We did this primarily to understand the optimization metrics available. We launched several videos and analyzed them based on cost per impression, cost per view, cost per click, and click-through rate. With this knowledge in hand, our team started automating campaigns using performance metrics such as cost per acquisition (CPA) and conversion rate.
The next step was to develop a conversion tracking and attribution system. Whereas we previously attributed most ad conversions based on click alone, we now added view-based attribution for YouTube ads using Google Campaign Manager.
With this change, the demonstrable value of YouTube as gcash database an advertising channel, particularly at the customer acquisition stage, became much clearer. And these placements took pride of place among our other paid channels.
We found that custom audiences worked better for promoting a specific product than for a generic brand. In cases with frequently fluctuating budgets and long conversion funnels, a Maximize Conversion bidding strategy proved to be more effective than a Target CPA strategy. And we settled on the skippable in-stream ad as the most effective ad format for your product.
►Makes your creativity simple and clear
Once the preliminary results were in, we set about optimizing our ad creatives and landing pages.
Olga Denisova, our VP of Digital Marketing, observed that a better conversion rate is achieved at each stage by having videos of a certain length, type of content, degree of complexity and format, i.e. animated or with human actors.
For example, in the upper stages of the funnel, we used short videos of up to 15 seconds that introduced the product and encouraged trial. For lower-funnel marketing, we featured longer videos, up to a minute long, with more details about a specific product feature or customer success story.
Based on this insight, Denisova and her team began optimizing our video creatives according to the audience’s purchase consideration phase.
Our previous video creatives had basic, text-heavy graphics but lacked voiceover and pizzazz.
►Before
We corrected this in newer videos, focusing on more eye-catching visuals with human actors. The CTA was also modified for different stages of the funnel and the video length was optimized.
After
►Scaling using analytics
Once we had our optimal YouTube ad settings figured out, we set out to expand the Semrush channel.
Based on past experience, we propose an optimal frequency of seven impressions per day and 10 views per month.
For remarketing campaigns, we base our learning strategy on the size of the target audience. In the upper remarketing funnel, with your large available audience of website visitors and registered users, we implement Google's Maximize Conversions bidding strategy.