In order to keep your campaigns relevant and optimize performance, it’s important to regularly introduce new creative assets to your campaigns.  When either new campaigns or new items to existing campaigns are added, Taboola's algorithm uses machine learning to test different bid pricing to explore the most efficient way to deliver your campaign.  Once the learning phase is complete, you can begin to optimize and scale your new campaign. 

The faster the learning process can be completed, the more frequently you can add new items and the faster you can scale new campaigns.  Creative Expedition, a feature of SmartBid, enables an Expedited Learning process that accelerates exploration of new creative assets. In many cases, the learning and exploration process will be completed in under 24 hours.

This allows you to keep campaigns fresh and relevant while substantially decreasing the duration of the learning process per creative asset.

How Does Creative Expedition Work?

Creative Expedition takes advantage of Taboola’s publisher placements while balancing the  eligibility criteria specific to your campaign. New items that qualify for Creative Expedition will automatically be placed in “Expedited Learning” instead of the standard “Learning” process.  

When an item is in Expedited Learning, your Campaign Bid may be increased in order to win the right number of impressions and clicks to accurately estimate CTR, but SmartBid will continue to manage bids to ensure they are the lowest required to secure the impression during the learning process. 

Once the right number of impressions and clicks have been achieved across the target publisher websites, the exploration process will end and the Expedited Learning phase is complete.  At this time, your campaign will revert back to your original campaign settings automatically.

What Is the Eligibility Criteria for Creative Expedition?

Your Smart Bid campaign will undergo Creative Expedition under the following criteria:

• You have a remaining campaign budget of at least $200

• If your Daily Cap allows for 200 clicks.

• It has been less than 10 days since your Ad was approved.

• Your campaign is NOT hyper targeted using postal codes/ regions / publishers / audiences

** New creative assets are eligible to undergo expedition for up to 10 days. It’s important to note only one item will be expedited at a time. 

Tracking Items Through the Creative Expedition Process

Progress status for all Learning Phases is available in Backstage in the Top Campaign Content view. 

Campaign Learning Phases

• Expedited Learning:  Creative Expedition items only.  Taboola’s algorithm has begun an accelerated process of collecting data to explore how to optimize the performance of your campaign.
• Learning: For items or campaigns that don’t qualify for Creative Expedition, Taboola’s algorithm has begun the standard process of collecting data to explore how to best optimize the performance of your campaign.
• Learning Completed: Taboola's algorithm has completed its learning process and your ads are now scaling across Taboola’s network.
• Learning Limited: Your ad has stopped learning due to insufficient optimization events (this can occur on the item level or at the campaign level). Adjustments need to be made to your targeting, budget or bid amount.

During either the Expedited Learning or the Learning phase, it is normal to see CPA fluctuations while your item undergoes the learning process. Once the learning process is complete, your campaign will automatically revert back to your original campaign settings.

An item can only undergo Expedited Learning once, and Expedited Learning will only be applied to one item at a time. 

Creative Expedition Best Practices

In order to maximize the benefits of Creative Expedition and the Expedited Learning process, we recommend the following best practices:

• Wait at least 1 day after notification of Learning Completed before evaluating the performance of your item.  Performance during the learning process is incomplete and should not be used to assess future performance.
• Exclude results from the “Expedited Learning” phase when evaluating overall performance