Cross-platform recommendations framework for Meta Ads Manager
Role: Product design lead | Timeline: ~1.5 years
Team: Cross-org V-team. Product Design (5+), Content (3), Engineering (10+), Product Management, Data Science, Marketing, Sales, Design Systems, Art & Illustration
Making sense of fragmented recommendation systems across Meta’s ads ecosystem to help millions of advertisers prioritize the highest-impact opportunities.
Overview
Meta’s advertising ecosystem generated a rapidly expanding volume of optimization recommendations across Ads Manager. As recommendation inventories scaled, advertisers increasingly struggled to identify which actions mattered most. Opportunity Score introduced a unified prioritization framework that consolidated recommendations into a ranked, system-level optimization experience, helping advertisers focus on the highest-impact opportunities across the account.
I led product design across core experience strategy, prioritization UX, orchestration models, and scalable guidance frameworks used across the ecosystem.
Existing landscape
Ads Manager is used by over ten million advertisers to publish ads to almost four billion active monthly users across Facebook, Instagram, WhatsApp, Messenger, and Threads. After publishing ads, advertisers monitor their active ad campaigns, making optimizations and adjusting campaign characteristics to improve performance in key metrics.
Ads Manager makes real time recommendations to improve performance, which appear throughout the advertiser experience, across multiple platforms and surfaces.
Impact snapshot
Users
~10 million
Inventory growth
Over 10x
Alerts
Thousands of unique issue codes
Org scale
Cross-functional V-team collaboration
Problem and opportunity
The problem
“How do I know which insight is truly valuable? How do I ensure I’m not missing the biggest opportunities?”
Before
- Recommendations operated independently of one another
- Guidance lacked prioritization
- Recommendation inventory competed for attention
- Advertisers struggled to identify highest-impact actions
- Complexity gap grew as guidance inventory scaled
The opportunity
“Show me my next biggest opportunity to drive meaningful performance improvement for my ad campaigns.”
After
- Recommendations operate under a cohesive system
- Guidance is prioritized for business value
- Recommendations are focused on the biggest opportunities
- Highest-impact actions are always a click away
- Scalable frameworks enabled over 10x inventory growth
Objectives
For advertisers
Personalized insights
Create a scalable insights system that curates opportunities to make meaningful campaign performance improvements with personalized guidance orchestrated under a simple scoring model.
For Meta
Scalable framework
Establish a framework that helps product teams across the organization to build their product knowledge into a personalized insights system deployed across the platform.
For me
System-level design
Design a product system that eliminates guesswork, and transforms campaign optimization from fragmented maintenance work into more outcome-focused and decision-oriented workflows.
Program details
My role
Responsibilities
- Product experience leadership
Led product design across core Opportunity Score experiences within Ads Manager. - Recommendation mechanics
Ranking and scoring the highest-impact optimization opportunities and related feedback and learning loops. - Guidance orchestration
Framework for surfacing the highest-impact opportunities. - Cross-platform integration
Integrated across multiple platforms and surfaces, internal- and advertiser-facing. - Design governance
Guidance frameworks, prioritization standards, and consistency models supporting long-term ecosystem scale.
Design philosophy
Fundamental principles
- Prioritize clarity over volume
More guidance does not automatically create better outcomes. - Surface highest-impact actions first
Advertisers needed confidence in what mattered most and a simple path to act on it. - Create scalable patterns
The system needed to support rapidly expanding recommendation inventories built across the org. - Reduce cognitive overhead
Optimization workflows should feel focused, not overwhelming. - Never punitive
Advertisers feel validated for optimal setups, not punished for deviating from them.
Challenges & constraints
Organizational complexity
Fragmentation
Opportunity Score required aligning independently shipped recommendation systems shipped across product teams, org charts, surfaces, and platforms. Errors, warnings, performance recommendations, budget guidance, audience insights, and creative suggestions were all built independently, with their own adoption goals, and unintentionally competed with one for the advertiser’s attention. The challenge was designing a unified experience, and creating a scalable system that could centralize diverse recommendation types into a cohesive, trustworthy, cross-platform decision-making framework.
Anatomy
Opportunity Score can be simplified down to two atomic elements
The “what”
A global account score
On a 0-100 point scale, the score represents how optimized an account’s campaigns, ad sets and ads are overall. It updates in real time, giving a clear understanding of the account’s health at a glance. A score of 100 represents an optimally-performing ad account, following all best practices.
Available points represent the value of recommendations to improve the score. This is important, because it explains the value of the recommendation in the context of its impact to overall account health.
The “how”
Insights to help raise the score
While campaigns are active, the system monitors for opportunities to improve ad delivery. When issues like fragmentated audiences or fatiguing creative arise, the score drops below 100, and personalized resolutions are recommended to advertisers.
As the advertiser is guided through actionable steps, the account returns to a more optimal state, and the score updates based on the point value of the recommendation.
