The Conversion Diagnostic Framework: How to Find What's Killing Your Conversion Rate

Most e-commerce stores guess at their conversion problems.

They see a 1.4% conversion rate, assume the product pages need work, redesign the product pages, and watch the number stay at 1.4%.

The problem wasn’t the product pages. They just assumed it was.

The Conversion Diagnostic Framework is a 5-step method for finding exactly where in your purchase funnel you’re losing customers — and why. Not a checklist of “best practices.” A diagnostic process that starts with your actual data.

Want us to run this framework on your store? Book a free 30-minute diagnostic preview. We’ll identify your top 3 conversion blockers before you commit to anything. Book a free diagnostic →

Why Most CRO Fails

Standard CRO advice gives you a list of things to fix. Update your CTA copy. Add reviews. Reduce checkout steps.

These are fine tactics. But they’re not diagnostic. They don’t tell you whether these are the specific things costing your specific store.

Two stores in the same category, with similar traffic quality, can have wildly different conversion problems. One store loses 40% of potential customers at the shipping cost reveal. Another loses them at the payment step because iDEAL is buried. A third has a mobile checkout with broken form inputs.

Same symptom (low conversion rate), completely different causes, completely different fixes.

The Conversion Diagnostic Framework tells you which store you are.

The 5 Steps

Step 1: Establish Your Funnel Baseline

Before you can diagnose a problem, you need to see the whole funnel.

Map conversion from entry to purchase across each major step:

Landing page → Product page: What % of landing page visitors reach a product page?
Product page → Add to cart: What % of product page visitors add to cart?
Add to cart → Checkout initiated: What % of cart visitors start checkout?
Checkout initiated → Purchase: What % who start checkout complete it?

Pull these numbers from Google Analytics (Conversions → Ecommerce → Funnel Exploration in GA4). You need at least 30 days of data, ideally 90 days, for statistically meaningful numbers.

What you’re looking for: Which step has the largest percentage drop-off relative to what’s normal for your category?

Baymard’s 2024 EU benchmark: cart-to-checkout completion averages 55–65%. Checkout-to-purchase averages 45–55%. If either of your numbers falls more than 15 percentage points below these ranges, that step is your primary diagnostic target.

Step 2: Segment the Drop-Off

Once you’ve identified the highest-drop-off step, segment it.

Split by:

Device: Desktop vs mobile drop-off at that step
Traffic source: Organic vs paid vs email vs direct
New vs returning visitor: New visitors abandon at different rates and for different reasons
Product category or price tier: High-ticket items behave differently from impulse purchases

This segmentation reveals whether the problem is universal (broken for everyone) or specific (broken for mobile users only, or for first-time visitors only).

Why this matters: A problem affecting only mobile users (which is 60–70% of your traffic) needs a different fix than a problem affecting all users. A problem specific to new visitors might be a trust issue. A problem specific to one product category might be a pricing or information gap in that category.

In GA4: Use the Funnel Exploration report with breakdowns. Add User Type, Device Category, and Session Source as breakdowns on the drop-off step.

Step 3: Watch What Actually Happens

Data tells you where. Session recordings tell you why.

For the step with the highest drop-off, watch 25–50 session recordings of users who abandoned at that step. Specifically:

What was the last thing they did before leaving?
Where did their cursor or scroll position go before they closed the page?
Did they interact with anything and then stop?
Are there visible moments of confusion, re-reading, or hesitation?

Tools: Hotjar, Microsoft Clarity (free), FullStory.

What you’re looking for:

Rage clicks (clicking something that doesn’t respond as expected)
U-turns (navigating forward, then back, indicating uncertainty)
Scroll to specific element then abandon (often indicates a deal-breaker discovered)
Form field interaction followed by exit (form friction)
Fast exit without scrolling (first impression or load issue)

Document patterns. If 18 out of 25 recordings show users scrolling to the shipping cost line and then leaving, you’ve found your problem.

Step 4: Run the Friction Audit on Your Target Step

For the specific step you’ve identified as the drop-off point, run a structured friction audit.

