The Conversion Diagnostic Framework

Your conversion rate is 0.8%.

You know something’s wrong. But what?

The checkout page? Product photos? Shipping costs? Mobile experience? Trust signals?

So you start guessing.

Add some trust badges. Change button colors. Install a popup. Offer a discount. Copy what your competitor is doing.

Nothing improves.

Here’s what I’ve learned from 100+ audits: most conversion problems aren’t what you think they are.

The client thinks: “We need better product images.” The actual problem: No one trusts this is a real company.

The client thinks: “Cart abandonment is too high, we need exit popups.” The actual problem: Surprise shipping costs at checkout.

The client thinks: “Button should be green instead of blue.” The actual problem: Safari mobile checkout is completely broken.

You can’t fix the right thing if you don’t know what’s actually broken.

This is how professionals diagnose conversion problems. Not by guessing. Not by copying tactics. By following a systematic process that reveals what’s actually stopping people from buying.

Why Random Fixes Don’t Work

Let me tell you about a jewelry client.

They came to me convinced they needed better photos. Spent weeks talking to photographers. Got quotes. Planned a whole product shoot.

I watched 10 session recordings.

Zero people zoomed into product photos. Nobody seemed confused about what the jewelry looked like.

What I DID see:

People hesitating at checkout
Refreshing the payment page multiple times
Abandoning right before clicking “complete order”

Exit survey asked: “What almost stopped you from buying?”

Top answer: “I wasn’t sure if this was a real company or a scam.”

The problem wasn’t photography. It was credibility.

We added:

“Family business since 1987” badge
Team photos with names
Physical showroom address with Google Maps embed
Customer service phone number prominently displayed

No new photos. Same products. Same prices.

Conversion increased 34% in eight weeks.

If they’d followed their gut and hired the photographer, they would’ve wasted €5,000 and still had a trust problem.

This is why diagnosis matters.

The Six-Step Diagnostic Process (ResearchXL Framework)

Peep Laja from CXL built a framework that works. I’ve used it on dozens of stores.

It’s called ResearchXL.

Six steps. Each one answers a different question about why your store isn’t converting.

Step 1: Technical Analysis

Question: Does your site actually work?

Start here. Always.

You’d be shocked how often the problem is technical.

Checkout broken on iPhone Safari (30% of your traffic). Forms don’t submit on certain browsers. Page loads in 4.2 seconds instead of 1.8 seconds. Mobile buttons too small to tap. Images don’t load on 4G connections.

I audited a furniture store last year. They complained about terrible conversion.

Opened their site on my phone. Tried to check out.

Payment button didn’t work. At all. Literally couldn’t complete a purchase on mobile. Mobile checkout issues are the most common and most damaging bugs I find in audits.

They’d been losing 68% of potential revenue for three months because nobody tested mobile checkout after the last Shopify update.

How to do technical analysis:

Pull up Shopify Analytics. Go to Reports → Behavior → Sessions by device.

See conversion by device type. If mobile is 0.3% and desktop is 2.1%, you have a mobile problem.

Then actually use your site on:

iPhone Safari
Android Chrome
Desktop Chrome
Desktop Safari

Go through the full checkout flow. Every step. Use a real credit card (you can cancel the order).

Does everything work? No broken buttons? Forms submit correctly? Images load? No weird scroll behavior?

Check page speed. Google “PageSpeed Insights.” Enter your URLs.

Anything over 3 seconds on mobile is slow. Over 7 seconds and people are leaving before the page even loads.

This step finds low-hanging fruit. Fix these issues first. Everything else is pointless if your site doesn’t function.

Step 2: Heuristic Analysis

Question: Does your page make sense from a visitor’s perspective?

This is where you evaluate your store using established conversion principles.

I use the LIFT Model from Chris Goward. Six factors that influence conversion:

Value Proposition Can someone tell within 5 seconds what you sell and why they should buy from you instead of 47 other stores?

Most stores fail here. Generic taglines. No differentiation. Unclear benefit.

Relevance Does your page match what the visitor expected when they clicked?

