View-Through Conversions: Should You Count Them?

Your Performance Max campaign reports 41 conversions. Your CRM shows 12 new inquiries. Nobody is lying. The gap is usually a metric almost no agency explains to the person paying for it: the view-through conversion.

A view-through is credit the platform awards itself for an ad the person never clicked. It is defined in the vendor's own documentation. What the vendor does not do is tell you whether to include it in the return you report to yourself. Here is the honest answer, and the exact columns to use.

What is a view-through conversion, exactly?

A view-through conversion is recorded when someone is served your ad, does not interact with it, and later converts on your site. Google's own definition is blunt about that: the column tells you when customers view, but don't interact with your ad, and later complete a conversion, per Google Ads Help. No click. No form fill on the ad. No hover.

The crediting rule matters more than the definition. For Display Network ads, Google gives the view-through to the last viewable impression. Under Active View, an impression counts as viewable when at least 50% of the ad is onscreen for at least 1 second. That is the entire bar for a conversion to be booked in your name.

One thing Google does get right: view-through conversions automatically exclude anyone who also interacted with another of your ads, and they are kept out of the main Conversions column. They only appear in the View-through conversions column and in All conversions.

Does a view-through require the person to have seen the ad?

No. It requires a viewable impression, which Google defines as 50% of the ad's pixels onscreen for 1 second — a machine-measurable event, not a human one. A 300x250 unit that scrolls past a knee in someone's feed while they thumb toward a recipe clears that bar. Eyeballs are not part of the specification.

Video is looser still. For video campaigns, the last impression of the ad gets credit, and Google spells out that an impression is not a view: a view is counted when someone watches 30 seconds, or the whole ad if it is shorter, or clicks on it. So a pre-roll that ran for two seconds before the skip button got hit can carry a view-through.

There is one more quiet asterisk. Google says view-through conversions from browsers that do not allow cross-site cookies cannot be reported. So the number is not just soft — it is soft and unevenly measured across your audience.

Why do Performance Max and Advantage+ lean on this metric so hard?

Because both campaign types spend most of their budget on impression inventory — Display, YouTube, Discover, Gmail, feeds — where clicks are rare and impressions are effectively unlimited. View-through credit is the only mechanism that lets that inventory look like it produced leads.

Meta runs the same play under different names. Its standard attribution setting counts a view-through as an event that happened within 1 day after an impression of your ad, and an engage-through as an event within 1 day of a non-link click — including a video played for 5 seconds, per Meta's attribution documentation. Click-through is the only setting that requires a link click.

This is not a conspiracy. It is the metric doing exactly what the docs say. The problem starts when a reporting deck sums that column with your Search conversions and calls the total your ROAS. Our read on Performance Max for service businesses and on Meta Advantage+ is the same: judge them on click-through credit first, then decide if the rest is real.

Which campaign types inflate worst on view-throughs?

Display and video inflate worst, because their credit needs only an unclicked impression and Google's default view-through window is 1 day, while a click-through conversion gets a 30-day window by default. Same 24 hours of exposure, wildly different evidence behind the credit.

Verdict: the metric that should scare you is not the view-through. It is the engaged-view conversion. Google says an engaged-view conversion is counted when a skippable in-stream ad is watched for at least 10 seconds (5 seconds for in-feed and Shorts) with no click, and that engaged-view conversions appear in both the Conversions and the All conversions columns — across Video, Display, Demand Gen and Performance Max campaigns (Google Ads Help). That is non-click credit sitting inside the column your Smart Bidding strategy optimizes toward.

Which means a Target CPA that looks healthy may be steering toward cheap 10-second video watches. If you are running automated bidding, read that sentence twice, then read our take on bid strategy automation.

Should you include view-throughs in the ROAS you report to yourself?

No. Report click-through and view-through as two separate lines, never one blended number. The board number for a service business is booked calls and closed revenue — a metric that survives contact with your calendar, not one that depends on a 1-second impression.

That is not a moral position, it is a decision-quality one. You use the number to move budget. If the number contains credit for conversions that were going to happen anyway, you will move budget toward the campaign that is best at being present, not the one that is best at causing revenue. Every dollar you shift on a blended number is a dollar shifted on a metric the platform grades itself with.

So the working rule we use: no view-through, no engaged-view, and no modeled credit goes into a cost-per-booked-call calculation. If it did not produce a click or a call, it does not get to price your media. See why most service businesses cannot tell which ad dollar made them money for the fuller version of this argument.

When is a view-through actually informative?

In exactly one situation: when you are running a controlled test and the view-through number is a supporting signal, not the verdict. Meta ships an incremental attribution model that optimizes and reports on conversions its models predict were caused by the ad — a tacit admission that its standard 1-day view-through setting does not do that.

So use it as a directional trace on high-frequency, long-consideration channels: if you turn a video campaign off in one region, hold everything else flat, and total booked calls drop, the view-through column was telling you something. If you turn it off and nothing moves, the view-through column was telling you a story. The test is the evidence. The column is a hint.

The same logic drives a brand bidding incrementality test — the only honest way to price a channel that mostly gets credit for demand it did not create.

How do you separate click-through and view-through in reporting?

In Google Ads, use the Conversions column for click-based credit and read view-throughs only in the View-through conversions column — Google keeps them out of Conversions by design. For engaged-view credit, which is not kept out, segment by Ad event type: Google's documentation says this splits click-through conversions (labelled Clicks) from engaged-view conversions (labelled Interactions).

• Build every performance table on Conversions, not All conversions — All conversions is the column that carries view-through credit.
• Segment Video, Display, Demand Gen and Performance Max by Ad event type, so engaged-view credit is visible instead of buried.
• Set your view-through window deliberately. It defaults to 1 day and can be pushed to 30 days or 4 weeks — a longer window does not create more truth, only more credit.
• Reconcile against a system the platform does not own: CRM, call tracking, or booked appointments. If a campaign's leads do not appear there, the conversions were not leads.
• Keep click-through and view-through as two rows in the monthly report, permanently. The moment they are summed, the argument is lost.

For anything with a sales cycle longer than a week, the reconciliation step matters more than the columns. We walk through it in tracking Google Ads conversions with long sales cycles.

What happens to your channel ranking when you strip them out?

The arithmetic is unforgiving: if 40% of a campaign's All conversions came from unclicked impressions, its cost per counted lead is 1.67x what the dashboard showed. A campaign at a reported $90 cost per conversion is really at $150 — and it just fell below the Search campaign you were about to cut budget from.

Rebuild the table once on click-through credit and the order usually changes. Search and Local Services Ads climb. Display, YouTube and the broad-reach half of Performance Max fall. Nothing about the underlying business changed. You just stopped grading the media on a metric the media supplier authored.

Then apply a kill rule. A channel that produces no qualified, click-attributed leads in 90 days gets cut — no matter how good its All conversions column looks.

If your reports still show one blended conversion number, you do not have an attribution problem — you have a reporting-integrity problem, and it is costing you real budget. Our paid ads service is built on click-attributed, CRM-reconciled numbers, month-to-month, with no lock-in. Get my free audit and we will show you which of your campaigns is living on view-through credit.