Demand Gen

Sample data

November 20, 2026

•
Sample data

{{minute}}

min Read

View-Through Conversions: How to Measure B2B Video Without Fooling Yourself

Waqas Khokhar

Founder at ScalixAI

On this page

What Is Centralized Notification Infrastructure?

The Default State: The Big Notification Mess

Building Workflows Without Engineering Dependency

Easy Debugging with Logs 

Embedded & Branded In-App Inbox

Multi-Tenant Aware Preference Center

Want to know what your YouTube spend actually adds?

Book a Video Measurement Audit

Key takeaways

1.

A view-through conversion shows exposure came first, not that the ad caused it.

2.

Engaged-view conversions need watch time, so they deserve more weight than VTCs.

3.

Upload CRM conversions within 14 days, or Google drops the engaged-view credit.

4.

GA4 cannot show impression-based VTCs, so Google Ads and GA4 will disagree.

5.

Only holdouts or lift studies tell you what YouTube actually added.

A view-through conversion (VTC) is a conversion that happens after someone sees your video ad, does not click or engage with it, and converts later within a set window. In B2B, VTCs prove exposure came before the conversion. They do not prove the ad caused it. Measuring B2B video means separating the two, then checking whether those conversions turn into pipeline in your CRM.

ScalixAI sees the same issues in audit calls every month. YouTube shows 150 view-through conversions. Search shows 200 conversions. GA4 shows a third number. The CRM shows 12 qualified opportunities. The founder asks which number is true, and the honest answer is that each one measures something different.

This guide explains what each number measures, which settings inflate or hide video's contribution, and how to test whether YouTube is adding pipeline or taking credit for it.

What is the difference between view-through, engaged-view, and click-through conversions?

Google Ads counts video conversions in three ways. A click-through conversion follows a click. An engaged-view conversion (EVC) follows a meaningful watch with no click. A view-through conversion follows an impression with no click and no meaningful watch. They carry very different evidence, so its important to read them separately.

Google's engaged-view conversion rules set the bar for "meaningful" by format:

[table]
Format | What counts as an engaged view
Skippable in-stream | 10 seconds watched, or the whole ad if shorter
In-feed video | 5 seconds watched
Shorts | 5 seconds watched, a CTA click, or a like, comment, or share
[/table]

That difference matters in B2B. An EVC means a person chose to watch your founder explain the problem for 10 seconds. A VTC means your ad loaded somewhere on their screen. Neither proves cause, but an EVC is a much stronger signal of attention.

Format also changes the numbers. Shorts count likes and shares as engagement, so a Shorts-heavy campaign can show more EVCs than an in-stream campaign with the same buyers. Make sure you compare EVC volume within a format, not across formats.

Which Google Ads column shows view-through conversions?

VTCs sit in the "View-through conversions" and "All conversions" columns. They stay out of the main "Conversions" column, according to Google's VTC documentation. The exception is Demand Gen campaigns opted into VTC optimization. Their biddable VTCs land in "Conversions".

EVCs work differently. They appear in "Conversions" and "All conversions" by default. So the "Conversions" number on a YouTube campaign already blends clicks and engaged views.

Three things founders often miss in that column:

  • Demand Gen VTC optimization is off by default. It works only on YouTube-only Demand Gen campaigns. If someone turned it on, Smart Bidding now chases impression-based conversions.
  • Privacy limits undercount VTCs. Google says VTCs from browsers that block cross-site cookies cannot be reported. Safari-heavy audiences will show fewer.
  • Google removes overlap. A person who also clicked one of your other ads is not counted as a VTC.

To see the split, segment the campaign view by "Ad event type". That shows how much of a YouTube campaign's conversions came from clicks, engaged views, and impressions.

What conversion window should B2B use for YouTube?

Start from your real-time lag, not your sales cycle length. Google's defaults are a 30-day click window, a 3-day engaged-view window and a 1-day view-through window. A 30-day VTC window does not mean a demo booked 29 days after a skipped ad deserves credit.

Google's conversion window rules allow 1 to 30 days for engaged views and view-throughs. Changes apply only going forward, so a longer window will not recover past conversions.

How ScalixAI picks windows:

  1. Pull the time lag first. Open Attribution, then Path metrics, and check days to conversion for the conversion action you bid on.
  2. Set the click window to cover the journey to that action. Fyxer's early tests used a 30-day click window while trial-to-paid took about eight weeks. The window ended before most paying users converted. The rebuild moved it to 60 days (Fyxer Demand Gen case study).
  3. Keep EVC windows short unless the data says otherwise. Someone who watched 10 seconds and booked a demo a few days later is a reasonable claim. Three weeks later is weak.
  4. Keep VTC windows tight for reporting. Every extra day adds people who would have converted anyway.

A short window undercounts video. A long window flatters it. Neither error is fixed by picking a bigger number. Fix it by testing incrementality; something we cover below.

Why do Google Ads, GA4 and the CRM show different video conversions?

They measure different things, so they should disagree. Google's attribution report documentation says conversions credited to impressions are not supported in Google Analytics. Engaged-view conversions are supported in both. So a GA4 report will never match a Google Ads report that includes VTCs, even when both are set up correctly.

