This release brings five updates to Zeal 2.0 — two to how we report on enforcement, and three to how we catch infringements faster. Here's what's changing, who it's for, and where to read more.

In this release: Organisation Dashboard (rebuilt) · Websites & Domains Dashboard (new) · Rule Suggestions (new) · Reinstated Domains (new) · Screenshot Logo Detected (new).

Proving the value of your program and spotting trends used to mean piecing numbers together from different places. The Platforms Dashboard brings marketplaces, social media, app stores and paid ads into a one clean view, so you can go from a high-level snapshot straight into the detail behind it.

What's inside: key figures and estimated value removed, a 12-month Infringements vs Removals trend, the Removal Lifecycle funnel and Time to Action, keyword Visibility and Cleanliness Score, a geographic risk overview, and offender and product deep-dives.

📖 Helpcenter article: Platforms Dashboard

Webpage and domain enforcement never had a reporting home of its own, so it was hard to see how it was performing or which hosts and registrars were genuinely cooperating. This new dashboard brings it all into one place, with separate Webpage Enforcement and Domain Enforcement tabs you switch between with a single toggle.

What's inside: key figures, trends, Time to Action and the Removal Lifecycle funnel on both tabs, plus an Entities Breakdown ranking your most and least compliant hosts and registrars. The Domain tab adds domain-level views — Top 10 TLDs Enforced, Top 20 Domains by Enforced URLs, and a 60-day escalation queue.

📖 Helpcenter article: Webpages & Domains Dashboard

Auto-moderation only grows as fast as someone can build new rules by hand, and good rule ideas from our Machine Learning had no easy way to reach the product. Rule Suggestions changes that: Zeal now proposes ready-made rules for you to review and switch on, so you can extend automated coverage with almost no manual effort. Each suggestion sits in the auto-moderation area, switched off by default — review it, and if it looks right, enable it in one click to make it a live rule.

📖 Helpcenter article: Rule Suggestions

When we enforce an infringing domain it goes offline — but bad actors often quietly bring it back with new content, and because it was already “handled” it can slip under the radar. The new Reinstated Insight filter automatically flags previously enforced domains the moment they move from Offline back to Online, so repeat offenders are caught early.

📖 Helpcenter article: Reinstated Domains

You can only act on the infringements you know about, so a site you weren't already tracking could be misusing your brand unnoticed. Screenshot Logo Detected widens the net: whenever any domain's homepage is screenshotted, the image is checked against your logos, and any match is automatically tagged to your org and surfaced in your queue. It's the single highest-confidence filter in Zeal today.

Coming next: an in-screenshot bounding box showing exactly where the logo was detected. The tagging works today; the visual indicator (bounding box around the logo) is still to come.

📖 Helpcenter article: Screenshot Logo Detected

Alongside the features above, our engineering and machine-learning teams shipped a lot of work under the hood this past month. You won't find new buttons for these, but they make Zeal more accurate, faster and more secure. Here's a look at what changed:

Zeal automatically works out what kind of product each post is about. The model behind this had two upgrades: it learned 108 new product categories — recognising them correctly 92–98% of the time, with no drop in accuracy on the ones it already knew — and a follow-up sharpened how it tells similar clothing items apart. On tricky, disputed cases it now picks the right category about 76% of the time, compared with 13% for the previous version. Both updates fixed real miscategorisations that clients had reported.

Ask Cortex, the assistant that answers questions about your data, was rebuilt on a more secure foundation, with protection that ensures each organisation only ever sees its own data. It answers questions about posts, enforcement, moderation and takedowns — and will politely say when something, like follower counts or domain registrars, is outside what it covers. Across extensive testing it answered correctly 97.7% of the time, in about four seconds on average, while costing roughly 19% less to run than before.

The technology that spots brand logos learned 14 new brands and got more accurate on 45 it already knew, so more logo misuse gets caught automatically. This is the same detection that powers Screenshot Logo Detected, above.