If your Facebook ad shows “handmade leather bags” but your homepage hero image is random accessories, the visitor’s brain goes “wrong place” and they leave.

Clarity Can someone understand how to buy without thinking?

Is the next step obvious? Is pricing clear? Are product details readable? No jargon?

Urgency Is there any reason to buy now instead of later?

Limited stock. Sale ending. Seasonal relevance. Something that makes waiting feel like a bad idea.

Anxiety What makes people nervous about buying?

No reviews. Unclear return policy. Sketchy checkout design. Missing contact information. Site looks unprofessional.

Distraction What pulls attention away from buying?

Popups. Autoplay videos. Blinking banners. Seven different calls-to-action. Social media icons everywhere.

How to do heuristic analysis:

Open your homepage in an incognito window.

Set a timer for 5 seconds. Look at the page.

What did you understand? What’s the main message? What’s the next step?

If you can’t answer clearly, your visitors can’t either.

Then go through your funnel:

Homepage
Collection page
Product page
Cart
Checkout

For each page, score it against the LIFT factors. Where does it fail?

Write down every confusion point. Every unclear element. Every moment where you’d hesitate if you were a first-time visitor.

Don’t just do this yourself. Get other people. Your designer sees the site differently than your accountant. Different perspectives catch different problems.

Step 3: Analytics Deep Dive

Question: What is actually happening?

Now you look at data to see where people drop off.

Open Shopify Analytics or Google Analytics.

Find your biggest leaks:

Look at the funnel:

How many people visit?
How many view products?
How many add to cart?
How many reach checkout?
How many complete purchase?

Where’s the biggest drop-off?

If 1,000 people visit but only 50 view products, your homepage isn’t working.

If 500 view products but only 20 add to cart, your product pages aren’t convincing.

If 100 reach checkout but only 8 complete, something’s broken in checkout.

Device breakdown:

Mobile vs. desktop conversion.

If mobile is way lower, you have a mobile problem. If mobile is 70% of traffic but 0.5% conversion while desktop is 30% of traffic at 2.8% conversion, fix mobile first.

Traffic source breakdown:

Email converts at 9.6%. Facebook ads convert at 1.2%. Google organic converts at 2.8%.

Which sources are bleeding money? Which are working?

Different traffic sources have different intent. Facebook is cold traffic (just browsing). Google is warm traffic (actively searching). Email is hot traffic (already know you).

Don’t expect Facebook to convert like email. But if Google organic is at 0.4%, something’s wrong with your SEO traffic or landing pages.

Browser issues:

Sometimes one browser just breaks.

I’ve seen stores where Safari was 65% lower conversion than Chrome because a form field didn’t work on Safari. Easy fix once you know.

Analytics tells you WHAT is happening. It doesn’t tell you WHY.

Step 4: Session Recordings

Question: How are real people actually using your site?

This is where everything clicks.

You WATCH real visitors navigate your store.

Use tools like Microsoft Clarity (free), Hotjar, or Lucky Orange.

They record actual sessions. You see the cursor move. Clicks. Scrolls. Where people hesitate. Where they rage-click. When they leave.

What to look for:

Confusion moments

Clicking things that aren’t clickable
Hovering over elements trying to figure out what they do
Scrolling back and forth looking for something
Refreshing the page multiple times

Friction points

Form fields they can’t fill out
Buttons they can’t find
Information they’re searching for but can’t locate
Mobile users trying to tap tiny buttons

Breaking points

Where they give up
What they were doing right before abandoning
Which page causes them to leave

I watched a session where someone spent 4 minutes on a product page, added to cart, went to checkout, then left.

Why?

They scrolled to the bottom of the checkout page looking for the total price including shipping. Couldn’t find it. It was at the top but below the fold on mobile.

They didn’t scroll back up. They just left.

€180 order lost because shipping cost wasn’t visible without scrolling.

Watch 20 sessions. You’ll see patterns.

“Oh, EVERYONE is clicking that image thinking it’s a button. It’s not a button.”

“Huh, people keep abandoning at this exact form field. What’s confusing about it?”

