Shipping weekly
What's new in Bubbles1 — features, improvements and fixes.
PromptDNA™ outcome scoring v2
Prompt scores now weight citation quality by buying stage. Low-yield prompts auto-retire with full history retained. Average token spend per client down 38%.
Claude & Perplexity native integration
All four major AI platforms are now first-class citizens: daily share-of-answer, citation tracking and drift alerts across ChatGPT, Gemini, Claude and Perplexity.
Audit engine: 20 new GEO aspects
llms.txt validation, chunkability scoring and answer-shape detection join the audit — now 100+ defined technical aspects per crawl.
Client portals for white-label reports
Give clients a read-only login under your brand — scoped to their project, always up to date, no more PDF email chains.
Crawler improvements
Better JS rendering on SPA-heavy sites, smarter rate limiting, and fixed hreflang detection on multi-domain setups.
Always getting better
New here? Start where the momentum is.