How the count setting works
Each conversion action in Google Ads carries a count setting with two choices. Every records each conversion that follows an ad interaction, so a person who buys twice after the same click produces two conversions. One records at most a single conversion per interaction, however many times the action is completed afterwards.
The setting is applied per action, not per account, so a shop can count purchases with Every while counting newsletter sign-ups with One. It works within the conversion window: once that window closes, a new interaction starts the count again. It is a counting rule and nothing more — it does not change what is tracked, only how the tracked events are totalled.
Why the count setting matters
It is the difference between a lead count and a form count. On a service business, one person who fills in the enquiry form, hears nothing, then fills it in twice more has produced one lead and three submissions. Left on Every, the account reports three, the cost per lead looks a third of what it is, and every decision made from that number is wrong.
It matters just as much in the other direction. An online shop that sets a purchase action to One will lose the second and third orders from a returning customer, understate revenue and mislead value-based bidding into thinking the traffic is worth less than it is. Whatever you choose, the bidding system inherits it, because it optimises against the totals it is shown.
Common mistakes with the count setting
Accepting the default without thinking is the everyday one, and it usually bites lead generation accounts hardest, because forms are resubmitted far more often than people assume — a slow confirmation page or an impatient tap on a mobile connection is enough. Anyone who has watched enquiries arrive in Nepal on a patchy mobile network will recognise the pattern.
The second mistake is using the setting to solve a problem that belongs elsewhere. Genuine double counting caused by two tags firing for the same event is a tracking fault, and switching to One only hides it while leaving the underlying data wrong. The third is changing the setting quietly. Conversion totals shift from the change date onwards, so any before-and-after comparison across it is meaningless unless the change is recorded.
How to act on it
Ask a simple question of each action: if the same person does this twice, is that worth twice as much to the business? Purchases and orders usually are. Enquiries, quotation requests, sign-ups, bookings for a single appointment and call-back requests usually are not.
Set each action accordingly, note the date beside any change, and check the totals for a period afterwards to confirm the effect matches what you expected. If the numbers move in a way you cannot explain, the cause is more likely to be a tagging fault than the count rule, and that is where to look next.