“Wait, three people just tried to zoom into this product detail and it doesn’t zoom. They all left.”

Analytics tells you what’s happening. Session recordings show you why.

Step 5: Qualitative Research (Surveys)

Question: Why didn’t they buy?

Just ask them.

Run surveys. Two types:

Exit surveys When someone’s about to leave without buying, show a popup:

“Quick question: What’s holding you back from buying today?”

Keep it open-ended. Let them type whatever they want.

You’ll get answers like:

“Shipping is too expensive”
“Wanted to check other stores first”
“Not sure about sizing”
“Looks kind of sketchy”
“Can’t find my size”

These are real objections from real people. Not guesses.

Post-purchase surveys Email recent buyers (within 48 hours) and ask:

“What almost stopped you from buying?”

Seriously. That’s the question.

The answers are gold.

“I almost didn’t buy because I couldn’t tell if you were a real company.” “I was worried about return policy.” “Shipping took too long to calculate.” “Payment page looked weird on my phone.”

These people DID buy. But they almost didn’t. Fix what almost stopped them, and you’ll convert more of the people who currently don’t buy.

Step 6: User Testing

Question: Can a real person actually complete the task?

Find 5 people who match your target customer.

Not your mom. Not your designer. Actual potential customers.

Give them a task: “Buy a birthday gift for your partner under €100.”

Watch them try to do it. Don’t help. Don’t explain. Just observe.

They’ll get stuck in places you never expected.

“Wait, where’s the size chart?” “How do I know if this ships to my country?” “What does this icon mean?” “Is this price in euros or dollars?”

Things that are obvious to you (because you built the site) are confusing to them.

User testing reveals assumptions you didn’t know you had.

I watched someone try to buy a leather bag.

They wanted to know the dimensions. Couldn’t find them. Checked product description. Not there. Clicked everywhere looking for measurements.

Gave up.

The dimensions were in the photos (shown next to a ruler). The user never looked at photos because they were “just browsing sizes first.”

Added dimensions to product description. Problem solved.

You can use tools like UserTesting.com for remote testing. Or just grab 5 friends who fit your target market and pay them €50 each for 30 minutes.

Putting It All Together: The Diagnosis Sheet

After running through these six steps, you’ll have a list. Maybe 50-100 insights.

Now you need to turn insights into action.

I use a simple prioritization framework:

Test — Big potential impact, clear hypothesis, worth A/B testing Fix — Obvious problems, just fix them (broken buttons, missing info, unclear copy) Instrument — Need better tracking or analytics setup first Park — Interesting but low priority, revisit later

Example from a recent audit:

Test:

Add reviews to product pages (saw people searching for social proof)
Show shipping cost on product page (exit survey said “surprise costs”)
Simplify checkout from 28 fields to 14 (heatmaps showed drop-off at field 12)

Fix:

Safari mobile checkout broken (technical analysis caught this)
Trust badges removed (heuristic analysis showed they increased anxiety)
Product dimensions missing from descriptions (user testing revealed)

Instrument:

Set up conversion tracking for each checkout step (can’t optimize what you can’t measure)
Add event tracking for “add to cart” by device (need to know where mobile breaks down)

Park:

Loyalty program idea (good but not urgent)
Subscription option (need to fix core conversion first)
Live chat widget (would help but expensive right now)

Focus on Fix items first (they’re quick wins). Then Test items (biggest potential impact). Then Instrument items (better data for future decisions).

Common Misdiagnoses (And What’s Really Wrong)

Here’s what amateur diagnosis looks like vs. professional diagnosis:

Amateur says: “Cart abandonment is high. Add exit popup with 10% discount.”
Professional finds: 48% of abandoners said “just browsing” (unavoidable) and 35% said “shipping costs surprised me” (fix by showing cost earlier).

Amateur says: “Conversion is low. Need better button color.”
Professional finds: Button color is fine. Problem is mobile checkout doesn’t work on 40% of devices.

Amateur says: “Nobody’s buying. We need more traffic.”
Professional finds: Traffic is fine. Product pages have zero reviews, no trust signals, unclear return policy. Trust problem, not traffic problem.

