Attribution & Conversion Counting

Click-Through Conversions vs View-Through Conversions

In short: Both decide whether a platform gives an ad credit for a conversion, but they require completely different evidence of interaction. Click-through conversions require a recorded click - the person is tied to the ad through direct action - and any conversion inside that click's attribution window is credited back to it. View-through conversions require only that the ad was served; if the person converts later without ever clicking, it can still be credited, but only within a separate, typically shorter, view window. Bidding systems and most advertisers treat click-through as the trustworthy default and view-through as a weaker, supplementary signal. If you need a number to optimize spend against with confidence, use click-through; if you want visibility into whether unclicked impressions might be doing something, look at view-through, but verify it with a lift test before you believe it.

By the AdFlint research team · Fact-checked against current Google and Meta platform behavior · Last reviewed July 2026

Click-Through Conversions

Counts conversions from people who clicked your ad first, credited back to that click within the campaign's click attribution window.

The stricter of the two counting paths: a recorded click ties the person to the ad, then any conversion inside the click window is credited back to it. These are the numbers most advertisers optimize on, and platform bidding systems lean on them heavily. The subtlety is that a click conversion is not proof of causation either, it just documents a much stronger interaction than an impression.

Full definition

View-Through Conversions

Counts people who were served your ad, never clicked it, and converted later within the view attribution window.

Both Google and Meta report these separately from click conversions, and both apply a shorter default window to views than to clicks. They matter most for display, video, and upper-funnel social, where the ad is seen far more often than it is clicked. The persistent error is adding view-through conversions straight into blended ROAS; without a holdout or lift test you have no evidence the impression caused anything.

Full definition

Side by side.

The differences that actually change what happens in your account.

	Click-Through Conversions	View-Through Conversions
What has to happen first	A recorded click on the ad - the platform ties the person to that specific interaction.	An impression served to the person - no click required, and no engagement beyond basic delivery.
Attribution window used	The campaign's click-through window - defaults are a single number set by the platform (commonly 7 to 30 days) and can typically be extended further, up to 28 days on Meta or up to 90 on Google.	A separate, shorter view-through window - about 1 day on Meta with no way to extend it, versus configurable from 1 up to 30 days on Google Ads.
Strength of evidence	Strongest available signal short of an actual purchase - the person demonstrably engaged with the ad.	Weakest signal counted as a conversion at all - it proves delivery, not engagement.
Bid strategy reliance	Primary input that Smart Bidding, Advantage+, and other automated bidders train and optimize on.	Reported alongside bidding data, but commonly excluded from or down-weighted in what the bidder optimizes toward.
Reporting placement	The standard conversions column - what most dashboards mean by default when they say conversions.	Broken out into its own column or segment, separate from the headline conversions metric.
Where it dominates volume	Search, Shopping, and any format where the click itself is the primary action being sold.	Display, video, and upper-funnel social, where the ad is seen far more often than it's clicked.
What it costs you when misread	Treating it as fully causal ignores that people who click are already more likely to convert regardless of the ad.	Folding it into blended ROAS without a holdout inflates the ad's apparent impact with no proof of causation.

What actually separates them.

Click-through conversions require a logged click event tying the user to the ad; view-through conversions require only that an impression was served, with no click at all.

The two use separate attribution windows, and the view-through window is almost always shorter than the click-through window on the same platform.

Automated bid strategies are built to trust and optimize against click-through data first; view-through numbers usually sit outside that optimization loop unless you deliberately fold them in.

Click-through conversions show up in the standard conversions total by default; view-through conversions are reported as a distinct, separately labeled figure you have to go looking for.

Click-through conversions are harder to inflate by accident since a click is a discrete logged action, while view-through totals can grow just from running more impressions, regardless of whether anyone was influenced.

Which one should you use?

Use Click-Through Conversions when

You're optimizing Search or Shopping campaigns where the click itself is close to a purchase-intent signal.
You need a conversion number feeding a Smart Bidding or Advantage+ campaign that you trust represents real engagement.
You're reporting results to a client or to finance and need the more defensible, harder-to-inflate figure.
You're diagnosing why a campaign's cost-per-conversion moved and want to isolate the number tied to actual clicks.

Use View-Through Conversions when

You're running display or video creative meant to build awareness, not necessarily to earn a click.
You want to see whether people who saw but didn't click an ad later converted through another path, like a direct visit or a branded search.
You're comparing upper-funnel formats against each other and click volume alone would undersell the quieter ones.
You're building a full-funnel report and want the unclicked-impression number visible and labeled, not silently folded into your main conversion count.

Common questions.

Which one should I trust when the numbers disagree?

Click-through conversions by default - a click is a discrete, logged action that's harder to inflate, while a view-through conversion only proves the ad was delivered, not that it did anything. Treat view-through as a secondary signal to investigate, not a number to bid against without more evidence.

Why is my total conversions column lower than click-through plus view-through added together?

Most dashboards report view-through conversions as a separate column specifically so they aren't silently summed into your headline total. If a report you're looking at does add them together, check whether it's labeled as a blended or total view - that's a deliberate choice, and it should be labeled as such.

Can a single conversion be counted as both click-through and view-through?

No. Platforms use a hierarchy where a click takes priority over a view when both happened, so a person who saw the ad and later clicked it gets counted as a click-through conversion, not a view-through one. View-through only applies when no click was ever recorded for that person.

Do view-through conversions actually mean the ad worked?

Not on their own - they only show the ad was served before the conversion happened, not that it caused it. Confirming causation needs a holdout group or lift test comparing people who saw the ad against people who didn't; without that, treat view-through numbers as a hypothesis, not proof.

Or stop choosing between them.

AdFlint picks the setting, writes the ads, and keeps optimizing inside the Google and Meta accounts you already own.

Try AdFlint free See how it works

Related comparisons

All comparisons