Performance Spotlight: How Laura Geller Turned Creative Volume and AI Into a Competitive Edge

When Will Frank describes his company's Meta ad account, he doesn't reach for a spreadsheet metaphor. He calls it a fishbowl.

“It’s an ecosystem. You can look at any one thing in isolation, but everything affects everything else — creative, media, offers, channels, even what’s happening in retail. The real job is understanding how those pieces work together.”

It fits a brand that sells across its own site, retail partners, and brick-and-mortar. Each channel has different customer behavior, packaging, versions of what "converting" looks like. Laura Geller, a beauty company focused on products for mature women, grew by adapting its offers and creative to each channel instead of forcing one playbook across all of them.

And increasingly, Frank says, the marketer's job has fundamentally changed. The tools have gotten good enough that the mechanical work of configuring ad platforms has given way to deciding what the whole system should look like.

Creative as infrastructure

Frank credits Laura Geller’s growth to creative output. His team treats ad creative as ongoing production, not a last-minute asset.

"We have a creative team that just supports our ads," he says. "They do a really great job of building out a massive pipeline from our creators, from the agencies we work with, and then the ones we’re building out ourselves."

That pipeline keeps the system constantly supplied with fresh inputs. When Meta’s Advantage+ and AI tools handle delivery, targeting, and placements, the biggest differentiator becomes what you put into the system. "We all know it's super important nowadays with new AI powered models like Andromeda surfacing," he says. "It allows us to always have a constant flow."

The team keeps their media buying in-house, too. "No one knows or loves the business as much as we do," Frank says. “Everyone on the team comes from an agency background, which means they've been hands-on-keyboard before, but now they're also the decision makers. That's what lets them move fast.”

A testing culture that knows when to wait

Laura Geller runs a heavy testing agenda on Meta each quarter, aligning with their Meta team on which bets are worth making. But speed doesn't always mean pulling the plug early.

Frank points to their experience with value optimization. "It took probably a couple tests and about three weeks before we really started to see it gain life," he says.

Early signals were encouraging. New reach and higher average order values, but conversions took time to materialize. "We were seeing some really good movement with the upper funnel metrics, but not necessarily on that conversion," he says. "It made us kind of excited to keep going."

They kept going. Value optimization is now one of their workhorse campaigns. The principle behind it is straightforward: when you optimize for the outcome you actually care about rather than a proxy like conversion volume, the results follow. But you have to know your account well enough to understand the tradeoffs you'll see along the way, and have the conviction to let a test breathe when the data gives you reason to.

AI as a pressure-testing tool

Frank's team uses AI across the business, but the framing is practical. They use it to shorten the loop from analysis to decision, while keeping final calls with the team.

"The biggest thing is how we can automate a lot of the reporting and the understanding, so we can focus on how to take or build an action item out of it and actually go do it," he says. "Getting back to doing more of the actual job of marketing, rather than the reporting."

On the creative side, AI lets his team turn one ad into six variations to target a variety of audiences – new hooks, crops, captions – compounding the volume that already flows from their creator and agency partnerships. His team can test more visual concepts earlier in the process, before committing to a full production.

"We can iterate and come up with it and just test it," Frank says.

But the use case Frank seems most enthusiastic about is what he calls using AI as a "second council." His team runs their own analysis, forms a hypothesis, then challenges AI to see if the reasoning holds. "We have our assumptions, and then we challenge it to see if it's spitting back something similar," he says. They've used it to evaluate headlines, simulate customer reactions, and audit their own website for usability issues. "Almost like a secret shopper," he says.

For Frank, this is what it looks like when AI crosses a threshold, not replacing marketers, but removing enough friction that they can operate at a higher level. The repetitive production work gives way to strategy, creative problem-solving, and system design.

The customer journey is the strategy

Laura Geller's product shows up in very different contexts depending on where the customer finds it. Bundles perform well online. Single SKUs win in retail. The core message stays the same, but the packaging, the creative, and the offer structure all flex to meet the customer where she is.

That cross-channel thinking is what Frank spends most of his time on now. Not the mechanical work of setting up audiences (Meta's Advantage+ tools handle that) but the strategic work of designing how each channel connects to the next, what the measurement stack needs to look like, and where the next dollar should go.

"I'm not in there anymore doing targeting, which I think is amazing," he says. "I don't have to sit there and do the lookalikes or interest groups that I grew up doing. I love that part."

What replaced it is harder to hand off: reading the data, understanding what's actually incremental, and making the call on where to double down. Laura Geller has built what Meta calls a 'suite of truth' — multi-touch attribution for daily optimizations, a marketing mix model for budget-level decisions, and incrementality tests to validate both.

"There's a lot of things that claim they're incremental," Frank says. "How do you test them? There's no silver bullet."

It's exactly the kind of higher-order thinking that can't be automated. Frank has moved his value up the stack, from execution to architecture.

What's next

Frank sees the pace of change picking up. "Over the last year and a half, it feels like the industry has accelerated faster than it ever has," he says.

He's watching agentic media buying with keen interest. "I'd love for the next wave of AI developments to allow us to do even more at scale," he says. "It's evolving so fast. Talk to me in a year and it'll probably be a completely different conversation." In fact, around the same time we spoke with Frank, Meta was already preparing to launch AI-powered tools that enable agents to take actions directly in Ads Manager.

He references the famous Oreo Super Bowl moment, when a team of ten people scrambled to post "dunking in the dark" during a blackout. "Nowadays, or in the future, if something quickly happens, we can react very quickly with ads, landing pages, product bundles," he says. "AI would allow us to move that much quicker."

As for where he thinks things are heading? "We're gonna become shepherds of AI in the end."

But Frank isn't waiting around for that future. He's already designing the system. Linking a creator partnership to a campaign to a landing page to a measurement framework that tells him whether any of it worked. For Frank, AI has made the job what he always wanted. He spends less time in the weeds of targeting and reporting, and more time solving the actual puzzle of growth.

And as AI crosses more thresholds in the months ahead, that's the bet he's making. That the marketers who move up to the strategy layer now will be the ones shaping what comes next, not reacting to it.