Facebook  |  2020

Ads Manager: Automated App Ads

I led design of private beta and public launch versions of Automated App Ads- a streamlined ML-powered product that allows customers to set up a single campaign that will to do the work of 50-100 manual ads.

Challenge

Your typical mobile app advertising customer is constantly running dozens to hundreds of active campaigns- with each campaign being overwhelmingly complex to set up. Automated App Ads solves for this problem by delivering sustained performance with minimal inputs and on-going campaign management.

Business goal: Take the burden of campaign creation and maintenance, and put it on Machine Learning.

User goal: Simplify ad creation, without compromising demographic targeting context.

Ideation

By gathering inputs from prior research and related teams, I was able to pull together over a dozen unique different concepts via a dedicated design sprint.

Good principles form good UX

By aligning broadly with design principles first, I then leveraged those as criteria for selecting a UX model. The principles acted as a filter, allowing us to cull down to testable concepts.

Refined with usability

Once we identified the UX models to test, usability quickly helped identify friction points, and the feedback loop refined nuances.

Result

Customers enjoy an automated ad creation experience, where they just provide the content and we distribute it for them across the Facebook ecosystem.

“It took 3 AAA campaigns compared to 156 manual app ad campaigns to achieve similar or better results, with only 1% of the campaign management actions”

Outputs

I planned and led execution on the end-to-end advertiser journey from campaign setup to publishing, while leveraging usability testing and data from beta tests to make agile product pivots until product launched.

(sensitive content redacted)