E-commerce Funnel Analysis: How to Find Where Your Store Is Losing Revenue

Driving more traffic does not automatically create more sales.

An e-commerce store can receive thousands of sessions, generate strong product interest, and still lose a significant share of potential purchases before customers complete checkout.

The issue may start when visitors fail to reach product pages. It may appear between product views and add to cart. It may happen inside the cart, during checkout, or at the payment stage.

Looking only at the overall conversion rate will not tell you where the problem begins.

To understand what is actually happening, you need to analyze the full e-commerce funnel stage by stage:

Sessions → Product Views → Add to Cart → Begin Checkout → Purchase

This guide explains how to calculate the performance of each stage, interpret funnel drop-offs correctly, estimate commercial opportunities, and decide what to investigate next.

You can also use the free Mersad Funnel Leak Detector to calculate the main metrics automatically:

Analyze your e-commerce funnel with the free tool

---

What Is an E-commerce Funnel?

An e-commerce funnel represents the sequence of actions a visitor takes before completing a purchase.

A simplified funnel usually includes five main stages:

1. Sessions: People who visit the store.
2. Product Views: Visitors who reach and view product pages.
3. Add to Cart: Visitors who add at least one product to their cart.
4. Begin Checkout: Visitors who start the checkout process.
5. Purchases: Visitors who complete an order.

Every stage answers a different business question.

A weak session-to-product-view rate may indicate that visitors are not finding relevant products.

A weak product-view-to-add-to-cart rate may point to issues around the offer, product information, price, stock, variants, delivery, returns, or trust.

A weak add-to-cart-to-checkout rate may be connected to cart friction, unclear shipping costs, weak checkout calls to action, discount-field distraction, or technical problems.

A weak checkout completion rate may require investigation into payments, shipping fees, address forms, browser errors, delivery restrictions, or purchase tracking.

The funnel does not prove why customers leave. It tells you where measurable deterioration occurs.

That distinction is essential.

---

Why the Overall Conversion Rate Is Not Enough

The overall conversion rate is calculated as:

Purchases ÷ Sessions × 100

For example, if a store generates 420 purchases from 40,000 sessions:

420 ÷ 40,000 × 100 = 1.05%

The number is useful, but incomplete.

A conversion rate of 1.05% does not tell you whether the main issue is:

• Traffic quality
• Landing-page relevance
• Product discovery
• Product-page persuasion
• Add-to-cart friction
• Cart usability
• Checkout completion
• Payment failure
• Tracking quality
• Product or audience mix

Two stores can have the same overall conversion rate while having completely different funnel problems.

One store may lose most visitors before they reach a product page.

Another may generate strong add-to-cart activity but fail during checkout.

The overall result is similar, but the required action is completely different.

That is why e-commerce funnel analysis should begin with stage-level conversion rates.

---

The Main E-commerce Funnel Metrics

1. Product View Rate

The Product View Rate shows how many sessions progress to a product page.

Product Views ÷ Sessions × 100

Example:

18,000 Product Views ÷ 40,000 Sessions = 45%

This means 45% of sessions reached a product page.

A weak result does not automatically mean the homepage is badly designed. It may also reflect:

• Irrelevant traffic
• Campaign-to-landing-page mismatch
• Visitors landing on low-intent content
• Weak navigation
• Poor search functionality
• Category-page friction
• Tracking errors
• An unusual landing-page mix

The first validation step should usually be segmentation by source, medium, campaign, landing page, device, geography, and new versus returning visitors.

---

2. Add-to-Cart Rate From Product Views

This metric shows how many product viewers add an item to the cart.

Add to Carts ÷ Product Views × 100

Example:

1,800 Add to Carts ÷ 18,000 Product Views = 10%

This stage is often influenced by:

• Product relevance
• Price
• Offer clarity
• Product images and video
• Product descriptions
• Reviews and social proof
• Stock availability
• Variant and size selection
• Delivery information
• Returns policy
• Add-to-cart visibility
• Mobile usability
• Add-to-cart tracking

A weak result should be segmented by product, category, traffic source, device, customer type, and landing page.

A store may have a healthy overall product-to-cart rate while a small group of high-traffic products performs poorly and creates a major revenue opportunity.

---

3. Checkout Initiation Rate

This metric shows how many users who added products to the cart started checkout.

Begin Checkouts ÷ Add to Carts × 100

Example:

900 Begin Checkouts ÷ 1,800 Add to Carts = 50%

Potential friction may include:

• Unclear shipping costs
• Free-shipping threshold confusion
• Weak checkout CTA visibility
• Cart drawer issues
• Coupon-field distraction
• Unexpected fees
• Login requirements
• Out-of-stock items
• Upsell overload
• Device-specific technical errors
• Begin-checkout tracking problems

This stage should be reviewed carefully on the most important devices and browsers.

