You’re right in a sense. A “conversion” according to Shopify is not necessarily completing checkout with an order number. It could be any action worth noting. It can include filling out a form, signing up for a newsletter, and of course making a purchase. So it’s not always orders/sessions.
That’s a sharp observation, and your math is correct.
The discrepancy you’ve found isn’t an error, but a key difference in how Shopify’s dashboard calculates its conversion rate. Instead of the standard formula of (Total Orders / Total Sessions), the dashboard metric is based on unique converting sessions, using the formula (Sessions with at least one Order / Total Sessions).
This is because a single customer can place multiple orders within one 30-minute session, but Shopify only counts this as one “converting session.”
Your own data shows this perfectly in action: your 520 orders were actually generated from just 221 unique sessions (8,998 * 2.46%), which indicates you have a healthy number of customers placing multiple orders per visit.
So, you’re right that ‘orders’ and ‘converting sessions’ are different, and that’s the source of the discrepancy you’re seeing.
While you reasoning sounds nice, this particular shop has near 0% repeat sales, within the time period shown in the screenshots. So your logic cannot be the source of the discrepancy shown.