top of page
Meta • 2023
Allow advertisers to generate on-brand dynamic video ads at scale
With a big shift towards short form video in the industry, Meta pushed to unlock scalable, on-brand solutions for dynamic video ads.
Project Overview
Dynamic Ads is an important value driver for retail, e-commerce and CPG advertisers, delivering users the most relevant products from a large catalog. Meta lacks video solutions for dynamic ads, and given the previous success of dynamic image ads, there's a focus to meet advertisers needs bringing this to video. This case study outlines the design process and strategy for allowing advertisers to easily generate on-brand variations of catalogue video ads at scale.
My contributions
I led the full strategy and design for creating scalable dynamic video ad templates that adhered to advertisers brand constraints. This involved a lot of collaboration with designers across Meta Ads Manager, Instagram, and Facebook, along with research, product, tech, and leadership across Monetization.
Existing state
Dynamic ads make up over 10% of Meta Ads revenue. Although, there's a lack of video solutions, there's still a small percentage of revenue coming from suboptimal video ads.
Existing suboptimal solutions
Problem space
Scalability
Advertisers have limited creative resources to generate video ads creatives at scale.
Customization
Advertisers current creatives are too plain, with the majority of what most advertisers have in their catalog are product shots on plain background, and advertisers need to ensure the style is on-brand.
Performance
There's no way to efficiently promote their product in a large product catalog to drive DR performance through video ads.
How do we allow advertisers to guide our automation so they can benefit from scale while adhering to their needs for creative control?
Hypothesis
By introducing a new in-house solution that allows advertisers to easily generate on-brand variations of catalog video ads at scale, we can help advertisers scale video ads for large catalogs dynamically with low design resource cost while adhering to their brand constraints and drive direct-response performance.
Solution Concept 1
Import designed template with plug-in in design tool (e.g. After Effects) to Meta Ads Manager​
Designers are still creating their video ads, having control over brand constraints. Once creatives are imported to Meta as a template, it's paired with their catalog(s) and can specify replaceable layers in the template to replace with information from the catalog(s).
Key benefits
-
Saves time by scaling up video ad creation for large product catalogs.
-
Complies with advertisers brand guidelines.
-
Drives DR objectives with personalized and diverse set of creatives.
Risks
-
Requires creative team to create templates.
-
Creatives won't be designed with insights on how to drive performance across Meta platforms.
Solution Concept 2
Import created template with API from partner tools (e.g. Smartly) to Meta Ads Manager​
Designers create or edit templates through editor on third-party online platforms where their catalog(s) are uploaded and synced with Meta. Templates are then paired templates with catalogs, setup, and launch campaigns with the paired templates on the third-party platform.
Key benefits
-
Continue to use the third-party platform you normally use to create ad creatives and set-up and launch multi-channel campaigns.
-
Save time by scaling up video creation for large catalogs.
-
Comply with brand guidelines.
-
Drive DR objectives with personalized and diverse set of creatives.
Risks
-
Requires some creative skill to create or edit templates.
-
Creatives won't be designed with insights on how to drive performance across Meta platforms.
Solution Concept 3
Design a library of pre-made templates in Meta Ads Manager for advertisers to use​
Advertisers upload their brand constraints and have templates surfaced based on their goals. It's automatically paired with their catalog(s), layers in the template are replaced with information from the catalog(s), and styles reflect the brand elements.
Key benefits​
-
Save time by scaling up video creation for large catalogs.
-
Comply with brand guidelines.
-
Drive DR objectives with personalized and diverse set of creatives.
-
Creatives designed based on insights to drive performance across Meta platforms.
Risks
-
Requires additional work to get brand constraints integrated in Meta Ads Manager.
-
Need support to continue creating templates in house.
Solution Concept 3
Following inbound from past research, we decided to focus our efforts on Solution Concept 3, with the hypothesis: Advertisers would adopt a dynamic video ad solution if we help them get started on creation through video templates they can choose from and customize.
In order to gain adoption, three levers needed to be met; performance, control and scale. By validating this, the team focused on proving the value through a series of Alpha and Beta tests.
Alpha
Focused on proving the value of a template layout for single image ads across 3 key placements; IG Stories, IG Reels, and FB Feed, utilising past learnings to improve template performance while incorporating advertisers brand constraints.
Beta
Focused on scaling a variety of template layouts for multi-product, multi-image, and single image ads across 6 placements; IG Stories, IG Reels, IG Feed, FB Stories, FB Reels, and FB Feed, utilising past learnings to improve template performance, and incorporate advertisers brand constraints.
For the Alpha, we incorporated levers that we had high confidence in and could be used across templates, while introducing an additional key lever; key selling points (e.g. price, discount, product name or description, delivery information) that could vary for the advertiser and template, creating more creative variations. Through research and previous tests, we learned that adding key information about advertisers products in the creative can help end users make more informed purchasing decisions, leading to a conversion lift.
With a successful Alpha, and moving onto a closed Beta, we were able to incorporate additional template design levers; multiple products, multiple images, lifestyle images and video, and AI enhancements.
Multiple products
By showing more of the products upfront it will increase engagement and thereby, conversion. By showing minimal information with products will increase digestibility.
Advertisers need​
-
Advertisers want to have the ability to promote multi-product in video
Insights​
-
Past dynamic ads on IG Stories had a similar animation; vertical scroll, that outperformed BAU with an increase in ad score and clicks
-
For users, more products could mean both more choices and more distractions, so having minimal information is more digestible
Multiple images
By showing multiple images for one product it can provide consumers with visual clarity, boosting consideration signals. By showing minimal information with products will increase digestibility.
Advertisers need​
-
Advertisers would like animations to showcase multiple images of the same product (e.g. different angles, colors)
Insights​
-
Consumer research suggested that providing visual clarity by showing multiple angels improves ad engagement
-
Past dynamic ads on IG Stories had a similar animation; vertical scroll, that outperformed BAU with an increase in ad score and clicks
-
For users, more products could mean both more choices and more distractions, so having minimal information is more digestible
AI enhancements
By leveraging AI, like background generation, we can help advertisers scale their product images to have more creative variations including lifestyle and outcrop, making ads feel more native, increasing the creative quality, and reducing ad fatigue.
Advertisers need​​
-
Advertisers need more creative assets, specifically lifestyle, as they currently find it resource intensive to create variations of their assets​​
Insights​​
-
75% of BAU advertisers only have 1 asset live per campaign
-
The more native an ad looks the are more engaging it is and leads to higher performance
Dynamic video ads was launched as an Alpha with 10 recruited advertisers. There were performance gains and interest from advertisers, but the effort to continue developing a variety of performing templates was too high. We pivoted to focus purely on AI improvements in order to automatically scale creative variations and templates.
bottom of page