Amateur says: “Let’s copy what Amazon does.”
Professional finds: Amazon ships from one warehouse. This client ships from multiple suppliers with different delivery dates. Amazon’s checkout wouldn’t work for their business model.

Amateur says: “Add more trust badges.”
Professional finds: Trust badges are actually REDUCING conversion because they draw attention to security when security wasn’t the concern. Real problem was unclear value proposition.

The difference: amateurs guess. Professionals diagnose.

When to DIY vs. When to Hire

You can run technical analysis and analytics yourself.

Anyone can check if their mobile site works. Anyone can look at Shopify Analytics and see where drop-off happens.

Heuristic analysis? Also doable. Use the LIFT framework. Evaluate your pages. Write down problems.

Where it gets harder:

Session recordings require knowing what to look for. It’s easy to watch 20 sessions and think “huh, people click around a lot” without catching the actual pattern.

Qualitative research requires asking the right questions and interpreting answers correctly. People say things that aren’t literally true (“shipping was too expensive” often means “I didn’t expect shipping to cost anything”).

User testing requires moderation skills. Asking leading questions ruins the whole exercise.

And the biggest challenge: you’re too close to your own store.

You know how it’s supposed to work. You can’t see it like a first-time visitor anymore.

That’s why external audits help. Fresh eyes catch things you’ve stopped seeing.

When to hire someone:

You’ve fixed obvious technical issues but conversion is still stuck
You’re getting traffic but can’t figure out where it’s breaking
You’ve tried random tactics for 6 months with no improvement
You don’t have time to learn analytics, session recording tools, and research methods
You need someone who’s done this 100+ times and knows the patterns

When to DIY:

You haven’t checked if your mobile site works yet (do this first)
You haven’t looked at Shopify Analytics device breakdown (do this second)
You have obvious problems (slow site, broken checkout, no product info)
You have time to learn and you’re methodical about it

Either way, don’t guess. Diagnose.

The Uncomfortable Truth About Diagnosis

Most people skip this entirely.

They want the fix. The tactic. The “best practice” that’ll magically increase conversions.

So they install trust badges without knowing if trust is the problem.

They change button colors without knowing if the button is even the issue.

They copy competitors without understanding why competitors do what they do.

Then they wonder why nothing works.

I’ve done enough audits to know: 80% of conversion work is diagnosis. 20% is execution.

The stores that win aren’t the ones with the prettiest designs or the most A/B tests.

They’re the ones who actually understand what’s broken and fix THAT instead of random things.

If you’re stuck at 0.8%, you don’t need 15 new tactics.

You need to know which ONE problem is causing 80% of your leaks.

Fix that. Then move to the next biggest leak. Then the next.

Systematic diagnosis. Not guesswork.

That’s how you get from 0.8% to 3%. For a broader view of what optimization looks like once you’ve finished diagnosing, the ultimate guide to CRO for ecommerce covers the full execution side.

What To Do Next

If you’ve never diagnosed your store properly, start here:

Week 1: Technical check

Test your site on iPhone Safari, Android Chrome, and desktop
Go through full checkout flow on each device
Check PageSpeed Insights
Fix anything that’s broken

Week 2: Analytics audit

Open Shopify Analytics
Look at device conversion breakdown
Look at traffic source conversion
Find your biggest drop-off point in the funnel

Week 3: Watch 20 session recordings

Install Microsoft Clarity (free)
Watch 20 real sessions
Note where people hesitate, get confused, or leave
Look for patterns

Week 4: Run a survey

Send email to 100 recent buyers
Ask: “What almost stopped you from buying?”
Read every answer
Group similar themes

Do this before changing anything. Before testing button colors. Before hiring a designer. Before installing apps.

Know what’s actually broken first.

Then fix it.

Philip Wallage runs BTNG.studio, a conversion-focused design service for e-commerce stores. After diagnosing 100+ stores, he’s learned that most conversion problems aren’t what owners think they are. If you want an external diagnosis of what’s actually stopping your store from converting, check out the UX QuickScan.