A view-through conversion happens when someone saw the ad, did not click, and later converted. It may reflect a real effect, especially in video and display, but it can also credit ads with what would have happened anyway. It should be treated with caution.
The essentials in a definition
View-through conversions are conversions attributed to someone who saw an ad and converted later without clicking. They can reflect real influence, especially in video and display, but also inflate results with people who would have bought anyway. The conversion measurement documentation explains the options; it pays to treat them as a secondary signal, validated by tests.
What it is and why it exists
The ad influences without being clicked: the person sees it, remembers, searches for the brand and buys later. Platforms try to capture this effect by counting conversions that occur within a window after the view.
In reach channels such as video and display, ignoring this effect would underestimate the channel’s value. But counting everything blindly overstates it.
Where the risk lies
Someone who was going to buy anyway may see an ad by chance. The platform counts the view as influence even if it did not change the decision. In remarketing the risk is higher: the audience has already visited the site and is closer to purchase.
So view-through conversion numbers should not be added to click-through ones without analysis.
Treating views with caution
- Separate click and view conversions
- Use short view windows
- Validate the effect with holdout tests
- Cross-check with indirect brand signals
How to handle it in reports
Separate click and view conversions. Use short view windows. Weigh views carefully in budget decisions. Compare channels using the same criterion.
If possible, validate the effect with holdout tests: one group does not see the ads and is compared with the exposed group.
Seeing an ad and buying later does not prove the ad made the sale.
When to trust more
In discovery channels, with new audiences and no remarketing, view-through conversion is more likely to reflect a real effect. With already warm audiences, less so.
In any case, complement with indirect signals: brand searches, direct traffic, surveys and total sales by region.
Decision table
| Conversion type | Measures | Caution |
|---|---|---|
| Click-through | Purchases after clicking the ad | More direct, favours the last contact |
| View-through | Purchases after seeing, without clicking | May include those who would buy anyway |
| Assisted | Participation in the journey | Not causality |
| Incremental (test) | Estimated causal effect | Needs design and volume |
Common mistakes to avoid
- Adding everything into the same ROAS. Separate click and view to understand each one’s effect.
- View window too long. Attributes sales far from the ad.
- Cutting channels for few attributed views. Some channels have a diffuse effect that only tests reveal.
How to measure the result
Report click and view in separate columns, track the proportion of each over time and validate the effect with tests before relevant budget decisions.
Limits and exceptions
Holdout tests to validate views have costs (potentially lost sales) and limits (volume and comparability). In small-budget channels, it may not pay to run them. In those cases, use conservative criteria: give less weight to view-through conversions, prefer short windows and track the effect on brand searches and total sales.
Illustrative example
Illustrative example: a brand sees many view-through and few click-through conversions on video. Instead of accepting the number, it runs a four-week regional holdout of video. Sales in the region without video fall slightly, less than attribution suggested. The brand keeps video with a smaller budget and reports click and view conversions separately. Think with Google studies help think through video’s role.
Frequently asked questions
Are view-through conversions fake?
No, but they can overstate the effect. Use them with caution.
Should I add them to click conversions?
It is better to analyse them separately.
When do they count more?
In discovery channels and with new audiences.
Should I ignore view-through conversions?
No, but treat them as a secondary signal and validate the effect with tests before giving them the same weight as click-through.
What view window should I use?
The shortest that makes sense for your cycle; long windows raise the risk of attributing sales that would happen anyway.
View-through conversions suggest influence, not causality: report them apart from click-through and validate the real effect with tests before reallocating budget.
Sources and recommended reading
Last editorial review: October 2026. Informational content; always confirm current conditions in the official sources before deciding.