Funnel Analysis: Finding and Fixing Where You're Losing Customers

The Leaky Bucket Problem

Imagine pouring water into a bucket with holes in it. You can pour faster (more traffic), or you can fix the holes (funnel optimisation). Most businesses keep pouring. Few examine the bucket.

Here's a typical scenario:

• 10,000 people visit your website this month
• 3,000 view a service or product page
• 500 start filling out a contact form or begin checkout
• 80 complete the form or purchase
• 25 become paying customers

That's a 0.25% visitor-to-customer conversion rate. Not unusual — but where exactly are people leaving, and why?

Funnel analysis is the process of mapping each step a visitor takes from first touch to final conversion, measuring how many people progress through each step, and identifying the stages with the largest drop-offs.

Fix the biggest leak first, and everything downstream improves.

---

Mapping Your Funnel

Before you can analyse drop-offs, you need to know what your funnel actually looks like.

The Website Conversion Funnel

Most websites follow a variation of this:

Stage 1: Landing Visitor arrives on the site (homepage, blog post, landing page, product page).

Stage 2: Engagement Visitor explores (views additional pages, scrolls, clicks on elements).

Stage 3: Interest Visitor shows intent (visits pricing page, views case studies, checks specific service/product pages).

Stage 4: Action Initiation Visitor begins converting (starts filling out a form, adds to cart, clicks "Book a call").

Stage 5: Conversion Visitor completes the action (form submitted, purchase completed, booking confirmed).

Stage 6: Post-Conversion Customer follows through (shows up to the call, pays the invoice, doesn't return the product).

E-commerce Funnel

More granular:

1. Product listing / category page view
2. Product detail page view
3. Add to cart
4. Begin checkout
5. Enter shipping info
6. Enter payment info
7. Complete purchase

B2B Lead Generation Funnel

1. Landing page visit
2. Content engagement (scroll depth, time on page)
3. CTA click ("Book a call," "Get a quote")
4. Form view
5. Form start (first field interaction)
6. Form submission
7. Qualified lead (meets criteria)
8. Sales conversation
9. Proposal sent
10. Deal closed

The more granular your funnel definition, the more precisely you can identify where the problem is.

---

Setting Up Funnel Tracking

GA4 Funnel Exploration

Google Analytics 4 has a built-in funnel analysis tool.

Setup:

1. Go to GA4 → Explore → Create new exploration
2. Choose "Funnel exploration" template
3. Define your steps (e.g., page_view on /pricing → click on "Contact" button → form_start → form_submit)
4. Set the date range and any segments

What you get:

• Visual funnel showing the number and percentage of users at each step
• Drop-off rate between each step
• Ability to segment by device, traffic source, geography, etc.
• Breakdown of where users go after abandoning each step

Open vs. closed funnels:

• Closed funnel: Users must complete steps in order (step 1 → 2 → 3). Useful for defined processes like checkout.
• Open funnel: Users can enter at any step. Useful for general website behaviour.

Event Tracking Prerequisites

Your funnel is only as good as the events you track. Essential events:

| Event | What It Captures | |-------|------------------| | page_view | Pages visited | | scroll | Scroll depth (25%, 50%, 75%, 90%) | | click | Button clicks, CTA clicks | | form_start | User begins filling a form | | form_submit | Form successfully submitted | | add_to_cart | Product added to cart | | begin_checkout | Checkout process started | | purchase | Transaction completed | | generate_lead | Lead form completed |

If these events aren't set up, you're flying blind. Implement them via Google Tag Manager before attempting funnel analysis.

Heatmaps and Session Recordings

GA4 tells you what's happening. Heatmaps and session recordings show you why.

• Heatmaps (Hotjar, Microsoft Clarity, Lucky Orange) show where people click, how far they scroll, and what they ignore
• Session recordings let you watch real visitor sessions to see confusion, hesitation, and rage clicks
• Form analytics (specific tools or Hotjar) show which form fields cause abandonment

Use quantitative data (GA4) to identify where people drop off. Use qualitative data (recordings, heatmaps) to understand why.