Here is how ScalixAI reconciles the four numbers from the opening example (hypothetical figures):

[table]
Source | Number | What it actually counts
Google Ads, YouTube "View-through conv." | 150 | Form fills after an impression, no click, in the window
Google Ads, Search "Conversions" | 200 | Form fills after a Search click. Some are the same people
GA4 key events | Different again | Sessions it can attribute. No VTCs, its own lookback
CRM opportunities | 12 | Leads sales accepted. The only number tied to revenue
[/table]

The steps:

  1. Match conversion actions and dates. Compare one action, such as "Demo booked", over one date range. Use the "by conv. time" columns when comparing to the CRM, which logs the conversion date, not the ad interaction date.
  2. Check for double counting. If a HubSpot form action and an imported GA4 key event both track the same demo as primary, every lead counts twice.
  3. Accept the overlap. The 150 VTCs and 200 Search conversions are not 350 leads. A buyer can see a YouTube ad, then search your brand and click.
  4. Trace the 12 back. Pull the opportunities from the CRM and check which ones had a YouTube click, engaged view, or impression before the first conversion.

The CRM number is the one the business runs on. The others help you understand how buyers actually got there.

How do you see YouTube's role in the buyer journey?

Turn on YouTube impressions in the attribution reports. Go to Goals, then Attribution, then Conversion paths, and switch the YouTube impressions toggle on. Paths will then show video impressions from YouTube and Demand Gen campaigns next to Search clicks.

A typical B2B SaaS path looks like this:

  1. YouTube impression: a founder-led video about the problem
  2. Brand search a few days later, organic click
  3. Category search, paid click
  4. Pricing page visit
  5. Demo booked

Last click gives Search all the credit. The path shows YouTube came first. The question to answer is whether Search would have won that buyer without the video.

Fyxer's conversion paths showed Demand Gen in 70 of 96 paths analysed. That is strong evidence YouTube touched most buyers. On its own, it is not proof YouTube changed their decision, so ScalixAI paired it with lift data.

Is YouTube creating demand or taking credit for it?

Attribution tells you which touchpoints got credit. Incrementality tells you whether conversions would have happened without the ads. VTCs answer the first question. Only an experiment answers the second.

YouTube can take credit it did not earn. If you already get 500 branded search conversions a month and launch YouTube to your site visitors, many of those people would have converted anyway. They just saw an ad first.

Here are three ways to test it, from strongest to weakest:

1. Conversion Lift studies. Google splits people into an exposed group and a held-out control group, then reports incremental conversions, relative lift, and incremental CPA (spend divided by incremental conversions). Google says Conversion Lift is not available in every account, so ask your Google rep. Geo-based studies can include offline data.

Fyxer qualified. Google's studies on the account showed a 359% rise in daily average conversions in the incrementality study (September to November 2025). A 28-day Search Lift study, measuring £60K of spend, showed searches for "fyxer" up 171%.

2. Geo holdouts. If you can't get Conversion Lift, hold YouTube back in matched regions and compare qualified demos, pipeline, and brand searches. Pick regions with similar past demand, and don't launch other campaigns in only one group.

3. Brand search trends. Watch brand impressions in Google Ads and brand clicks in Search Console before, during, and after a flight. Fyxer's monthly brand impressions rose from a 62K average to 169K in November 2025. After Demand Gen stopped, they fell to 51K by June 2026. The rise and the fall line up with the spend, making our case stronger. It is still a correlation, so treat it as support, not proof. Our post on how offline campaigns lift branded search covers this method.

How do you connect video exposure to pipeline and revenue?

Import CRM stages back into Google Ads so a video conversion can be judged by what happened to the lead. A form fill is not an acquisition. Send back SQLs, opportunities, and closed-won deals, then compare how EVC and click-sourced leads progress.

The setup is covered in our offline conversion tracking guide and our HubSpot and Google Ads integration guide. Three video-specific rules sit on top of it:

  • The 14-day rule. For EVCs, Google requires offline events to be uploaded within 14 days of the conversion time. Upload later, and the engaged-view credit is not recorded, even inside a 30-day window. A weekly manual CSV upload loses credit for any qualified lead that took more than a week to reach.
  • Enhanced conversions for leads. Google says it supports engaged-view and cross-device conversions that GCLID-only imports miss. A buyer who watches on a phone and books from a work laptop has no click ID to match. Hashed email can bridge that.
  • Use the current pipe. From June 15, 2026, the Google Ads API stopped accepting new offline conversion import setups. The Data Manager API is now the main route. Older integrations keep working during migration, but anything new should be built there.

Delve, a compliance automation platform, had CRM-integrated tracking and SQL imports from day one. Campaign-level pipeline attribution was visible within 30 days (Delve case study). Build this before the video budget goes live, not after the first bad report.

What B2B measurement problems make video look worse or better than it is?

