What’s new in Moistur AI. We ship improvements to how we measure and improve your brand’s visibility across up to 8 AI assistants, including ChatGPT, Claude, Gemini and Perplexity, on a regular cadence.
Last updated: October 2, 2026
A rebuilt public site and two report fixes
Features page rebuilt on real product screenshots, a contact page with the booking calendar, and a sample client report: the weekly playbook and monthly brief exactly as a client receives them.
A guide to how the AI visibility score is built, and Peec AI and Otterly AI alternatives pages on the same ten-client benchmark as the Profound page, every price re-checked on the day.
Every guide now has its own share card, and signed-in customers can read the pricing page again.
Monthly brief: the headline names the assistant (Claude, ChatGPT) rather than an internal model id. Weekly playbook PDF: the verdict cards grow with their text instead of cutting it off.
A session revoked on another device no longer traps the browser in a sign-in loop; one honest bounce clears it.
Homepage loads with a lighter hero video; text contrast lifted to AA across the site.
Sign-in fixes and honest portfolio deltas
Password-reset links now work even when you are already signed in.
Sign-in lockouts are scoped to the email and its source together, so one failed source can no longer lock an account out for everyone.
Agency portfolio: a client with no measured change this period is shown as unmeasured, not as flat.
Comparison pages
Side-by-side comparisons with Ahrefs, Peec, Profound, Semrush, Otterly and Scrunch, every rival figure dated and sourced, plus a Profound alternatives guide and a white-label reports page.
Author pages for both founders, and one consistent company record across the site.
Security hardening
Every server action now checks authorisation before it reads or writes data, and an automated gate in CI blocks any change that breaks the rule.
Next.js updated to 15.5.25 with 17 dependency updates, closing every reachable advisory.
Security & Trust page corrected to match the code: SOC 2 not certified and no auditor engaged, SSO no longer listed as live, and the complete sub-processor list.
Dry, Soggy, Moist
The Hydration card on the dashboard and the cover of prospect reports use one ladder: Dry, Soggy, Moist.
No invented zeros
An unmeasured metric shows a dash everywhere: weekly digests, CSV exports, client report links, playbook deltas and the agency table. Some of these used to print 0.
Prompts are archived, not deleted
Removing a prompt archives it and keeps its scan history. The menu item now says Archive.
Billing emails and reliability
Clearer billing emails across the subscription lifecycle: receipts, renewals, and payment-issue notices.
A provider-health watchdog that catches AI-model outages early, so a model going down doesn't quietly corrupt your scan data.
The Weekly Playbook, rebuilt
Redesigned Weekly GEO Playbook: a clear verdict, a what-changed breakdown, a short forecast, and prioritized, consulting-grade actions, delivered every week.
White-label reporting: agencies can put their own logo and colors on shareable client reports and PDFs.
Sharper scoring: better handling of partial agreement between models, intent-weighted prompts, and more accurate brand detection (including aliases and misspellings).
More honest metrics
Brand visibility now reports a confidence interval instead of a single, noisy number, because one AI answer isn't a measurement.
Sentiment scoring expanded to multiple dimensions for a truer read on how models talk about your brand.
Cross-model disagreement is now a first-class signal: see not just where you stand, but how consistently the major AI models agree about you.