A lower checkout-completion percentage does not by itself mean Shopify checkout is losing more customers. Before you change design, payment methods, or shipping, check that the old report and the new one count the same people, steps, channels, and sequence.
This comes up most often after a migration or an analytics change. A team sees a familiar label like 'reached checkout', assumes the definition held, and starts fixing a denominator problem as though it were a customer experience problem.
Shopify changed session measurement and default bot filtering from September 21 to 23, 2026. If the break begins there, use our September 2026 Shopify Analytics guide before comparing the new conversion rate with an older baseline.
What does Shopify count as checkout conversion?
Shopify counts checkout conversion rate as sessions that completed checkout divided by sessions that reached checkout. Its behaviour reports can use an open funnel, which includes sessions entering at any step, or a closed funnel, which requires cart addition, checkout, and purchase in that order.
Shopify's behaviour-report documentation explains the open and closed funnel definitions used in Analytics.
Read Shopify's funnel definitions →Why reached checkout can rise without more buying intent
Accelerated checkout buttons can skip the cart, and an open funnel still counts the checkout visit when the earlier storefront steps never happened. Shopify also notes that checkout counts can exceed add-to-cart counts when customers arrive through channels such as Instagram Shopping or Facebook Shops.
- ✓Buy it now and wallet buttons shorten the recorded path
- ✓Open funnels count people who enter at checkout
- ✓External sales channels do not always reproduce the storefront cart sequence
- ✓Consent rules and bot filters can move the session denominator
How to compare a migration without inventing a checkout crisis
Start with the outcomes that survive a change of platform: paid orders, net sales, refunds, and unique purchasing customers over matched periods. Then line up timezone, traffic sources, devices, markets, consent rules, and campaign mix.
Then rebuild the funnel from equivalent steps. Do not compare the old platform's payment-step arrival with Shopify's checkout-page arrival, and do not compare visitors with sessions. Write every definition down next to the number, so the next reporting change cannot quietly rewrite your benchmark.
When the drop is real, where should you look first?
Open the recent abandoned checkouts and group them by the last reliable state. No address usually means friction early in checkout. An address with no shipping method points at rates or eligibility. A shipping method selected with no payment points at the gateway, trust, authentication, or payment-method fit. If a newer store has almost no orders at all, start earlier with why you are not getting sales on Shopify, which checks traffic and setup first.
- ✓Test the highest-volume market and device yourself
- ✓Check shipping-rate availability with a real address
- ✓Confirm that the local payment methods you expect actually appear
- ✓Compare the express and standard checkout paths
- ✓Look at payment declines separately from customer abandonment
Which Shopify issues distort checkout diagnosis?
The usual false starts are shipping unavailable errors, currency changing at checkout, and payment methods hidden by market eligibility. A missing confirmation email lands after the purchase, so it can look like a failed checkout when the order actually exists.
Bot activity is the other source of noise. Shopify's analytics field reference says conversion rate can appear lower because of bot traffic, and recommends filtering for human traffic where that is available.
Build the measurement contract before the CRO backlog
Write down which report, filters, funnel type, and event definitions the business will use. Then go after the largest loss point you have actually verified. That turns CRO from a reaction to a percentage into a sequence of tests you can defend.
Once the measurement is trustworthy, your support inbox is the cheapest ranked list of what to fix on the page.
Build the CRO backlog from support →If your checkout numbers and customer behaviour tell different stories, we can isolate the measurement problem and the real conversion blockers in one scoped sprint.
Scope a Shopify CRO sprint →