Surfaces
Campaign table
The Campaigns page is the highest traffic surface in Ads Manager. Opportunity score is prominently featured on this extraordinarily high intent surface (arguably ranked among some of the most valuable screen real estate on the internet), anchored to the page header at all times.
Account overview
Opportunity score is positioned “above the fold”, as a focal point to the account overview page. Even in an empty states (or perhaps especially in an empty state), when no recommendations are available, it sets the context for account-wide health.
Campaign table
Prioritization model
Prioritization is a central principle that is carried throughout the framework, including in visual summarization layers
Top recommendations
Clicking on the score at any time shows the next best opportunity to drive true business impact.
Campaign actions
Campaign-level insights, using a ranking and prioritization logic to highlight the biggest opportunities.
Account overview
Ranking
The score contextualizes each recommended opportunity. Advertisers can understand the value of a recommendation by its’ point value
Contextualization
A key design investment was explainability, ensuring that insights are never generic, and always presented in context.
Trust
Building advertiser trust was critical to the adoption goals. Insights are always accompanied by statistical proof of value.
Visual design
I led art direction of the Opportunity Score brand across:
- Promotional marketing
- Social ads (Reels ads on Facebook and Instagram)
- Documentation and help center
- Illustration and art assets
- Design system integration
- Iconography, colors, and typography
- GTM assets
An ad campaign with teaser videos was published from the official Meta account to Meta platform users with advertising accounts (~10 million advertisers) to introduce the new standard for ad optimality and to announce the launch of Opportunity Score.
Design decisions
Orchestration
Insights ecosystem
To reduce conflicts and fragmentation, recommendations were designed to operate as a system, rather than as independent packages.
Framework scaling
The system was designed to support inventory growth without requiring entirely new interaction models or surface patterns.
Unified scoring model
Unified guidance under a centralized framework that prioritized opportunities through a shared scoring and ranking system.
Advertiser trust
Confidence through explainability
Makes system prioritization understandable and trustworthy. Recommendations included reasoning and actionable next steps.
True business outcomes
Recommendations were framed around projected performance improvements rather than abstract system health alone.
Validation, not punishment
Advertisers should feel rewarded for healthy campaign setups instead of pressured by punitive scoring mechanics.
Prioritization & ranking
Impact over volume
Emphasizes the relative performance impact of each opportunity to guide focus on the most meaningful actions.
Progressive prioritization
Progressively elevates the most impactful opportunities first, simplifying decision making even in complex advertising environments.
Identify the next best action
Eliminates guesswork by helping advertisers quickly understand which action would create the highest-value incremental gain.
Outcomes
Advertiser outcome
Higher-confidence decision-making
Opportunity Score reduced guesswork by giving advertisers a prioritized plan, not just disconnected recommendations.
It empowered advertisers to make more informed decisions and understand the value of the system’s personalized insights. More importantly, advertisers no longer had to guess which optimizations would drive the greatest impact. They could focus on incremental optimizations that are proven to be the most impactful.
On average, advertisers who used Opportunity Score in their advertising strategy saw a 5% lower cost per acquisition, allowing budget scaling to high-performing ad campaigns.
Business outcome
Product adoption lift, >10x scaling
Opportunity Score created a scalable recommendation framework that product teams could plug into across Ads Manager.
The framework exceeded adoption targets in alpha, and drove measurable increases in campaign creation completion rates and advertiser engagement, growing from roughly ~10 inventory in alpha to to over 150 by global launch. The performance gains felt by advertisers had a significant effect, and led to statistically significant gains throughout the funnel (adoption → outcomes → revenue). In other words, as advertisers followed more recommendations, their ads performed better, and their spend on Meta ads grew.
Personal outcome
Promotion to staff product designer
Opportunity Score fundamentally changed how I think about product design at scale.
Working on this framework highlighted the importance of system thinking over product thinking, and that designing effective recommendation systems depends not only on reducing cognitive overhead and reducing underlying complexity, but on building trust in the system.
I learned how often products unintentionally “ship the org chart” to users, and the long term consequences this can have at scale. Simplifying complex organizational structures into cohesive user experiences prevents this common pitfall.
More work
Media Customization Controls for Ads Reaching Billions
Media Customization Controls for Ads Reaching Billions
Ad Placement Settings in Meta’s Ads Manager
Ad Placement Settings in Meta’s Ads Manager
Campaign Scoring for Optimal AI-Powered Ad Performance
Campaign Scoring for Optimal AI-Powered Ad Performance
Risk Management App for Verifying Insurance Coverage
Risk Management App for Verifying Insurance Coverage
COVID-19 Monitoring Solution for Essential Workers
COVID-19 Monitoring Solution for Essential Workers
Workforce Management App for Optimizing Cost in New Markets
Workforce Management App for Optimizing Cost in New Markets
Improving Restaurant Operational Efficiency