The Women in Experimentation Summit. 3rd Edition.

Open a dozen AI-built landing pages. Same hero. Same gradient. Same “clean” card grid. Victoria calls it the Claude UX Factory - and she’s here to shut down the assembly line. Working across UX, UI & CRO at Miso, she’ll show you the visual tells of AI-generated design, where AI-built flows fail real users, and the exact point in your process where AI helps most. Leave knowing what to automate, what to keep manual, and how to keep shipping fast - without surrendering the human touch that makes a brand worth noticing.

AI can write your email, personalize your push, and A/B test your subject line before your morning coffee. The marketing that persuades? A hand-written note you don’t want to bin. “Do things that don’t scale” is everyone’s favorite advice. Yet nobody’s playbook! Lauren has scaled consumer brands from 0 to hundreds of millions - founding team at HelloFresh Canada, plus Huggies, Spin Master, Sano Genetics, and Tonies. Her case: offline is a performance channel. Measurable. Testable. Almost entirely under-resourced. She’ll show you how to run it like one - just in time for Gen Z bringing back landlines.

A winning experiment isn’t a winning business decision! Read that again. Teams incentivised to grow at any cost rarely ask what a stat sig uplift costs in brand credibility, trust, and retention. This is no metrics sermon. Ester’s framework helps you see familiar numbers in a new light: as hiding places for trade-offs that sabotage CRO programs. Balance primary metrics against guardrails, without allowing cherry-picking.

Want to watch an LLM design and deploy an A/B test in real time? Us too! But the AI flourish isn’t the point. When testing gets (almost) free, half-baked ideas ship at scale - and “tests run” stops being a flex. Everyone senses the fix; nobody wants to say it out loud. Matt will: slow down, intentionally. He’ll show you where to add friction, how to build learning loops, and which decision criteria make results trustworthy so the restraint can sharpen momentum, instead of stalling it.

Before Julie helped Wikipedia teams learn faster through privacy-first testing at billion-user scale, she was a marine biologist. Running experiments. With dolphins! In the wild, there’s no “just rerun it” - preparedness and discipline are how you get any signal at all. Julie learned to carve out pockets of control inside genuinely uncontrollable conditions. Sounds like testing in the AI flux, no? She brings the toolkit to navigate variance, chaos, and stubbornly small effects - for everyone who’s ever wondered whether their control condition is actually controlled.

We treat data as neutral. It isn’t. Every KPI carries assumptions about who the “average user” is - and as AI takes over ideas and execution, those assumptions harden into fact, at scale. Suddenly your program optimizes exclusively for heteronormative, able-bodied, cis, white, middle-class users. Sabine uses queer and feminist theory to surface what’s baked into your test metrics - asking not just who your data talks about, but who it alienates. People in experimentation hold power over choices. That power shouldn’t be allowed to manipulate.

Every experimenter has a plan for losing tests. Almost no one has a plan for the winners. A stat sig uplift is a promise your whole business now has to keep - demand, capacity, support, fulfillment. Himani has seen wins scale faster than the systems built to hold them, and she’ll show you why mature programs in Fortune 500s ask “what breaks when this works?” before launch - to avoid an unforeseen collapse.

When we chase expansion revenue, we default to new - new features, new markets, new ad spend. Nikki, host of the Churn Flags podcast, says the expansion you’re after is sitting in the product you already have and the customers you’ve already paid (dearly - CAC has never been higher) to acquire. Churn and upgrade decisions don’t happen on AE calls. They happen in-product, every time a change in your app, your user, or the market forces a re-evaluation. Nikki maps the 9 moments where you can sway that decision in your favor.

Last year, Hesh Fekry stood on this (virtual) stage and declared: “We are drowning in frameworks! Period.” Iqbal arrives with the rescue plan. “Don’t outsource your thinking to AI” is popular advice. He asks: weren’t we already outsourcing it to acronyms? Iqbal has spent serious time training language models and studying how they “think” - and why that isn’t anything like human intelligence. His session: practical techniques to rebuild deep understanding, the one capability that will stay premium.

Who’s your best customer? Easy - the biggest spender, right? Liliann disagrees. That model is dying, and with CACs shooting through the roof, you can’t afford the blind spot. Behavioral signals - recency, frequency, website visits, email engagement, purchase patterns beyond gross AOV - paint a different portrait: customers displaying real intent while flying under the radar. Liliann shows you how to spot these subtle signals, act on them, and re-segment your existing base for better targeting, experimentation, and decision-making across the board.