B2B buying journeys break the assumptions behind platform reporting. Five problems come up in most ScalixAI audits:

  1. Immature cohorts. A video flight from last month has not had time to close deals. Compare cohorts at the same age, such as 90 days after first exposure, not this month against last year.
  2. Many buyers, one account. The engineer watches the ad, the manager books the demo, the CFO signs. Google sees three people. Your CRM sees one deal.
  3. Low volume. Ten opportunities a month cannot support a strong conclusion either way. Look at trends over quarters.
  4. Overlap with LinkedIn and Search. The same buyer may see your LinkedIn ad, your YouTube ad, and click a Search ad. Every platform claims them.
  5. Dashboards that undercount. Oneleet's LinkedIn dashboard reported a 0.36x return. Multi-touch attribution showed $1.63M in influenced closed-won revenue across 150 deals (Oneleet case study). Video has the same gap, which is why our LinkedIn attribution breakdown applies here too.

Fyxer also shows the gap between modelled and booked value. About £674K of predicted lifetime value is attributed to its two Demand Gen campaigns. That is a model output, not booked revenue, and it should be reported that way.

The ScalixAI YouTube conversion measurement checklist

ScalixAI runs this checklist before judging any B2B YouTube or Demand Gen campaign. Each line is a setting or report you can open today.

Conversion actions

unticked
  • The primary conversion is a qualified event (demo held, SQL, activated workspace), not a page view or newsletter signup.
unticked
  • Lead conversion actions count "One" per click or interaction, not "Every".
unticked
  • No business outcome is tracked as primary by two actions (for example, HubSpot form and imported GA4 key event).
unticked
  • Demand Gen VTC optimization is off unless you chose it, and you know whether VTCs are in "Conversions".

Windows

unticked
  • Click window covers the journey seen in the Path metrics days-to-conversion data.
unticked
  • EVC window was set on purpose, not left at 3 days by accident or stretched to 30 by habit.
unticked
  • VTC window is 1 to 3 days for reporting unless a lift test supports more.
unticked
  • Window changes are logged with dates, since they don't apply backwards.

Reporting

unticked
  • YouTube campaigns are segmented by "Ad event type" in every review.
unticked
  • Columns include Conversions, View-through conv., All conv. and Conversions (by conv. time).
unticked
  • The YouTube impressions toggle is on in Attribution, Conversion paths.
unticked
  • Google Ads and CRM are compared on the same action and the same conversion dates.
unticked
  • Nobody adds VTCs to click conversions and calls it total leads.

CRM and data pipe

unticked
  • GCLID, GBRAID, and WBRAID are stored on the contact record, not only the first session.
unticked
  • Enhanced conversions are on, with hashed email sent on the form.
unticked
  • CRM stages upload daily, so EVC credit beats the 14-day limit.
unticked
  • New upload integrations use Data Manager, not a legacy Google Ads API path.
unticked
  • Every lead source can be traced to an opportunity ID in the CRM.

Validation

unticked
  • SQL rate is compared for click, EVC, and VTC-sourced leads.
unticked
  • Brand impressions and Search Console brand clicks have a pre-launch baseline.
unticked
  • Your Google rep has been asked about Conversion Lift eligibility.
unticked
  • If Lift is unavailable, a geo holdout is planned with matched regions.
unticked
  • Cohorts are judged at the same age, never this month against last quarter.

Want to know what your YouTube spend actually adds?

ScalixAI will run this checklist on your account. We check conversion actions, windows, column setup, CRM uploads, and attribution paths, then show how many of your video conversions became pipeline and what a lift test would need. It's the same review behind the Fyxer and Delve programs.

Book a Video Measurement Audit

Still deciding on YouTube? Read whether YouTube ads are worth it for B2B SaaS or our Demand Gen and YouTube playbook.

Frequently asked
‍questions

What is a good view-through conversion rate for B2B?

There isn't a reliable benchmark. VTC rates depend on window length, audience warmth, and how many people already know your brand. Compare your own VTC-sourced SQL rate with click-sourced leads instead.

Are view-through conversions real?

They are real events, but not proof of impact. A VTC means someone saw your ad and later converted. They might have converted anyway. Lift studies and holdouts show how many were caused by the ad.

Should you count view-through conversions in ROAS?

Not in your main ROAS. Report VTCs separately. Base budget decisions on click conversions, EVCs that turn into pipeline, and incremental results from lift tests.

What is the default view-through conversion window in Google Ads?

One day. The engaged-view default is 3 days, and the click-through default is 30 days. Google allows 1 to 30 days for view-through and engaged-view windows.

Why doesn't GA4 show my YouTube view-through conversions?

GA4 does not support conversions attributed to impressions. It does support engaged-view conversions. So Google Ads will always show more video conversions than GA4 when VTCs are included.

What is the difference between view-through and engaged-view conversions?

A VTC needs only an impression. An EVC needs real engagement, such as 10 seconds of a skippable in-stream ad or 5 seconds of an in-feed or Shorts ad.

How long should you wait before judging YouTube for B2B?

Long enough for leads to reach opportunity. For most B2B SaaS, that is at least one quarter. Oneleet's closes came months after first exposure on LinkedIn.
Book a Call