Zefr: Atrium

Bad things happen in our world often and when they do, it becomes risky to run ads that could accidentally land in front of content that is damaging to a brand's image and standards.

A brand's presence on social media isn't just about placement, it's also about representation. Anyone can reference or parody a brand in a way that is demeaning without providing a way for the brand to even know. These trends make it more difficult for brands to only depend on ad manager's tools for placing their ads.

Gartner estimates that by 2030, 90% of the internet could be AI-generated. Brands can't afford to block all AI content without impacting scale. AI content ranges from artistic expression and entertainment to misinformation and AI sexualized imagery. Additionally, if a brand does run against content that was undesirable what action should be taken? These are all very complex problems that I worked with various teams to navigate.

Zefr's social partnerships distribute the data. An intuitive UI turns it into a decision.

Our content analysts and data science team classify and label content in meaningful ways. Now customers can be aware of when their brand is misappropriated with generative AI or flagged due to any of the twelve GARM standards, like social issues or sexual content they might not align with brand values. The Zefr UI provides tools for setting up brand safety and suitability preferences, targeting, measuring live and previously run campaigns with reporting features allowing customers and their account managers to be more hands-on with their data ensuring that future campaigns run even more smoothly. Or if customers want to be less hands-on, reporting features provide schedule reporting and alerts giving peace of mind to them and the brands they service. I owned the UX end to end reporting to a VP of product, initially managing an external team of designers and later bringing that team in-house. Next, I'm helping the team focus on an AI driven solution that quickens the way customers can gain insights and also work with it via a Claude MCP connection.

Our measurement tool, now known as Zefr Atrium, did not generate any revenue when I started in 2021; today it generates roughly 10M+ in additional revenue.

While I was also very focused on improving our targeting and settings features, my biggest win at Zefr has been helping the team find a product/market fit that is making a difference in the industry and for the business. I've been able to achieve this by being collaborative across global time zones, being diligent with using product analytics to check heat maps, dead clicks, user behavior, etc., while leveraging Pendo, an in-application guidance support and notification system, to build guidance and support and gain insights via user funnels and their analytics. This allowed me to help shape our product road maps and understand what to build next.

(Note: All of the data in these mocks shown is dummy data for illustrative purposes only. None of it is actual platform or client data.)

Continue to expand reach and quicken the pace.

With a marketing brand update, Zefr will soon change it's skin. I've prepared designs for this process and currently working with a team to improve upon claude skills and our style guide. While it is important to maintain component integrity in Figma, it is no longer the source of truth for our design system. Leveraging storybook with claude has allowed me to have stronger alignment with my teams.