Low conversion does not mean every page should add more claims, trust blocks, and buttons. Collections, product pages, guides, cart, and checkout each own a different job. Define the job first so CRO has evidence and boundaries.
22 parts · 1 checkpoint · 1 free / 21 member parts →
Note Do these three things today
- Write one sentence for each of the five positions — collection, product, guide, cart, checkout — saying what it must help the shopper finish. 2. Find the matching events (view_item_list, view_item, begin_checkout, purchase) and mark where the break is. 3. Change one task on one page, and note the review date plus the pages left untouched.
Key point In one line
Assign page jobs before optimizing modules
- Collection and search: view_item_list / select_item
- Product page: view_item / add_to_cart
- Guide / comparison page: scroll / internal click / select_item
- Cart: view_cart / begin_checkout
- Checkout / payment: add_shipping_info / add_payment_info / purchase
How to diagnose a Shopify conversion funnel: verify data before assigning page roles
Find a trustworthy funnel break, then give one page one repair or experiment job.
This lesson answers
Locate the weak step before assigning a page a learning job
The 14-day cohort tells three different stories. Eighty add-to-carts from 1,000 product views gives an 8.0% PDP-to-ATC rate. Fifty-two checkout starts from 80 add-to-carts gives 65.0%. Thirty-one purchases from 52 checkout starts gives 59.6%. Each step uses the prior step as its denominator. The 31/1,000 total conversion rate cannot replace three page questions.
Try it first
Narrow the choice set so users can find likely-fit products faster.
- Key event
- view_item_list / select_item
- Evidence
- Collection-to-PDP click-through, filter use, backtracking, onsite search terms.
- Common friction
- Cannot find: categories are broad, filters are weak, product cards lack clues.
- Wrong move
- Putting full sales copy on the collection page and making users reread the PDP.
- First move
- Start with filters, collection intro, and key product-card fields.
Complete persuasion: fit, reason to buy, proof, and risk reversal.
- Key event
- view_item / add_to_cart
- Evidence
- ATC rate, image/video interaction, FAQ opens, review clicks, support questions.
- Common friction
- Cannot understand or trust: wrong benefit order, late proof, unclear risk.
- Wrong move
- Adding more modules without answering the earliest doubt.
- First move
- Reorder first screen, specs, real proof, reviews, FAQ, and return promise.
Explain differences and move users from comparison to a specific product.
- Key event
- scroll / internal click / select_item
- Evidence
- Comparison clicks, recommendation-card clicks, post-click PDP quality.
- Common friction
- Cannot choose: comparison is too promotional and lacks scenarios or limits.
- Wrong move
- Turning the guide into another broad brand story.
- First move
- Organize comparison by scenario, budget, size, usage frequency, and constraints.
Reduce pre-checkout uncertainty: cost, inventory, promo, shipping, returns.
- Key event
- view_cart / begin_checkout
- Evidence
- Cart-to-checkout start, promo-code attempts, shipping questions, inventory notices.
- Common friction
- Cannot calculate: total cost, shipping, timing, or promo terms appear too late.
- Wrong move
- Retelling the brand story when the user is ready to buy.
- First move
- Surface cost, delivery timing, free-shipping threshold, return summary, support entry.
Help users complete the action by reducing form, error, payment, and mobile friction.
- Key event
- add_shipping_info / add_payment_info / purchase
- Evidence
- Payment failures, address errors, field exits, mobile conversion, error codes.
- Common friction
- Cannot complete: hard fields, unavailable payment, unclear errors.
- Wrong move
- Testing PDP titles while payment is failing.
- First move
- Check payment methods, address fields, localization, errors, and mobile input first.
Note Open case: one path, four page jobs
Collections help the shopper choose; the product page confirms the item solves the problem; the cart makes total price and promise clear; checkout completes. When one page takes another page's job, the break gets recorded in the wrong place.
Pro lesson · free with an account until 2027-04-26 · Parts 2-22 · 21 parts
Members: full exercises and records
This part includes
- Turn a funnel break into one page job
- Verify the reading with events and orders
- Write a stop line for the repair or experiment
1 self-check question
An account is free; sign in to read this part.
Checking your membership…
FAQ
- When conversion rate is low, should I change the homepage or landing page first?
Not automatically. Read the GA4 event chain first. Weak select_item may point to collection choice; weak add_to_cart may point to PDP persuasion; weak begin_checkout to purchase may point to cart or checkout friction.
- If several pages look broken, which one should this CRO round change?
Choose one page role and one event break. If this round is about collection choice, do not also change PDP hero, cart shipping, and checkout fields. Otherwise the 7-14 day readout will not explain what worked.
What to learn next
After page-role diagnosis, continue with the entry promise and product-page handoff
Sources
New Store Launch Self-Check
Free accountWrite payment, policy, test-order, mobile checkout, and flow failures back into the diagnosis table so system issues are not misread as page-copy issues.
Sources: 6Last checked
Source check: 2026-10-03 (opens in a new tab)Checked
Supports
These sources define event chains, evidence boundaries, and diagnosis steps. Actual page changes still need your GA4, Shopify orders, behavior samples, and test orders. Use events to separate list exposure, product view, add-to-cart, checkout, and purchase. These sources define event chains, evidence boundaries, and diagnosis steps. Actual page changes still need your GA4, Shopify orders, behavior samples, and test orders. Use events to separate list exposure, product view, add-to-cart, checkout, and purchase.
GA4 ecommerce events and item fields (opens in a new tab)Checked
Supports
Keep item, variant, list, coupon, and order fields on the same event evidence chain. Keep item, variant, list, coupon, and order fields on the same event evidence chain.
GA4 Funnel exploration (opens in a new tab)Checked
Supports
Use funnel steps to see where users complete or leave. Use funnel steps to see where users complete or leave.
Shopify Sales reports (opens in a new tab)Checked
Supports
Use orders, products, channels, and discounts to test whether page promises carry into real transactions. Use orders, products, channels, and discounts to test whether page promises carry into real transactions.
Microsoft Clarity masking (opens in a new tab)Checked
Supports
Before behavior samples enter review, confirm email, phone, address, payment, and form inputs are masked. Before behavior samples enter review, confirm email, phone, address, payment, and form inputs are masked.
Baymard checkout usability research (opens in a new tab)Checked
Supports
Break cart and checkout issues into specific friction, not only price. Break cart and checkout issues into specific friction, not only price.