The challenge of scale
One of the hardest lessons I learned at Meta was that the more users you have, the less you can design for any single one of them. A local bakery owner spending $20 a day on ads doesn’t think like a Fortune 500 marketing team managing millions in spend. A first-time advertiser launching their first campaign doesn’t behave like an agency managing hundreds of accounts. Yet they’re often using the exact same product. As designers, we’re taught to deeply understand our users and design around specific needs. But when your users number in the tens of millions, designing an experience that represents the needs of all users becomes very tricky.
The trap of designing for everyone
The natural response to diverse user needs is to add more flexibility. More settings, more controls, more customization, and more surfaces. Over time, every edge case earns a feature, every workflow gains an exception, and the product becomes capable of doing almost anything. Many enterprise products don’t suffer from a lack of functionality, they suffer from an abundance of it. The pursuit of serving everyone often creates experiences that feel overwhelming to everyone, and are optimized for no one.
From experiences to systems
What I adapted in my approach is to stop designing individual experiences and start designing systems. Instead of asking what a specific user needs, find the patterns that exist across millions of users. At Meta, one recurring pattern became obvious: advertisers wanted better results, but they didn’t always know what action to take next. The recommendations varied, the business goals varied, and the account sizes varied, but the underlying problems and unanswered questions remained remarkably consistent.
Why prioritization matters more than information
This challenge became especially apparent while working on Opportunity Score. Over time, Meta accumulated hundreds of recommendations, thousands of errors and warnings, all built by dozens of disconnected product teams throughout Ads Manager. Each recommendation was useful on its own, but collectively they created a new problem: advertisers struggled to determine what mattered most. The issue wasn’t a lack of guidance. It was a lack of prioritization. Advertisers didn’t need more recommendations, they needed help deciding which recommendation deserved their attention first. Identifying the common patterns is what helped the team design solutions that all 10 million users can benefit from.
