June 15, 2026
Paid analytics that break down what 'Results' really means, a rebuilt Google Reviews, and playbook deploys that fix themselves
This release makes your paid analytics genuinely answer the question, rebuilds Google Reviews from the numbers up, and teaches playbook deployment to recover from failures on its own.
Paid Analytics That Answer the Question
Meta rolls every outcome into one blended "Results" number, so five leads and two hundred reach impressions show up as two hundred and five results, which tells you nothing. Now the Results card breaks that number into its real types.
What's new:
- The Results card splits the blended total into its real types (Leads, Messages, Calls, Reach, Purchases, and more), shown as a donut for three types or fewer and a bar chart beyond that.
- The Trends chart adds a button per result type so you can plot any one over time with its own true cost and rate.
- The Workspace Totals table labels each location's Results with what it actually produced, like "200 Leads, 223 Link Clicks."
- A new Compare to dropdown shows the percent change against a prior period, month, quarter, or year, with directional color and a low-sample badge on short windows.
- Every workspace now carries a FIRE score: a single 0 to 100 health number that compares the selected period to the workspace's own history and its brand peers, so a small location and a flagship are each judged against themselves.
Where to find it: Paid Social Ads > Analytics.
Google Reviews, Rebuilt
Google Reviews got a top-to-bottom rework this release, starting with the numbers themselves.
What's new:
- The KPI cards (Average Rating, Total Reviews, Response Rate, Average Response Time, Reviews per Week) now show all-time lifetime totals instead of recomputing against the selected date range. The trend chart still respects the date picker.
- Bulk reply lets you select individual reviews, a whole page, or every review matching your current filters, generate an AI template response, edit any draft, and post safely through a rate-limited background queue that survives a closed tab.
- A data-quality fix scopes the hourly review sync to each workspace's claimed listings, and a cleanup migration removed over seven thousand duplicate review copies across affected multi-location hubs without losing a single distinct review.
Where to find it: Google Reviews.
Deploys That Fix Themselves
A failed playbook launch used to be fire-and-forget. Now it recovers on its own.
What's new:
- A failed deploy schedules its own retry on an escalating backoff: thirty minutes, then one, two, four, and eight hours.
- After five failed attempts the system escalates once, emailing the playbook author and the person who opted in.
- Permanent failures like payment problems or revoked access skip the retry queue and escalate immediately.
- A daily digest lists every failed or stuck opt-in grouped by hub so nothing sits unnoticed.
Where to find it: Paid Social Ads > Playbooks deploy flows.
Disconnecting an Account No Longer Loses Your Posts
A temporary token expiration or an accidental disconnect should not cost you a week of planned content.
What's new:
- Disconnecting a social account now pauses and hides your future scheduled posts instead of deleting them.
- Reconnect the account and future-dated posts are automatically restored and re-queued.
- Posts whose scheduled time has already passed are dropped cleanly.
Where to find it: Connected Accounts and the content calendar.
Smaller Fixes
- Google Business posts that failed with "location not accessible" are unblocked, and the media upload step retries automatically on a transient Google image-fetch failure instead of failing the whole post.
- Another performance sweep sped up the hub ad accounts endpoint, ad creative extraction, the connected accounts page, and media thumbnails.
- Locations endpoints now enforce workspace management scope checks, closing a permissions gap for multi-tenant access.
- Expired LinkedIn tokens warn once and stop instead of looping as retryable analytics failures.
This release was about making the data surfaces you rely on genuinely trustworthy, and recovering your work when something fails.