Senior Director of Product · 2023 to 2025
I launched GLP-1s at Hims & Hers and rebuilt payments underneath them
When Hims & Hers started selling GLP-1s for weight loss, an order went from $30 to $1,500 and fraud rose with it. I split checkout into a risk path and a fast path, matched onboarding to how ready each person was, and let the team test prices without engineering. The launch brought in over $100M of new revenue in its first year.
Hims & Hers2023 to 2025
- new revenue in year one
- $100M+
- less payment fraud
- 23%
- higher lifetime value
- 31%

Where things stood
Hims & Hers grew on hair loss and erectile dysfunction treatments, and the platform was built for small orders. GLP-1s changed the size of every order. My first fix was tighter rules at checkout: card checks, velocity limits and a manual review queue. Real customers waited up to 12 hours for a first order, conversion fell 6%, and fraud barely moved.
Calls I made
I split checkout into two paths
One set of rules for every order either let fraud through or blocked real customers. A real patient leaves a care history: they message providers, refill and answer follow-ups. So I scored payment risk and care history apart. New customers with a thin payment profile pay for a consultation first. Customers with a care history skip the extra checks.
What happened: Fraud fell 23% and payment conversion rose 8%.
Two checkout paths
Real customers waited up to 12 hours for a first order. Conversion fell 6% and fraud barely moved. I matched onboarding to how ready someone is
Everyone got the same intake, whether they were ready to start or just looking. With the data science team I shipped a model that reads intent at intake. A wrong guess sends someone who is just looking into a fast checkout for a prescription drug, so I tuned it to be sure before it skips anything. Ready people go straight to checkout. Unsure people get more to read first.
What happened: Onboarding conversion rose 18% and acquisition costs fell.
I took engineering out of pricing tests
Every pricing test needed engineering work, so finding the right price for each group of customers was slow. I launched an incentives system on our customer data platform that serves targeted discounts by itself.
What happened: Lifetime value rose 31%, and pricing experiments ran 40% faster.
What changed
- new revenue from the GLP-1 launch in its first year
- less fraud, while payment conversion rose 8%
- higher onboarding conversion from an intake that reads intent
- higher lifetime value from targeted pricing
- faster pricing experiments, with no engineering for each test