Safety scores for the everyday products parents buy.

Renn Labs built R3 from scratch in early 2026: a new site that scores everyday products for safety, from air fryers and water filters to car seats and baby lotion, and helps US parents choose between them, positioned as the Wirecutter for clean parenting. It launched with 750+ scored products across 162 categories.

The brief called for speed and leverage over headcount: launch fast, scale content with AI instead of a writing team, and reach parents without the years of traffic an established review site already has. I joined in February 2026 as R3's only hands-on marketer, reporting to the founder and the Chief of Staff, expected to run search, content, social, email and recall alerts myself, as systems rather than as tasks done by hand.

A brand-new domain starts with nothing to rank on. Nobody searched for R3 by name, and parents already trusted established review sites, so every visit had to come from a product or category search. By May 2026, Google showed R3 about 90,000 times a month, and parents clicked about 0.12% of the time.

Two problems sat underneath. The pages changed almost every day, so Google kept seeing a moving target, and bots and team visits mixed into the analytics, so nobody could tell which channel worked.

Where it started

1. Built from zero, expected to move fast

   Renn Labs set out to launch R3 fast and scale it with AI instead of a writing team, positioned from day one as the Wirecutter for clean parenting.

2. A new domain starts with nothing to rank on

   R3 had no brand searches, in a category where parents already trusted established review sites. By May, Google showed R3 about 90,000 times a month and parents clicked about 0.12% of the time.

3. Numbers nobody could trust

   Team visits, bots and tracking that changed with every redesign mixed into the analytics, so nobody could tell which channel was working.

4. One marketer, expected to run a team's channels

   Search, content, social, email and recall alerts all needed running every week, with no team behind them, so the work had to run as systems.

I started by sorting R3's search demand by intent. About 89% of it was shopping: "best", "safest", "X vs Y". Only about 11% was safety research. A parent searching "best air fryer" is still worried about what's in the coating, they just won't type the chemistry, so I made safety the deciding detail inside the comparison and category pages they already look for.

From late May I stopped the daily page changes and gave every template one fixed structure, so Google had a stable site to rank and the team had numbers it could trust. Then I built the rest as systems one marketer can run: pages that grow with the catalog, a weekly search loop, a social post factory, recall watch and lifecycle email.