Ulta Beauty’s Next Growth Engine: Data-Driven Discovery at Scale

Josh Friedman, SVP of E-Commerce and Digital at Ulta Beauty, on personalization, marketplace expansion, and why loyalty is the real infrastructure.

At Shoptalk 2025, Josh Friedman, SVP of E-Commerce and Digital at Ulta Beauty, laid out a growth strategy that’s not about chasing trends—it’s about delivering relevance. In an increasingly complex beauty landscape, Ulta is using AI, first-party data, and a fast-scaling digital platform to deliver on a clear mission: to be the beauty destination of choice for every customer, every time.

From a soon-to-launch marketplace to deeper personalization and retail–digital fluidity, Friedman’s keynote painted a picture of a brand building for long-term loyalty—not just digital convenience.

A Marketplace Built for Growth—And for Gaps

Ulta isn’t launching a marketplace to be everything to everyone. It’s launching to go deeper—into emerging categories, wellness, and underrepresented brands.

With 600+ brand partners already in-store and online, the marketplace will allow the team to test new categories, trial innovative product lines, and onboard emerging partners without compromising curation. Wellness will be a key vertical here, reflecting what Friedman called “the blurring of lines between beauty and wellness”—a trend driven directly by customer demand.

The platform will be powered by Mirakl, with strong operational support to accelerate the launch. “They bring not just tech, but the services and standards we need to scale right,” Friedman said.

Personalization That Starts with Trust

In a category as personal as beauty, generic recommendations aren’t just ineffective—they’re insulting.

Friedman emphasized that Ulta’s personalization roadmap is grounded in real data and real expectations. With 44+ million active members in its loyalty program, the company doesn’t just collect data—it’s expected to use it.

That includes:

  • Ingredient preferences

  • Brand affinities

  • Shopping modes (pickup, ship, return-in-store)

  • Category exploration across price points

In 2024, Ulta formalized a strategic partnership with Adobe to drive this next phase of personalization, integrating zero-, first-, and third-party data to ensure every touchpoint delivers something meaningful.

AI That Supports the Associate and the Shopper

Friedman didn’t shy away from AI—but he didn’t overstate it either. Ulta is actively exploring agentic AI to remove friction from operational tasks and unlock human creativity.

Think less “virtual assistant” and more internal supercharger—helping teams onboard products faster, process returns efficiently, and spend more time curating experiences, not chasing manual workflows.

For the customer, that means better recommendations, faster service, and a more fluid experience from screen to store.

Measuring What Matters

Ulta’s attribution strategy is grounded in use-case specificity. As Friedman noted, “Different objectives require different models.”

For campaign effectiveness, they use:

  • Media Mix Modeling (MMM)

  • Multi-Touch Attribution (MTA)

  • Click-based models for ecommerce evaluation

What’s changed, he noted, is speed. Where MMM once required long lag times, newer systems are becoming more agile—providing actionable insights closer to real-time, allowing Ulta to move faster on budget allocation and performance optimization.

The Store Is Still the Star

Despite Ulta’s digital scale, in-store remains the brand’s foundation—and its superpower. Over 95% of purchases come through loyalty members, allowing Ulta to see how fluidly customers move between online and offline.

Buy online, pick up in store. Return in store. Browse in store, ship to home. The customer journey is channel-agnostic—and Friedman’s team measures success not by which channel wins, but by how blended experiences drive higher lifetime value.

What Ulta Is Teaching the Market

💄 Personalization without trust is just noise
🧪 Marketplace isn’t just about scale—it’s about testing what’s next
🔁 Omni-channel fluidity drives deeper loyalty than channel silos
🛠️ Agentic AI can unlock creativity—not just efficiency
📈 Attribution models must fit the objective—not the org chart

In a beauty industry obsessed with flash and fast growth, Ulta’s approach is refreshingly mature: listen to the customer, build for real behavior, and scale what works—slowly, then all at once.

Disclaimer: The above podcast episode was generated using AI based on an interview transcript. While the content remains true to the original conversation, the voices, tone, and delivery were synthesized and do not represent actual recordings of the speakers. This AI-generated format is intended to enhance accessibility and provide an alternative way to engage with the discussion.

SPONSORED BY
sponsored by Fospha

Independently Created. Not affiliated with Shoptalk.

ClickZ is a Contentive publication in the Events division