Smarter Notion sync, safer marketing workflows, live quiz previews, and cash planning from real payments
Filter which Notion tasks sync to Hub, let agents edit tasks directly, test your quiz before it goes live, make marketing emails respect every stop signal, and drive cash budgets automatically from your payment register.
What's new in Hub
This update covers five areas across task management, marketing, lead capture, and finance.
Notion task sync: choose what comes in, map any field
Previously, connecting Hub to a Notion database imported everything — mixing personal to-dos with work tasks and carrying over every property whether you needed it or not. That changes with two new controls.
Import filters let you define up to 20 rules per database pairing. For example, you can tell Hub to only import tasks that have a specific status value, or that belong to a named project. Rules combine with AND; multiple selected values within one rule combine with OR. Tasks you create in Hub continue to sync back to Notion regardless of filters.
When a task no longer matches a filter, Hub can either keep the Hub copy but stop syncing it, or automatically archive it — whichever you prefer. If the Notion task later matches again, Hub restores the archived copy automatically.
Field mapping now covers every Notion property type that has a compatible Hub equivalent: rich text, select and status dropdowns, multi-select tags, numbers, URLs, email addresses, phone numbers, and calendar dates. If the Hub collection doesn't yet have a matching field, you can create one directly from the mapping settings — Hub generates the field with the right type and Notion's existing choices pre-loaded.
Unsupported types (formula, rollup, relation, files) remain visible in settings with an explanation, so you always know what isn't mapped and why.
Agents can now edit tasks, not just create them
Hub's Operations, Business (Dashboard), and Executive agents could already create tasks when given permission. They can now edit existing tasks too — updating the title, description, due date, priority, assignee, and completion status — all with the same permission controls that govern creation.
Task editing can be enabled separately from task creation, so you decide how much autonomy each agent has. The edit is version-checked: if a task changed between when the agent read it and when it tries to update it, the edit is rejected rather than overwriting your work.
You can now also apply task permissions to all your existing agents in one step — setting the shared collection and default assignee once rather than visiting each agent individually. Disabled agents stay disabled; new agents need their own explicit grant.
Agents cannot archive or delete tasks — that boundary is enforced at every level.
Marketing emails now respect every stop signal, including across delays
Workflow email actions now check each recipient's current marketing consent and stop signals individually at the point of delivery — including after a scheduled delay. This means that if someone books a meeting, makes a purchase, replies, unsubscribes, or withdraws consent between a workflow starting and a delayed step running, they won't receive the follow-up.
Stop decisions are recorded on the journey itself, so a recipient who was correctly stopped earlier in a sequence cannot be re-enrolled by a later delayed step.
Email bounce and complaint events now automatically create suppression records, and Hub's Outlook mailbox sync flags automated and bulk messages so they're distinguished from genuine human replies when evaluating whether to continue a sequence.
For forms, marketing consent is now an explicit checkbox destination in the form builder. A checked, visible consent box saves the contact's consent, timestamp, form source, and response reference atomically with the form submission. An unchecked or hidden checkbox never modifies an existing consent choice.
Test your quiz before it goes live
Quiz draft previews previously showed disabled radio buttons — you could see the questions but couldn't interact with them or check whether scoring worked correctly. The preview is now fully interactive.
From the quiz builder, you can now select answers, submit, and see the exact result band that a real respondent would see — including any content or call-to-action that appears in that band. A restart button lets you run through different answer combinations.
The preview is clearly labelled as a local draft test: no submission is saved, no contact is created, and no emails or workflows run. It's designed to help you verify that questions, scoring weights, and result copy work as intended before you publish.
Drive cash budgets automatically from your payment register
Hub's Banking and cash budgets now share a single payment register and editor. Whether you're looking at the banking ledger or a cash budget's Payments tab, you're working with the same payment schedules — no duplication.
From the banking ledger, you can now select one or more historical transactions and either:
- Create an upcoming payment — turning the historical transaction into evidence-backed future commitment, pre-filled with its amount, direction, and account.
- Match to an existing payment — allocating the transaction against an existing schedule occurrence, including partial allocations across recurring entries.
On the budget side, cash budget lines can opt into an automatic payment plan: instead of entering planned amounts manually, the line reads its figures from linked, reviewed payment occurrences. The approved baseline is preserved as a reference so you can see what changed. Settlements don't reduce planned amounts — they reduce outstanding cash.
The cash review workspace gains an accounting reference projection that bridges movements after a reference date, a downside projection that excludes receipts without confirmed confidence, and clearer labelling of unknown amounts and undated outflows.
Note: The payment register and automatic plan features are pending deployment. Deployment status will be confirmed separately.
Retire goals cleanly and browse complete check-in history
Goal owners can now retire a goal — closing it, stopping its check-in cadence, and removing its linked KPI alert rules — in a single, safe action. Previously, this operation could fail partway through if Hub attempted to delete check-ins before removing alerts. Retirement now happens in one server-side transaction: either everything updates together, or nothing changes.
Retiring a goal keeps all check-ins and KPI definitions intact. Only the alert rules linked to that specific goal are removed.
Check-in history is now paginated with proper page controls (Older, Newer, Latest), shows the author's name alongside each entry, and recovers gracefully from load failures — the current page stays visible if a subsequent page request fails, and a Retry button lets you try again without losing your draft.
More improvements are in progress. We'll keep you updated as they roll out.