step 3: Workspace Configuration

Healthy: >70% completion

This is usually the first "real setup." High drop-off is normal.

Fix:

• Add tooltips explaining each setting
• Provide sensible defaults (no blank state)
• Allow skipping: "Configure this later"
• Show what happens after: "Next, you'll create your first project"

Step 4: First Action (Create First Thing)

Healthy: >50% completion

This is where users start using your product. High drop-off means they don't understand what to do.

Fix:

• Add a template: "Here's an example project. Click to explore."
• Add step-by-step guide: "1. Click 'New Project' 2. Name it 3. Click Create"
• Show what happens next: "You'll run your first analysis here"
• Video: 60-second screen recording of the action

Step 5: Activation (Experience Aha)

Healthy: >30% completion (from step 4)

If <20%, users complete action but don't see the value.

Fix:

• Make results obvious: "Your first analysis is ready"
• Add explanation: "This shows you X, which means Y"
• Next steps: "Now try filtering by Z"
• Celebration: Show a success message

💡 Emily's take: I worked with a SaaS where 50% of users completed step 4 but only 10% activated. Turns out their "first action" was setting up an integration, but they didn't show the result. Changed it to "create sample data and view results." Activation jumped to 45%. Same onboarding, different aha moment.

Compare Onboarding by Segment

Your blended onboarding funnel hides problems. Compare by segment:

By traffic source:

Organic users already understand your value. Paid ads users are confused. This tells you: your ad messaging doesn't match your product positioning.

By signup method:

OAuth and magic link have better conversion (less friction). Consider making them the default.

By time of day:

Evening signups have worse activation (maybe tired users, maybe distracted). Consider pausing ads in evenings if performance is bad.

Track Time-to-Completion

How long does each step take?

gtag('event', 'onboarding_step_completed', {
  'step_number': 2,
  'time_to_complete_seconds': 240, // 4 minutes
});

Then analyze:

If step 2 is taking 4 minutes but step 1 is 90 seconds, something in step 2 is confusing people (they're reading instructions, getting stuck, re-reading).

Fix: Simplify step 2, add tooltips, or skip it entirely.

Optimize the Onboarding Flow

Once you know where the cliff is, optimize:

1. Remove steps - Do users really need to fill in company size? Skip it.
2. Combine steps - Merge profile setup and workspace config if they're sequential.
3. Simplify steps - Profile setup: just name. Everything else can be settings later.
4. Add guidance - Video, tooltip, or instruction for hard steps.
5. Move aha moment earlier - Instead of "activate after full onboarding," let users see value in step 3.

Frequently Asked Questions

Q: How long should onboarding take?

A: Under 5 minutes for B2C. Under 10 minutes for B2B. If it's longer, you're over-explaining.

Q: Should onboarding be mandatory?

A: Mostly yes, but allow skipping. "Skip setup, configure later" gives advanced users an escape hatch.

Q: What if onboarding is a funnel within your product (not a separate flow)?

A: Same approach. Track events for each step. Build funnel. Find cliff.

Q: How often should I optimize onboarding?

A: Monthly. Small changes (word choice, button color) can improve completion by 2–5%. Run A/B tests: "One-click setup" vs. "Traditional form." Measure activation impact.

Q: Should I show a progress bar during onboarding?

A: Yes. "Step 2 of 4" reduces drop-off (users know they're almost done). "Step 1 of 7" causes drop-off (looks long).

The Bottom Line

Measure onboarding step-by-step. Find the cliff (where 50%+ drop). Fix that step.

Most cliffs are at "first action" because users don't know what to do. Add a video, template, or step-by-step guide.

Move your aha moment earlier. Don't make users complete setup before they experience value.

Emily Redmond is a data analyst at Emilytics — AI analytics agent watching your GA4, Search Console, and Bing data around the clock. 8 years experience. Say hi →