Laurence

performance marketing is fundamentally a trading problem: while trading has evolved from arbitrary heuristics to systematic, quantitative algorithms performance marketing is stuck at if-else rules from the 1990s @trylaurence is a dinosaur killer

We grew a brand's total sales by 2x within a week. Read the full case study at trylaurence.com/case-studies/h…

The reason why brands choose @trylaurence to manage their Amazon PPC is dead simple: we are miles better than any other option. Results like these need no explanation: - 4x ROAS - 13% TACOS And we've doubled organic sales at the same time ;)

@Monetisedev @ycombinator @mlaurencechen we use contribution margin to set profitability boundaries for every SKU ;)

@ycombinator @trylaurence @mlaurencechen Automating bids is nice, optimizing contribution margin per SKU is the real win. When ads become a control system for unit economics, not just spend, CPG teams scale with way less guesswork. Congrats to the team 🚀

.@trylaurence automates performance marketing for CPG brands, starting on Amazon. They use reinforcement learning to set bids, budgets, and keyword targets like quant hedge funds trade markets: continuous optimization on autopilot. Congrats on the launch @mlaurencechen and @LGierhake! ycombinator.com/launches/PNP-l…

perpetua, pacvue, and teikametrics simply don't work. outsourced workers copy pasting to and from spreadsheets tanks performance after a few months. @trylaurence lets your brand advertise how hedge funds trade billions: with math and statistics

@mlaurencechen You've got to give a bonus to the engineer who thought of that

@trylaurence now uses light mode 🌞

trylaurence.com does it for you :)

We're sharing the secrets behind how we quadrupled this struggling Amazon product's top-line sales while barely increasing spend (5.7x ROAS):

Others pay offshore teams $5 an hour to run playbooks on your Amazon Ads. They're guessing on your dime. We explicitly calculate risk, knowing when to pull back spend and when to scale it back up. We laid low over Christmas and New Year because that's what the math was telling us. Once our forecasts were confident, we doubled our spend while tripling sales, then did it again. The result is 28% ACOS for an $8M/year seller whose previous agency delivered 40% ACOS on a good day. Calculated, not lucky.

Holiday volatility exposes bad ad systems. We managed risk through the volatility, scaled spend, and still delivered ROAS above historical baselines while keeping TACOS in check. When data gets noisy, guessing is expensive.

With less than $20k in my bank account, I quit my job at Google and lost $140k... Leaving was the hardest decision I've ever made- my whole life has been spent chasing financial security and safety, and that's exactly what I got at Google. But that safety came with a ceiling: I wasn't growing, I wasn't learning, and I was instead optimizing for comfort. So I left it all behind to start Laurence: a thinking machine that surfaces deep insights from any datasets backed by statistical proof. The hardest part of working with data is knowing what questions to ask, and Laurence both finds those questions and gives you answers. I don't know where this journey will lead. But I'd rather build something of my own than stay comfortable pushing paper for big tech.