Each page in the purchase funnel has a known set of friction categories. Here are the most common, by step:

Product page friction:

No or few reviews visible without clicking through
Price unclear for variant products (doesn’t update dynamically)
Size guide missing or hard to find
Return policy not accessible from the page
Add-to-cart below the fold on mobile

Cart friction:

Shipping cost revealed for the first time
No progress bar or visual confirmation of cart state
No trust signals (payment logos, return policy)
Cross-sell placement that interrupts rather than assists

Checkout friction:

Account creation required before purchase
Payment method not showing iDEAL prominently (Dutch stores)
Form labels disappear on focus (placeholder-only labels)
Inline validation only fires on submit, not on field exit
Mobile keyboard doesn’t match field type (phone/postcode)
Order total changes between cart and checkout confirmation

Mobile-specific friction:

Page load over 3 seconds
Touch targets under 44px
Non-dismissible cookie banner taking 40%+ of viewport
Checkout doesn’t persist cart data if session times out

Score each friction point on your target step. Note which are present. Prioritize by frequency (how many users encounter it) and severity (does it typically cause abandonment or just delay?).

Step 5: Prioritize and Test

Every friction point you identify needs a severity rating and an effort estimate.

Severity: How often does this cause abandonment vs just slow users down?

Critical: Directly causes a significant percentage to abandon
Major: Causes abandonment for a meaningful segment (e.g., mobile-only)
Minor: Adds friction but rarely prevents completion

Effort: How hard is this to fix?

Quick (under 1 day of developer time)
Medium (1–5 days)
Heavy (5+ days or requires platform-level changes)

Fix Critical + Quick items immediately. These are your highest ROI changes.

For Critical + Heavy items, build the business case: what is the estimated revenue recovery from fixing this? If moving from 1.4% to 1.7% checkout conversion on your traffic volume generates €8,000/month, a 5-day development sprint at €3,000 has a 2-week payback.

For Major items, A/B test when possible. Confirm the friction exists in your specific context before committing to a fix. What Baymard identifies as a universal pattern may affect your specific audience differently.

A Real Diagnostic Example

Here’s how the framework plays out in practice.

A Dutch kitchenware store, €2.8M annual revenue, came to BTNG with a flat 1.3% conversion rate.

Step 1 — Funnel baseline:

Product → Cart: 8.2% (below-average but not outlier)
Cart → Checkout start: 52% (low end of normal)
Checkout start → Purchase: 28% (significantly below the 45–55% benchmark)

Primary target: checkout completion.

Step 2 — Segment:

Desktop checkout completion: 41% (near normal)
Mobile checkout completion: 18% (severely low)

The problem is mobile-specific.

Step 3 — Session recordings: Watched 40 mobile checkout recordings. 31 of 40 showed the same pattern: user reached the payment step, scrolled, appeared to look for iDEAL, could not find it immediately, and exited.

Step 4 — Friction audit (checkout, mobile):

iDEAL was present but appeared as the 4th item in the payment list, below the fold on an iPhone 13 screen
“More payment options” disclosure was collapsible — iDEAL was inside it
Apple Pay and Google Pay were not enabled

Step 5 — Prioritization:

Move iDEAL to first position in payment list: Critical severity, Quick effort
Enable Apple Pay/Google Pay: Critical, Medium effort

Both fixed within one week.

Result: Mobile checkout completion rate moved from 18% to 34%. Overall conversion rate: 1.3% to 1.9%. Revenue impact: approximately €47,000 additional annual revenue. One week of work.

This is what the Conversion Diagnostic Framework surfaces. Not “improve your mobile UX.” Specifically: iDEAL is the fourth item in your payment list and it’s costing you €47K/year.

Running the Framework Yourself

Time required: 4–6 hours for the full 5 steps, assuming you have analytics and session recordings set up.

Tools needed:

Google Analytics 4 (Funnel Exploration report)
Session recording tool (Hotjar free plan covers 35 daily sessions; Microsoft Clarity is free and unlimited)
A spreadsheet for the friction audit log

Prerequisites:

Enhanced ecommerce tracking is working in GA4 (add-to-cart events, checkout steps, purchase events)
At minimum 2 weeks of session recordings running — you can’t watch recordings that weren’t captured
At least 500 sessions per funnel step for meaningful data (below this, individual sessions skew the numbers)

If your analytics aren’t set up correctly, that’s step zero. The framework doesn’t work on bad data.

When to Get Help

The framework is designed to be self-serviceable. But there are situations where outside perspective adds value:

You’ve run the framework and found issues but aren’t sure which to prioritize first. The impact estimates require judgment about what’s causal vs correlational. Getting it wrong means fixing the wrong thing.

Your session recordings show clear friction but you’re not sure how to fix it. Identifying the problem and knowing the right UX fix are different skills.

You’ve fixed the obvious things and conversion still hasn’t moved. This usually means there’s a second-order problem the initial diagnosis missed. Often a trust issue or a product/price fit issue, not a UX issue.

BTNG runs this framework professionally. We’ll identify your specific blockers and tell you exactly what to fix, in what order, with implementation-ready specs.

Book a free diagnostic preview →