A store may appear healthy on desktop while losing a much larger share of mobile users in the cart.

---

4. Checkout Completion Rate

Checkout Completion Rate shows how many users who started checkout completed a purchase.

Purchases ÷ Begin Checkouts × 100

Example:

420 Purchases ÷ 900 Begin Checkouts = 46.67%

A weak checkout completion rate may be related to:

• Payment failures
• Missing payment methods
• Shipping-cost shock
• Address-form friction
• Mobile keyboard or field issues
• Delivery-area restrictions
• Cash-on-delivery rules
• Promo-code errors
• Browser problems
• Purchase-event tracking issues

Checkout analysis should combine analytics with payment logs, platform orders, customer support feedback, session recordings, and technical QA.

Do not assume that every checkout drop-off is a UX issue.

Some losses may come from payment rejection, stock changes, delivery restrictions, fraud controls, or incorrect analytics implementation.

---

5. Average Order Value

Average Order Value shows the average revenue generated by each completed order.

Revenue ÷ Purchases

Example:

SAR 105,000 ÷ 420 Purchases = SAR 250

AOV helps translate additional purchases into estimated revenue impact.

It also helps identify situations where purchases increase while revenue declines.

That can happen when:

• Customers shift toward lower-priced products
• Discount usage increases
• Bundles perform differently
• Product mix changes
• Upsells weaken
• Higher-value products go out of stock

A conversion uplift is not automatically a business win if AOV, margin, cancellation rate, or return rate deteriorates.

---

6. Revenue per Session

Revenue per Session measures the amount of revenue generated from each store session.

Revenue ÷ Sessions

Example:

SAR 105,000 ÷ 40,000 Sessions = SAR 2.63 per session

Revenue per Session is useful because it connects traffic volume, conversion efficiency, and order value.

A store may increase its conversion rate while Revenue per Session remains flat if AOV falls.

Another store may maintain the same conversion rate while Revenue per Session improves because customers purchase higher-value products.

This is why funnel analysis should not focus on conversion rate alone.

---

How to Calculate Funnel Drop-Off

Drop-off represents the share of users who did not progress to the next stage.

Drop-off Rate = 1 − Stage Conversion Rate

Using the previous example:

Sessions: 40,000
Product Views: 18,000

Product View Rate:

18,000 ÷ 40,000 = 45%

Drop-off:

100% − 45% = 55%

The lost volume is:

40,000 − 18,000 = 22,000 sessions

This gives you two different views:

• Percentage drop-off: How inefficient the transition is.
• Volume loss: How many users did not progress.

The highest percentage drop-off is not always the most important commercial problem.

A late-stage issue may affect fewer users but have higher purchase intent.

An early-stage issue may affect a large number of visitors but include low-quality traffic that was unlikely to convert.

Priority should consider:

• Lost volume
• Stage conversion rate
• Revenue proximity
• Change versus previous period
• Traffic and audience quality
• Data reliability
• Business impact
• Ease of validation
• Implementation complexity

---

Use the Free Funnel Leak Detector

You can calculate the metrics manually, or use the free Mersad Funnel Leak Detector.

The tool allows you to enter:

• Sessions
• Product Views
• Add to Carts
• Begin Checkouts
• Purchases
• Revenue
• Currency
• Optional previous-period data

It then calculates:

• Stage conversion rates
• Stage drop-offs
• Lost volume
• Overall conversion rate
• Average Order Value
• Revenue per Session
• Estimated opportunity scenarios
• Priority scores
• Possible hypotheses
• Required validation
• Recommended next actions

You can also export the analysis as PDF or Excel.

Open the Mersad Funnel Leak Detector

The tool is available in Arabic and English and supports multiple currencies.

---

How to Compare Two Periods Correctly

Comparing funnel performance over time can help identify where a decline or uplift started.

However, the comparison must use consistent data.

Use:

• The same event definitions
• The same attribution logic
• Comparable date ranges
• Consistent tracking
• The same currency
• Similar promotional context when possible
• Comparable stock and pricing conditions

When comparing periods, answer these questions:

What changed?

Calculate both the absolute and percentage change in:

• Sessions
• Product Views
• Add to Carts
• Begin Checkouts
• Purchases
• Revenue
• Conversion Rate
• Average Order Value
• Revenue per Session

Where did the change start?

Identify the first funnel stage where performance materially changed.

If sessions increased but product views did not, review traffic quality and landing pages.

If product views increased but add to carts declined, investigate product mix, price, content, stock, offers, and device performance.

If checkout starts increased but purchases declined, focus on checkout, payments, delivery, and tracking.

Which segment contributed most?

Break results down by:

• Source and medium
• Campaign
• Landing page
• Device
• Browser
• Geography
• New versus returning users
• Product
• Category
• Payment method
• Shipping method
• Customer type

An overall decline may be caused by traffic shifting toward a weaker segment, not by the whole website getting worse.