---

Diagnosing Drop-Offs

Stage 1 → Stage 2: Landing to Engagement

Symptom: High bounce rate. People land and immediately leave.

Common causes:

• Mismatch between ad/search and landing page. They expected one thing and got another. If your Google Ad says "Free SEO Audit" and the landing page doesn't mention it above the fold, they leave.
• Slow page load. Every additional second of load time increases bounce rate by 7-10%. Check Core Web Vitals.
• Unclear value proposition. Visitor can't figure out what you do or why they should care within 5 seconds.
• Poor mobile experience. The site works on desktop but is unusable on a phone.
• Visual credibility. Outdated design, stock photos, or amateur layout signals untrustworthiness.

Fixes:

• Match landing page headline to the ad or search intent
• Achieve sub-3-second page load
• Put a clear, specific value proposition above the fold
• Test on real mobile devices (not just browser dev tools)

Stage 2 → Stage 3: Engagement to Interest

Symptom: People browse but never reach key pages (pricing, services, products).

Common causes:

• Poor navigation. People can't find what they're looking for.
• No clear next step. The page doesn't guide them toward conversion-focused pages.
• Content doesn't build confidence. Blog posts are interesting but don't connect to services/products.
• Too many choices. Paradox of choice — too many options leads to no decision.

Fixes:

• Add clear CTAs on every page pointing toward the next step
• Simplify navigation — key pages within 1-2 clicks
• Add internal links from content pages to service/product pages
• Use "most popular" or "recommended" labels to reduce choice overload

Stage 3 → Stage 4: Interest to Action Initiation

Symptom: People visit pricing/service pages but don't start converting.

Common causes:

• Pricing isn't visible or is confusing. Uncertainty about cost prevents action.
• No social proof on the page. They're interested but not convinced.
• CTA is weak or buried. "Contact us" at the bottom of the page after 2,000 words of text.
• Too much commitment required. "Book a 60-minute strategy session" feels like a big ask. "Get a free 5-minute quote" doesn't.
• Missing information. They have unanswered questions that prevent them from taking the next step.

Fixes:

• Show pricing or pricing ranges (hiding pricing increases anxiety)
• Add testimonials and case studies near the CTA
• Make the CTA visible above the fold and repeat it throughout the page
• Reduce perceived commitment ("Free," "No obligation," "Takes 2 minutes")
• Add FAQ sections addressing common objections

Stage 4 → Stage 5: Action Initiation to Conversion

Symptom: People start the form/checkout but don't finish.

Common causes:

• Too many form fields. Every field beyond the essential ones reduces completion by 10-15%.
• Unexpected costs (shipping, taxes, fees appearing at checkout).
• No trust signals. No security badges, reviews, or guarantees near the payment step.
• Technical issues. Form errors, broken validation, page crashes.
• Forced account creation. Requiring signup before purchase or form submission.

Fixes:

• Reduce form fields to the absolute minimum
• Show all costs upfront (no surprises at checkout)
• Add trust badges, secure payment icons, and guarantees near conversion points
• Test the form/checkout yourself on multiple devices
• Offer guest checkout or alternative submission methods

---

The Prioritisation Framework

You've identified multiple drop-offs. Which do you fix first?

Fix the biggest leak closest to the conversion point first.

Why? Improvements at the bottom of the funnel have immediate revenue impact. Fixing a 50% drop-off at the form submission stage doubles your conversions from that step — today.

Fixing a 30% drop-off at the landing page increases traffic into the funnel but doesn't help if downstream stages are still leaking.

Priority order:

1. Form/checkout completion (Stage 4 → 5)
2. CTA/action initiation (Stage 3 → 4)
3. Interest development (Stage 2 → 3)
4. Landing engagement (Stage 1 → 2)

Impact Estimation

Before fixing anything, estimate the impact:

Current: 10,000 visitors → 500 form starts → 80 completions (16% form completion rate)

If we improve form completion to 25%: 10,000 visitors → 500 form starts → 125 completions

That's 56% more conversions without spending a cent on additional traffic.

