View-Through Conversion

Three conditions, and every platform applies all three:

1. An impression. The platform served the ad and, by its own definition, it was shown. On the Google Display Network and video ads, and on Microsoft's Audience ads, the last impression before the conversion takes the credit.
2. No click. The person did not click the ad. On Google Ads and Microsoft Advertising a click on any of your other ads on that network also removes the view-through, because the conversion is then counted as a click.
3. A conversion inside the window. The person converted on your site, recorded by the platform's own pixel or conversion API, within the view-through window that started at the impression.

Video platforms add a class in between, the engaged view. Google counts an engaged-view conversion when somebody watches at least 10 seconds of a skippable YouTube ad (5 seconds on in-feed and Shorts ads) and converts inside the engaged-view window, 3 days by default for web conversions, and it puts those in its Conversions column. Meta renamed its version engage-through in March 2026: a non-link interaction with the ad, or 5 seconds of video, followed by a conversion within 1 day. TikTok's engaged view-through starts at 6 seconds watched.

Every platform runs the same four rules. What differs is the numbers inside them.

• The windowA view-through only counts inside a period that starts at the impression. It is a setting on the conversion action or the ad set, and the longer it is, the more view-throughs the platform records, as Google's own help text says. See attribution window.
• The no-click ruleA click wins. Google and Microsoft exclude anyone who clicked any of your ads on their network, and TikTok attributes each conversion to one touchpoint: a click, a view or an engaged view.
• Deduplication inside one platformWithin one platform a conversion is normally counted once, for the last impression or click. LinkedIn is the one to check: under its "Last touch, each ad set" model, every ad set that reached the member inside the window is credited with the conversion.
• No deduplication across platformsMeta does not know that Google served the same person an impression, and neither knows what LinkedIn did. Each platform counts its own view-throughs in full, so one lead can be a view-through conversion in three dashboards at once.

Where the view-throughs land in the report differs as well, and it decides whether the headline figure includes them. Google Ads keeps them out of its Conversions column for Display and video and shows them under View-through conversions and All conversions, except in the campaign types that bid on them. Meta's Results column includes them whenever view-through is part of the attribution setting. LinkedIn's Conversions metric counts actions after clicking on or seeing the ad. Microsoft adds them to All conv. by default. Before you quote a platform's number, check which column it came from.

The defaults as read on each platform's own documentation on 28 September 2026, with the range you can set in brackets where the platform documents one. Every one of them can be changed; on Google Ads, LinkedIn and Microsoft a changed window only applies to conversions recorded after the change.

Some notes on the rows. Meta's view window is 1 day or nothing: the 7-day and 28-day view windows stopped being returned by its Insights API on 12 January 2026, and the default setting is 7-day click, 1-day view and 1-day engagement. StackAdapt's 30 days is from its own blog, stated for DSPs in general. TikTok counts click-through, engaged view-through and view-through conversions with windows chosen per ad group, and Basis tracks both click-through and view-through by default with a lookback window set per tactic, a click beating an impression.

Two readings of the table. First, the long windows are where the view-throughs pile up: LinkedIn's 90-day view window is the longest default of the major platforms by a wide margin, and on a B2B account it credits a lead to an impression served a quarter earlier, while programmatic platforms commonly start at 30 days. Second, a Meta number and a Google Ads number are not the same kind of number even at equal windows, because Meta puts its view-throughs in the headline figure and Google keeps them in a column of their own.

There is no single standard definition, because the phrase names two different numbers. In video, view-through rate was Google's name for what it now calls the TrueView view rate: views divided by impressions, a measure of how many people watched the ad, with nothing to do with conversions. In conversion reporting it usually means the view-through conversion rate: view-through conversions per impression served.

A second ratio is more useful when you read a platform's claim: the view-through share, view-through conversions divided by all the conversions the platform reports for the campaign. The same campaign with 90 click-through conversions next to its 60 view-throughs has a view-through share of 40%, 60 of 150. A share that climbs while the leads in your CRM stay flat is the pattern worth investigating, and the click-through vs view-through post covers how to test it.

The honest case is simple: most people who see an ad do not click it, and some of them act on it later. A display banner, a LinkedIn post in the feed or a 20-second YouTube spot can put a name in somebody's head that they type into Google a week later. The click then goes to the search ad or the typed URL, and a report built only on clicks gives the impression no credit at all.

• Video and CTV have few clicks to count. A video ad is watched rather than clicked, and a television screen has no cursor. Without some form of view credit these channels report next to nothing, whatever they did.
• A click-only report favours the last channel. Brand search and retargeting sit closest to the conversion and collect the clicks, so they look best on any click report, including when an impression somewhere else started the journey.
• The bidding needs a signal. Where clicks are rare, a conversion after an impression is the only feedback a platform's algorithm gets, which is why Google lets YouTube-only Demand Gen campaigns opt into bidding on view-through conversions.

So the metric answers a real question. The trouble is that it answers with an association rather than a measurement: the platform knows the person saw the ad and later converted, not that they converted because of it.

• Retargeting reaches people already on their wayA retargeting audience is made of people who visited your site. Many of them would have come back anyway, and each one who converts inside the window after an impression is a view-through, whether the ad changed anything or not.
• Several platforms claim one conversionOne lead who saw a Meta ad, a LinkedIn ad and a programmatic banner in the same week is a view-through in three dashboards. Add the platforms together and they report more conversions than the business had.
• Windows outlast the decisionA 30-day or 90-day view window credits an impression served long before the person started looking. The longer the window, the larger the share of its conversions that were decided somewhere else.
• The bar for an impression is lowOn the Google Display Network, Active View counts an ad as viewed when half of it is on screen for one continuous second. A scroll past is enough to start the window.

None of this is fraud, and the numbers are not invented: the person did see the ad and did convert. What inflates is the story the number implies, that the ad caused the conversion, and the sum across platforms, which counts one conversion many times. Last-click attribution has the opposite bias, and neither is the answer on its own.

A view-through conversion exists only inside the platform that served the impression. The impression happened on Meta's, Google's or a DSP's inventory, the platform matched it to your conversion through its own pixel or conversion API, and nothing about it ever reached your website. There was no click, so there was no visit carrying a click ID or UTM parameters for any other system to read.

Your CRM records a lead with the source of the visit that produced it: the UTM parameters, the click ID, the referrer, or a field somebody typed. A lead who saw a LinkedIn ad and then searched for your brand arrives as search, because that is the visit that happened. The impression is not wrong. It is invisible to every system except LinkedIn.

GA4 is almost the same. Its attribution models credit clicks, and its attribution settings are a reporting model, the channels that can receive credit and a key event lookback window, with no view-through window among them. The exception is YouTube: with Google Ads linked, GA4 records engaged-view key events, a YouTube ad watched for at least 10 seconds followed by a key event within 3 days, under google / cpc. A display banner, a Meta ad or a programmatic impression that was only seen does not appear.

That is why a click-based report of your CRM is always lower than the sum of the platforms, and why the gap is information rather than an error. Conversion tracking covers what each system records, and the customer journey is what a click-based record can follow.

A view-through conversion is an ad platform's record that somebody saw an ad, did not click, and converted inside a window the platform set by default. It is a real signal for video, CTV and display, where clicks are rare, and an inflated one wherever retargeting, long windows and several platforms claim the same person. Read it as exposure, keep it in its own column, and judge it against what your CRM recorded. For the decision itself, which number to use for bidding, budget and reporting and how to test whether the view-throughs are real, read click-through vs view-through conversions.