---

Traffic Increased, but Purchases Declined

This is one of the most important situations to investigate.

More traffic with fewer purchases may indicate:

• Lower-intent acquisition
• A change in campaign targeting
• Traffic going to weaker landing pages
• More mobile traffic with lower conversion
• Geographic mix changes
• Product availability issues
• Pricing or promotion changes
• Technical problems
• Tracking changes
• A real decline in funnel efficiency

Do not immediately blame the website or the advertising campaign.

Separate the problem into two questions:

1. Did the quality or allocation of traffic change?
2. Did the website convert the traffic less efficiently?

The answer may involve both.

---

Purchases Increased, but Revenue Declined

More orders do not always mean stronger business performance.

If purchases increase while revenue declines, review:

• Average Order Value
• Discounting
• Product mix
• Bundle performance
• Upsells
• High-value product availability
• Returns
• Cancellations
• Net revenue versus gross revenue

A higher conversion rate can still create a weaker commercial outcome if it is driven by low-value or low-margin orders.

---

How to Estimate a Revenue Opportunity

Opportunity scenarios can help estimate the commercial effect of improving a specific stage.

For example, assume:

Product Views: 18,000
Add to Carts: 1,800
Current Product View → Add to Cart Rate: 10%

If the stage improves by 10% relatively, the target rate becomes:

10% × 1.10 = 11%

Expected add to carts:

18,000 × 11% = 1,980

Additional add to carts:

1,980 − 1,800 = 180

To estimate additional purchases, apply the current downstream conversion rates.

If:

Add to Cart → Checkout = 50%
Checkout → Purchase = 46.67%

Then:

180 × 50% × 46.67% ≈ 42 additional purchases

With an AOV of SAR 250:

42 × SAR 250 = SAR 10,500 estimated additional revenue

This is a scenario, not a forecast.

It assumes:

• Downstream rates remain stable
• AOV remains stable
• Traffic quality remains stable
• Stock is available
• No operational constraint changes
• The improvement is real and sustainable

Use opportunity estimates to support prioritization, not to promise results.

---

Why the Largest Drop-Off Is Not Automatically the Top Priority

Every funnel contains drop-off.

A visitor who enters the store is not guaranteed to view a product. A product viewer is not guaranteed to add to cart. A cart user is not guaranteed to purchase.

The objective is not to eliminate all drop-off.

The objective is to find the stage where measurable deterioration, business impact, confidence, and validation potential justify action.

A useful prioritization process should consider:

Impact

How much revenue or customer volume could the issue affect?

Confidence

How reliable is the evidence?

Do you have consistent analytics, segment-level patterns, recordings, customer feedback, or technical confirmation?

Effort

How difficult is it to investigate and implement a change?

Urgency

Is this a severe technical issue, checkout blocker, payment failure, or tracking problem?

Business significance

Would an improvement affect revenue, margin, repeat purchase, operational cost, or customer experience?

Do not A/B test an obvious bug, broken payment method, or tracking failure.

Fix it directly and verify the result.

---

From Funnel Finding to Root-Cause Validation

Funnel data identifies the location of a problem, not necessarily the cause.

For example:

Observation: Product View to Add to Cart Rate declined.

That does not prove that the Add to Cart button is weak.

Possible hypotheses may include:

• Traffic shifted toward low-intent products
• Prices increased
• Discounts ended
• Delivery became less attractive
• Popular variants went out of stock
• Product reviews weakened
• Product content became less persuasive
• Mobile selection became harder
• Tracking changed

To move from observation to diagnosis, combine several sources of evidence.

Quantitative Analysis

Use GA4 or another analytics platform to segment by:

• Product
• Category
• Source
• Landing page
• Device
• Customer type
• Geography
• Browser
• Date
• Campaign

Behavioral Analysis

Use tools such as heatmaps and session recordings to review:

• Hesitation
• Repeated clicks
• Dead clicks
• Scrolling
• Variant interaction
• Form abandonment
• Navigation behavior
• Cart and checkout friction

Voice of Customer

Review:

• Customer surveys
• Exit-intent polls
• Support conversations
• Reviews
• Return reasons
• Search queries
• On-site feedback

Technical Validation

Check:

• Event tracking
• Browser errors
• Page speed
• Payment failures
• Inventory synchronization
• Coupon behavior
• Shipping calculations
• Mobile responsiveness

A strong CRO diagnosis connects the funnel data to user behavior, technical evidence, and business operations.

---

Common E-commerce Funnel Analysis Mistakes

1. Treating Event Counts as Users

A customer can trigger some events more than once.

If your funnel uses event counts, Add to Cart may exceed Product Views in certain implementations.

Define whether your data represents users, sessions, events, or transactions before interpreting the funnel.

---

2. Mixing Different Date Ranges

All funnel metrics must come from the same reporting period.

