Experiments
Operating experiments, in the open.
Public proof-of-work. Each experiment runs a property through the BEHOLDINGS operating system and reports what we changed, what we're measuring, and what we're deliberately holding back. Not a blog — a record of how the system actually runs, including older tests that didn't all work.
Running since 2019. Published now.
We've been building, breaking, and refining this operating system on real properties since 2019. What's public here is the start of that record — the most-tested apps, opened up deliberately. It is not everything we've run, and these properties are intentionally still proving themselves, not finished products.
Data confidence: Verified · Reconstructed · Remembered
Reviving BeTheHippy
BeTheHippy was once the strongest organic property in the portfolio, reportedly reaching ~25k GSC clicks in a 30-day window in 2021 before declining close to zero. This experiment tests whether an old content property can be revived into a buyer-assist engine.
WhatIsLeaving Alert Utility
Testing whether structured streaming-removal data can become a repeat-use utility through platform pages, monthly archives, title pages, leaving-soon hubs, and email alerts. The property has grown roughly 50× from launch baseline, has crossed the internal display-ad readiness threshold, and is now the anchor of a small Leaving Network alongside GamesLeaving.
Hearthbruk Local Partner Validation
Testing whether agents, inspectors, or home-service partners will pay for homeowner handoff and retention tools.
ZodiacStreak Ad-Arbitrage Postmortem
Testing whether a new astrology property could buy low-cost search traffic and monetize through ads and email capture. Early signs worked, but the arbitrage model proved too fragile to become durable acquisition.
Trackifya Postmortem
Tested whether carrier-tracking pages could capture package-tracking search demand and support ad monetization or paid arbitrage. The site reached the index but never reached ranking, producing only 9 organic clicks across roughly 15 months. The lesson: wrapper pages do not create durable acquisition unless they add original data or value the source-of-truth sites cannot provide.