---

Funnel Analysis by Segment

Average funnel performance hides important differences. Always segment.

By Device

Mobile funnels almost always underperform desktop. If mobile conversion is 50%+ lower than desktop, you have a mobile UX problem — not a traffic problem.

By Traffic Source

| Source | Typical Funnel Behaviour | |--------|------------------------| | Organic search | Higher intent, better progression through funnel | | Paid search | High intent but impatient — bounce if landing page doesn't match | | Social media | Lower intent, more exploration, fewer direct conversions | | Direct | Returning visitors with higher conversion rates | | Referral | Depends on source — warm referrals convert well |

If paid traffic bounces at 80% while organic bounces at 40%, the problem isn't your landing page — it's your ad targeting or ad-to-page match.

By Geography

International visitors may drop off at shipping costs. Local visitors may convert better. Different regions may need different messaging.

By New vs. Returning

Returning visitors typically convert 2-3x higher than first-time visitors. If returning visitor conversion is low, you have a trust or UX problem. If first-time visitor conversion is low, you may have an awareness or messaging problem.

---

Building a Funnel Dashboard

Track your funnel weekly with a simple dashboard.

Essential Metrics

| Stage | Metric | Target | |-------|--------|--------| | Landing | Bounce rate | Under 50% | | Engagement | Pages per session | 2+ | | Interest | Key page views (pricing, services) | 15-25% of visitors | | Action initiation | Form starts / Add to cart | 5-10% of visitors | | Conversion | Form completion / Purchase | 50-70% of initiators | | Overall | Visitor to customer | Set based on your baseline |

Monitoring Cadence

Weekly: Check overall funnel conversion and identify any sudden changes. Monthly: Deep-dive into segment-level performance. Compare to previous month. Quarterly: Full funnel audit with heatmap and session recording analysis.

When to Investigate

• Drop-off rate at any stage increases by 10%+ compared to the previous period
• One segment dramatically underperforms others (e.g., mobile 3x worse than desktop)
• A change you made (new page design, new form, new CTA) correlates with funnel performance change

---

Common Mistakes

1. Analysing averages instead of segments — your overall funnel hides dramatically different behaviour between mobile/desktop, traffic sources, and visitor types
2. Fixing the top of the funnel first — more traffic into a broken funnel just means more people having a bad experience
3. Not tracking micro-conversions — if you only track the final conversion, you can't see where the problem is. Track every meaningful step.
4. Making multiple changes at once — if you change the headline, form, CTA, and page layout simultaneously, you won't know which change had the impact
5. Ignoring qualitative data — numbers tell you where people leave. Session recordings and surveys tell you why.
6. One-time analysis — funnel performance changes constantly. Seasonal patterns, marketing changes, and product updates all affect the funnel. Monitor continuously.
7. No baseline — if you don't know your current conversion rates, you can't measure improvement. Establish baselines before making changes.
8. Optimising for the wrong metric — a lower bounce rate is meaningless if conversion rate doesn't improve. Always tie funnel metrics back to revenue.

---

Start Here

1. Map your funnel — write down every step from first visit to final conversion
2. Set up event tracking in GA4 for each step (use Google Tag Manager)
3. Build a GA4 funnel exploration with your defined steps
4. Identify the stage with the largest drop-off
5. Watch 10 session recordings of users who abandoned at that stage
6. Hypothesise why they're leaving (based on data and recordings)
7. Make one change to address the biggest drop-off
8. Monitor for 2-4 weeks and measure the impact

Funnel analysis transforms marketing from guesswork into a systematic process. Instead of asking "why aren't we getting more customers?" you ask "why do 70% of people who start our form not finish it?" — and that question has an answer you can find and fix. Every percentage point improvement in your funnel compounds across every visitor, every day, for as long as the improvement lasts.