Video action campaigns are a simple and cost-effective way to drive more conversions on and off YouTube, all in a single campaign, with the help of Google AI. These campaigns use Google's AI smart bidding strategies to drive conversions. In some cases, your campaigns might not be delivering the expected volume of conversions.
This article describes a few ways in which you can increase conversions in campaigns that aren’t generating the expected number of conversions.
Before you begin
- If you wish to learn more about how Smart Bidding works with YouTube, review the YouTube Smart Bidding Guide
1. Follow all the best practices for video action campaigns
- Conversion tracking: Use Google Ads web conversion tracking with gtag.js or Google Tag Manager for the accurate tracking and optimisation of your desired conversion event. In addition, track lighter actions to achieve your desired conversions faster. For example, use 'add to cart' to provide more signals to Google AI.
- Budget: Set a budget that is at least 15 times your tCPA for your tCPA campaigns and 10 times the expected CPA of your maximise conversions campaigns. If you aren’t sure of the expected CPA, set a budget that’s 2–3 times your Search CPA, or 1–2 times your Display CPA.
- Bidding: If you know your video CPA, start with tCPA bidding. If you don’t know your video CPA, use the UI CPA guidance to determine tCPA for video. If your primary goal is to drive as many conversions as possible through a flighted campaign, start with Maximise conversions bidding.
- Segment: Start with either 'Custom segments – search terms', your data, Customer Match or similar segments and opt in to 'optimised targeting' to show your ads to more people who are likely to convert. For custom segments, use 10–15 of the highest converting search keywords as search terms.
- Creative: Add sitelinks to your campaign or ad group, or run multiple creative variants (at least five if possible by changing the Call To Action (CTA) button, headline text or core video asset).
2. Understand the video conversion funnel and ensure that you’re considering engagement metrics
Engagement rate: engagements/impressions
View rate: views/impressions
CTR = clicks/impressions
Conversion rate (for video): click-through conversions/views
Engaged-view conversions (EVC): This is not a metric in Google Ads. To see reporting, click the segments icon 
Engaged-view conversion rate: Engaged-view conversions/engagements. This is not reported in Google Ads. You’ll have to calculate this manually.
View-through conversion rate: view-through conversions/impressions. This can be added as a custom column in reporting.
In video campaigns, the user has to be counted as a view for a conversion to be counted. This doesn’t require the user to click the ad to your site. If the user clicks the ad and navigates to your website, a click will be reported under the 'Clicks' column. The user will still be counted as a conversion if, for example, they saw an ad and then navigated to the site organically and converted.
It’s best to think of performance as a funnel to see what step contributes to low performance. You can add the above columns in this order to simulate this funnel visual and see where the biggest drop-offs happen.
In your performance report, click the columns icon 
- Impressions
- Engagements
- Views
- Clicks
- Conversion/view-through conversion
Next, identify where within these steps the drop-off is the highest and the ways you can improve:
View rate (formerly known as view-through rate or VTR): You can improve the view rate by improving the creative quality, optimising the video length, and more.
Click-through rate (CTR): You can improve your CTR by adapting the CTA text and creatives.
Conversion rate (CVR): CVR varies significantly depending on the events that are tracked as conversions (such as reaching a page, submitting a lead form, making an expensive purchase). The best way to judge your CVR is to compare it to other campaigns (including search, display) with the same goal.
Low CVR can also be caused by poor landing page quality, price and product availability.
Bounce rate (available with Google Analytics): High bounce rate can be caused by poor landing page quality. Having a high bounce rate can affect both CTR and CVR.
To see a breakdown of CTC compared to EVC, segment reports by selecting 'Conversions', then 'Ad event type'.
3. Set proper expectations for the 'Learning' phase and account for conversion delay
If you've just set your campaign live or made major changes to your bids, creatives, targeting, etc., wait for 7–14 days before you assess the campaign's performance or make more optimisations. You can determine the status of the campaign from the bid strategy status. If the campaign is in the learning phase, the bid strategy status will be 'Learning'. Learn more about bid strategy statuses
Why does it need this time to learn?
Bid strategies use Google AI to adjust your bids using signals that indicate whether a user is likely to convert. After a campaign is changed, it takes some time for Google Ads to optimise the bids for the best outcome.
This is why it’s important to resist the urge to make changes within the learning period. Any changes made mean the bid strategies have to be re-trained or re-calibrated to adhere to the new settings applied to the campaign.
How long is the 'Learning' phase?
It varies based on the volume of conversions that your campaign receives. If you receive a very high number of conversions (>100 conversions/day), the learning phase will take less time (1–3 days). If you receive a very low volume of conversions (<5 conversions/day), the learning phase will take 14 days. The bid strategy will keep optimising to deliver your campaign's goal, even after you exit the learning phase at the end of 14 days. Your campaigns will be less prone to volatility in reaction to changes after you’ve exited the learning phase.
Conversion delay
When looking at your campaign performance, ensure that you account for conversion delay in your reporting. After determining your 'days to conversion', you should remember not to pull reports that fall within that period. For example, if your 'days to conversion' is 3 days, it might not be a fair representation of your performance to look at reporting that includes the last 3 days. If you do this, you’ll be including a portion of spend where the retro-dated conversion reporting is not yet appearing.
4. Ensure that you’re capturing all conversions
Even when your campaign delivers conversions, it doesn’t mean that you’re able to capture all or most of the conversions.
Here are some scenarios where you would be missing out on conversions:
- Universal Analytics can measure click-through conversions but not EVCs, which represent an average of 50% of all conversions.
- Offline conversion import based on GCLID or wbraid or gbraid tracking parameters doesn’t work for EVCs. Support for wbraid or gbraid for iOS is very recent and requires you to update the back-end code.
- Cookie consent banners often found in the EU limit our ability to track conversions, especially if they're not properly implemented with consent mode.
- Some landing pages might drop GCLIDs or not have a tag to collect it.
- UA and GA4 don’t use the 'ads greedy' attribution logic. This contributes to an average loss of 40% of conversions.
- Last-click attribution models can disadvantage video compared to search/Shopping.
5. Consider creating 'shallower' conversions for the bid strategy to optimise towards
If you expect your campaign to generate less than 10 conversions per day, it's recommended to set it up to optimise for a more frequent event. For example, a campaign for a small e-commerce site might generate fewer than 10 purchases a day but 50 'Add to basket' events. It might be better to create a secondary conversion action to track the 'Add to basket' event and use campaign-level conversion goals to configure the campaign to optimise towards that goal.
6. Don’t let other performance metrics distract you
Google AI bid strategies ultimately optimise towards the goal metric of your campaign. The table below briefly describes how they work.
| Max conversions | Target CPA | Max conversion value | Target ROAS (return on ad spend) |
|
| Goal | More conversions for your budget | More conversions within tCPA | More conversion value for your budget | More conversions within tROAS |
| How | Less focus on avg. CPA or tCPA goals |
Focus on avg. CPA with strong tCPA goals | Doesn’t take value relative to ad spend into account | Balances spend and how much users purchase |
| Focus metrics | Conversions | Cost / conv. (CPA) | Conv. value | Conv. value/cost (ROAS) |
When you study the performance metrics of your campaign, don’t focus on the metrics that aren’t central to the goal of your campaign. For example, you may notice that the CPV of your campaign is low, but this is to be expected if the bid strategy isn’t focused on CPV as a success metric.
7. Keep an open mind about retesting Smart Bidding strategies
We’re constantly working to improve Smart Bidding’s ability to predict conversions. If it’s been a while since you’ve reviewed your bidding strategies, we encourage you to explore them again.
