MPMS 2026: Simon Whitcombe on the Next Frontier of Performance Marketing

Speaker: Simon Whitcombe, VP, Global Business Group, North America, Meta

Session: Opening Keynote - The Next Frontier of Performance Marketing

Timestamp: [00:26:43 - 00:41:14]

Key Takeaways

Whitcombe's entire keynote - and the framing for the whole day - is built on three architect questions: Who tells your story? Where does the transaction happen? How do you know it worked?

The answers: Creators. Everywhere. Incrementality.

Who tells your story? Creators.

70% of consumers now trust creator recommendations over traditional ads. Whitcombe said he would not have believed that stat possible when he joined Meta 13 years ago. It's not because creators got louder - it's because AI-powered discovery got good enough to match the right creator content to the right person at the exact moment they're ready to act.
Over half of all online shoppers say influencer content shapes what they buy. Creators are not a brand-awareness play anymore. They are a direct-response channel.
Discovery to purchase in 11 seconds. Creator wearing a jacket in a Reel, viewer taps, buys. No link in bio, no redirect, no friction. The ability for creators to link products directly in Instagram Reels is rolling out to everyone soon.

Where does the transaction happen? Everywhere.

Business AI agents are in testing now - trained on your catalog and your brand voice. They answer customer questions, guide high-consideration purchases, and act as your best salesperson. The framing: eCommerce has historically been a sub-$100, quick-click story. AI agents change the aperture entirely. A $200 jacket, a $5,000 vacation package, a B2B software contract - all now in play online.
"The conversation is the landing page." For high-consideration purchases, you've always been guessing the questions in the customer's head. With an AI agent, the customer just asks. The thread becomes a relationship, not a single transaction.
Your catalog metadata is the prerequisite for all of it. If Maya wants to know if the jacket comes in petite, the answer doesn't come from a product title - it comes from well-structured metadata. Whitcombe's exact words: "Without that, everything you're about to see today are just demos. With it, they're GMV engines for your business." Gayatri Ganu and Andrea Moore go deep on this later in the day.
Muse Spark - announced in this session. Meta's first large language model from Alexander Wang's Meta Superintelligence Labs. Built from scratch in nine months. Now powering Meta AI across all apps and AI glasses.