The Growth Signal Stack: Measure the Decision, Not the Available Data
A decision-first framework for choosing growth metrics without mistaking publishing, rankings, or TVL for attention, adoption, or customer value.
Loading
Original research 11 posts.
Hi Violet research and field notes filed under “original research.”
From Commit to Customer: A Shipping-to-Demand System for Technical Products
A practical operating system for translating meaningful product changes into clear explanations, useful distribution, measurable action, and market learning.
The Minimum Viable Growth Ecosystem: How to Design a Connected Public Surface
A practical system for deciding which public surfaces a consumer-crypto product should maintain, repair, consolidate, or deliberately decline.
The Measurement Dark Matter of ChatGPT Ads
A reported conversion, an observed site action, and a verified customer are different facts. Nine public ChatGPT Ads case records show why that distinction matters.
What Makes Indie Game Devlogs Go Viral on TikTok? 12,489 Videos
We tracked 12,489 TikToks from 314 indie game developers. The devlog checklist made no difference; a breakout lifted the next four posts by up to 1.35x.
AI 3D Gets a Fifth of the Reach on TikTok: 7,317 Game-Art Clips
AI 3D clips get 0.21x the views per follower of traditional 3D on TikTok. Character animation reaches 9.1x, Blender scenes 8.6x. 7,317 game-art clips measured.
Steam's Coming Soon List Skips 71% of Games. What Earns Followers
71% of upcoming Steam games show no release date and never appear in the default Coming Soon list. Demos earn 1.67x the followers. Data from 60,790 games.
AI Detector Robustness: Data and Methods from Our RAID Reanalysis
Full aggregate tables from Hi Violet's RAID reanalysis: detector accuracy under 11 attacks, per-document flip rates, per-domain false positives, methods.
Can a Paraphraser Beat AI Detection? (Data, Not Hype)
Measured on 480,000 generations: paraphrasing makes one detector 5.4pp more accurate and another 17.0pp worse. Evasion is a detector lottery, not a strategy.
Where Detectors Break: The Domain Map
One pooled 5% setting, eight writing domains of truth: measured per-domain detector false-positive rates range 0.96% to 20.4%.
The Flip Rate: Why Detector Accuracy Hides the Collapse
Measured on 480,000 generations: one symbol substitution flips 93.7% of a detector's per-document verdicts. Aggregate accuracy hides the collapse.
Bring the market problem into focus.
Want this kind of research run on your product? Tell us what you want to move and where it's stuck, and we'll reply by email.