Do not combine monthly sessions with weekly purchases.

---

3. Ignoring Tracking Quality

Incorrect event implementation can create false drop-offs or impossible funnels.

Validate the events before making UX or business decisions.

---

4. Comparing Different Traffic Mixes

A performance decline may be caused by more low-intent traffic entering the funnel.

Always review source, campaign, landing page, device, geography, and audience composition.

---

5. Using Universal Benchmarks as Targets

Benchmarks can provide context, but they do not define what your store should achieve.

Conversion performance depends on:

• Industry
• Price
• Product type
• Repeat purchase
• Device
• Geography
• Traffic source
• Brand strength
• Payment methods
• Delivery experience
• Customer trust

Your own historical and segment-level performance is often more useful than a broad industry average.

---

6. Assuming Correlation Proves Causation

A lower conversion rate after a design change does not automatically mean the design caused the decline.

Other variables may have changed at the same time.

Validate before concluding.

---

7. Optimizing the Highest Drop-Off Without Context

The largest percentage leak may not represent the largest commercial opportunity.

Consider volume, purchase intent, data quality, and downstream impact.

---

A Practical Funnel Analysis Workflow

Use this process when reviewing your store.

Step 1: Validate the data

Confirm that all funnel events are being tracked correctly.

Step 2: Calculate the main rates

Measure every stage transition, overall conversion rate, AOV, and Revenue per Session.

Step 3: Compare periods

Identify where the change began.

Step 4: Segment the funnel

Break performance down by traffic source, page, device, audience, product, and geography.

Step 5: Identify confirmed findings

Describe only what the data proves.

Example:

Mobile Product View → Add to Cart Rate declined from 9.2% to 6.8%.

Step 6: Create hypotheses

List potential explanations without presenting them as facts.

Step 7: Define required validation

Specify what data, research, recordings, or technical checks are needed.

Step 8: Prioritize actions

Use business impact, confidence, effort, and urgency.

Step 9: Fix obvious issues

Do not test severe bugs or broken tracking.

Step 10: Experiment where uncertainty remains

Use controlled testing when multiple viable solutions exist and sufficient traffic is available.

---

Analyze Your Store’s Funnel for Free

The Mersad Funnel Leak Detector provides a structured starting point for this workflow.

Use it to:

• Calculate stage-level conversion rates
• Identify the largest drop-offs
• Review lost customer volume
• Estimate opportunity scenarios
• Compare periods
• Download your analysis
• Build a validation plan

Use the free Funnel Leak Detector

The output should be treated as a diagnostic starting point.

A complete CRO analysis still requires segmentation, behavioral research, technical validation, and commercial context.

---

Frequently Asked Questions

What data do I need to analyze my e-commerce funnel?

You need sessions, product views, add-to-cart actions, checkout starts, purchases, and revenue from the same reporting period.

You should also confirm whether the numbers represent users, sessions, or event counts.

---

Is the Funnel Leak Detector free?

Yes. The Mersad Funnel Leak Detector is free to use.

You can analyze your funnel and download the results.

Open the free tool

---

Can I use the tool for Shopify, Salla, Zid, or WooCommerce?

Yes.

The tool is platform-independent. You can use it with data from Shopify, Salla, Zid, WooCommerce, or a custom e-commerce website, provided the funnel definitions are consistent.

---

Does the tool identify the cause of a drop-off?

No.

It identifies where measurable drop-off occurs and provides potential hypotheses and validation actions.

Root causes require additional evidence.

---

Can I compare two reporting periods?

Yes.

You can enter previous-period data to compare funnel efficiency over time.

Make sure both periods use consistent tracking and comparable definitions.

---

What is a good e-commerce conversion rate?

There is no single conversion rate that applies to every store.

Performance varies by industry, traffic quality, device, geography, price, product type, brand strength, and customer mix.

Start by comparing your store against its own historical and segment-level performance.

---

Should I optimize the stage with the highest drop-off?

Not automatically.

Consider lost volume, purchase intent, deterioration versus the previous period, revenue proximity, confidence, and effort.

---

Is an estimated revenue opportunity guaranteed?

No.

It is a scenario based on current downstream conversion rates and Average Order Value remaining stable.

It is not a guaranteed forecast.

---

Final Takeaway

E-commerce funnel analysis is not about finding one percentage and calling it the problem.

It is a structured process for answering:

• Where did performance weaken?
• Which users and segments contributed most?
• Is the issue related to traffic, conversion, product mix, tracking, or operations?
• What does the data confirm?
• What remains a hypothesis?
• What should be validated next?
• Which action has the strongest business case?

The funnel tells you where to investigate.

Behavioral data, customer research, technical validation, and experimentation help you prove why.

Start with your own data:

Analyze your e-commerce funnel with the free Mersad Funnel Leak Detector