Multi-Channel Template Sets land on a refreshed Templates page
For Operators and agency users who publish the same campaign across multiple channels and want to reuse a coherent multi-channel layout instead of rebuilding it every time.
The Templates page is restructured around a new Multi-Channel Sets section that sits above the per-channel templates. Users can build a set by picking one template per channel from a sidebar picker, name it (up to 120 characters), and save it; saved sets are then displayed on the page and can be re-opened or adapted. Channel cards now show a single representative preview, the page copy and labels were aligned with the latest Figma (set labels now read by channel rather than by aspect-ratio sizes), and an empty Home Banner section was removed. Most of the new flow ships behind a feature flag.
Why it matters: Putting together one piece of content for Instagram, LinkedIn and email used to mean opening three separate templates and matching them up by hand. There was no way to save that grouping for next time, and the channel cards talked about pixel sizes instead of the channels users actually publish to.
A new Home page surfaces what needs your attention
For Anyone collaborating across multiple campaigns or stores — particularly account managers, agency operators and marketing leads who used to start the day digging through notifications.
A brand-new Home page lives at /home and replaces the placeholder with a real activity feed wired to the engagement-service backend. Cards show approval requests, mentions, scheduled posts, published posts and insights, with unread/read styling, dynamic unread counts on tabs, and a one-click dismiss menu. Per the latest design, the personal-attention tabs (To approve, Mentions, plus an upcoming Assignments tab) now collapse into a single 'Needs your attention' tab, while Scheduled, Published and Insights remain split. CTAs link straight to the relevant epic or task.
Why it matters: Approvals, mentions and assignments used to be scattered across emails, Slack pings and the campaign UI. There was no single place to walk in to in the morning and see, in priority order, the things waiting on you.
Image Guardrails come to Visual Guidelines
For Brand-conscious merchants and agency users who want consistent, on-brand AI imagery without re-prompting on every generation.
Visual Guidelines now has an Image Guardrails section. Users see AI-generated suggestions for what their AI image generation should and shouldn't do, can edit those rules, and save them. The saved guardrails are then injected into image generation prompts so future visuals follow the brand's rules automatically.
Why it matters: There was no way to tell Glowtify, once and for all, that brand visuals shouldn't include certain elements (people, competitor logos, specific colour combos, etc.). Users had to remember to re-prompt or fix images by hand every time.
Magic wand only changes things when you ask it to
For Anyone editing social-media templates who has been bitten by surprise magic-wand edits — designers and operators in particular.
The magic wand on social-media template editing now requires an explicit click-to-apply rather than auto-editing as you hover or change context. Users see what would be changed and only commit it intentionally, so the editor stops surprising them with adjustments they didn't ask for.
Why it matters: The magic wand was previously triggering edits in situations where users didn't expect it, and they'd find their template had been changed without an explicit confirmation.
Event calendar suggestions get tag-aware
For Merchants and content planners who use the event calendar to plan campaigns and want suggestions that match the type of moment they're actually planning around.
Calendar event suggestions are now generated with a stricter, per-category prompt and respect the chip filter on the calendar — so picking 'Commercial' or 'Cultural' actually narrows what's suggested. Seasonal events are dropped from the grid view, legacy events fall back to the Commercial chip via a new helper, and a cloud function regenerates suggestion fragments via Gemini 3 Flash for stores active in the last 30 days.
Why it matters: Suggestions used to ignore which calendar tag the user was looking at, so a merchant browsing 'Commercial' moments would still get cultural or seasonal noise mixed in. Existing events also weren't tagged consistently.
Campaigns now move themselves through review and scheduling
For PMs, account managers and reviewers who rely on campaign status as the source of truth for what's queued for review or scheduled — particularly anyone running approval-based workflows.
Campaign status now auto-transitions through the full state machine — Scheduled → In Progress → In Review → Published — instead of only flipping to Published. The trigger also handles task creates and deletes, and reads/writes inside a Firestore transaction so a campaign's status stays consistent with the underlying tasks even under concurrent edits.
Why it matters: Campaign status used to lag behind reality. Even after every task in a campaign moved to In Review, the campaign itself stayed in earlier states until someone advanced it manually — making 'In Review' filters and dashboards unreliable.
Command Center polish for account managers
For Account managers and agency operators running Command Center across multiple client stores.
Command Center got a coordinated polish pass for account managers. The route is now gated to glowtify.com users, the default landing tab moved from Tasks to Calendar, calendar cards show the client logo next to the store name and surface a status badge without requiring hover, and store/status/assignee filters persist across sessions in localStorage. The Comments tab also got smarter: epics with unread comments float to the top, headers show the campaign as a subtitle, threads sort by latest activity, and only threads with new activity auto-expand.
Why it matters: Account managers running multiple client stores spent the start of every session re-applying the same filters, hovering over cards to read status, and digging through long comment lists to find the conversations that had moved since they last looked.
Catch publish-time media issues before they ship
For Anyone publishing posts cross-platform (especially Facebook + Instagram together) and LinkedIn carousel users who export PDFs.
Three publish-path improvements landed together: every social-post preview now shows an upfront warning when uploaded media exceeds the platform's size, duration or resolution limits (with the Facebook preview also checking Instagram limits when a post is going to both); the media-size error overlay was rebuilt as a slide-in/out alert with a typo fix in the dimension labels; and the LinkedIn carousel PDF export now preserves the source image quality instead of degrading it on export.
Why it matters: Media problems were being discovered too late — after hitting publish, or when a downloaded LinkedIn carousel PDF arrived with visibly fuzzy images. Even when the editor did surface an error, the dimension messages had a typo ('pxx' instead of 'px x') and popped in/out abruptly.
