When we launched Gigi in fall 2025, we gave our customers a living PDF of all the things Gigi could do and all the things she couldn't. We wrote our first line of code in January 2025, so the product was nascent and its capabilities were naturally limited. Every capability Gigi has maps to plumbing. An Amazon Ads API endpoint becomes a tool Gigi can call. Agentic skills built on those tools turn raw API access into media manager work. No endpoint, no tool. No tool, no skill. No skill, no capability. We knew Gigi was limited, but we wanted the product in customers' hands so they could tell us which net new capabilities to prioritize.

Their feedback and guidance was crystal clear. We were handed a litany of new capability requests, and customers flagged performance gaps immediately: latency and reliability. We did the heavy lifting, painstakingly integrating with almost every single API endpoint in Amazon's DSP. It took months. During that time, some customers stayed. Others churned quickly after adopting Gigi. They churned because the promise of Gigi didn't meet the reality of Gigi in that moment. Our PDF of the capabilities Gigi did not possess required too much mental bandwidth. Software gets forgiven for being incomplete. A point solution that does one thing well still earns its seat. An agent doesn't get that grace. You hire an agent the way you hire a teammate, and a teammate who does 60% of the job doesn't shrink your hiring problem. It adds to it, because now someone has to remember what the teammate can't do and cover the gaps. It was natural that we faced early churn. Gigi couldn't dramatically enhance the existing team's productivity or mitigate our customers' urgent hiring needs.

The learning for us: we needed to go a mile deep in a single media channel, the Amazon DSP. Our customers were hiring Gigi to do all of the tasks of the media manager, not a portion of the work, and they were only willing to reward us financially if they knew Gigi could do all of it.

Churn hurts. The winter was dark. Revenue was flat. Vibes were off. But we knew what we needed to do. The directive from our existing customers and our churned customers was the same: if Gigi can perform all of the tasks of the media manager in the Amazon DSP, fast and reliably, then I'll hire Gigi. It took time, but by the end of Q1 we were confident (hopeful) that the reality of Gigi met the promise. A new era for Gigi was upon us.

First impressions of new software products are critical. Your customer's first experience needs to delight. You need to build trust as fast as possible, and most importantly, you can't lose it early or you may never get it back. The first new agency we could delight in this new era was Bluebird Marketing. The second they got the product, the reaction was viscerally different from any other agency we'd worked with. The Bluebird team was engaged. They accepted Gigi's work, automated her work, and complimented her work (Bluebird is the most polite customer to Gigi, perhaps it's the southern charm). They quickly began prompting her to take action across as many clients as possible. Credit to the Bluebird team. If AI adoption is a proxy for how much you care about your job, Bluebird is one of the most passionately and intellectually curious teams we've come across. Their adoption curve was profound. They were the first agency to immediately see the value in the new Gigi, and we felt vindicated. We're proud to share our Bluebird case study this week. They will always hold a special place in Gigi lore.

Since then, almost every new agency we've onboarded has followed a similar adoption curve. The value we provide is demonstrable, and these agencies are quickly realizing that Gigi uniquely enables them to rearchitect their org structure in a way no other software could.

I often get asked why we're still only on the Amazon DSP. Other agentic media buying products have emerged in the last two years, and some offer surface level functionality across a variety of media channels out of the gate. Yet we're still on the Amazon DSP. Our perspective is simple: to hire an agent like Gigi, you need to know that Gigi does all the work, not a portion of the work. We went a mile deep on Amazon DSP because our customers told us to, and now we're being rewarded for it. We work with some of the largest agencies in the world, managing hundreds of millions in Amazon DSP spend on their behalf, and our revenue reflects that. We finally feel like we're in a position to expand to other media channels. When we do, we'll do it methodically and thoroughly. The next channel won't ship with a PDF of what Gigi can't do.