Skip to main content

Approval queues

Why this matters for your business

For agencies and large brands, automation power is a liability without governance. A junior marketer's well-intentioned "auto-discount returning customers" rule could give away hundreds of thousands in margin. A new client's misconfigured auto-bid rule could blow out a quarter's ad budget. Even the right rules can fire on the wrong customers when conditions drift.

Approval queues add a human checkpoint. For configured action types, every fire pauses at a queue — a designated approver sees the proposed action with full context (what customer, what data, what the rule says, what it would cost) and approves or denies in one click. Only approved fires execute.

The trade-off is latency: approval-gated rules don't fire in real time. The benefit is governance: nothing happens that a trained human didn't agree to.

What this typically unlocks

OutcomeResult
High-stakes mistakes per quarter0 with approval gates active
Junior team member autonomy with safety nethigh — they can build, approval gate catches mistakes
Audit-grade "who approved what?" trailcomplete
Multi-stakeholder sign-off (compliance, legal)possible without out-of-band tooling

What you actually get

CapabilityDescription
Per-action-category gatesConfigure approval requirements per action type (paid spend, mass tags, price changes, etc.)
Multi-reviewer chainsRequire 1, 2, or N approvers; first-to-approve OR all-must-approve
Approver groups"Marketing leads", "Legal", etc. — group-based assignment
Time-out behaviourDefault behaviour after N hours unanswered (auto-deny / auto-approve / escalate)
Audit trailEvery approval decision logged with reviewer, timestamp, reasoning
Bulk approval UIApprove 50 fires at once for similar items (with cap)
Mobile-friendly approval UXApprove from phone notifications
Slack integrationApproval requests post to designated channel

Configuration

Per-action approval matrix (example)

Action typeApproval required?ApproversTime-outAfter time-out
Paid ad bid changesYesAny of: marketing-lead24hAuto-deny
Mass-tag (> 1K customers)YesAll of: ops-lead, marketing-lead48hEscalate
Price changes (catalog)YesAll of: pricing-lead, finance72hAuto-deny
Send transactional emailNo
Tag single customerNo
Customer-facing copy changeYesAny of: brand-lead24hEscalate

You set this matrix once per shop; rules respect it automatically.

How it works

Real merchant scenarios

Scenario A — Agency client safety

Agency has a junior marketer building rules for client X. Approval matrix requires marketing-lead sign-off for any rule that touches paid spend.

Junior builds: "Auto-bid up 30% on top-LTV decile." Saves the rule live. First fire: queued for approval.

Marketing lead reviews preview: "$200/day spend impact, 850 customers affected. Rule firing weekly per LTV refresh." Lead sees the same runaway pattern that dry-run would have caught, denies the fire with note "rule fires weekly — needs first-time guard."

Junior fixes the rule. Re-submits. Lead approves. No money spent on a buggy rule.

Scenario B — Compliance gate for regulated brand

Pharma brand under regulatory oversight. Any customer-facing copy change requires compliance review.

Rule: "Send post-purchase email with usage tips." Rule fires on every order — but the action requires compliance approval the first time a new email template is used.

First fire: queued. Compliance reviews template, approves. Subsequent fires of same template auto-execute (template hash unchanged).

When the brand updates the template copy: hash changes; next fire re-queues for compliance. Always-current sign-off without slowing every send.

Scenario C — Time-out policies in action

Brand sets: "Mass-tag actions auto-deny after 48h unanswered."

Rule: "Tag VIP customers when LTV crosses threshold." Triggered overnight while approver on holiday. Queue fills. At 48h mark: 12 fires auto-denied with reason approval_timeout.

Approver returns: sees the denials in failures dashboard, understands what happened, manually approves backlog. No silent "did this happen?" mystery.

Best practices

Configure timeout policies explicitly. "What happens at midnight when nobody's around?" should have a documented answer.

Use Slack integration for high-volume queues. In-app queues miss things; Slack pings get attention.

Group approvers, don't list individuals. "Marketing-lead group" stays current as people join/leave.

Bulk-approve carefully. Bulk-approve 50 fires that look the same is fine; bulk-approving without inspecting is what approval queues are designed to prevent.

Don't require approval on everything. Approval friction slows real-time workflows; reserve gates for genuinely high- stakes actions.

Don't let queues grow. A 200-fire backlog means people will start rubber-stamping. Tune the rule, the matrix, or the approver capacity.

Plan tiers

CapabilityFreeStarterProAgencyEnterprise
Approval queues
Per-action matrix
Multi-reviewer (any-of / all-of)
Approver groups
Time-out policies
Slack integration
Mobile approval UX
Bulk approval UI
Approval analytics (avg time, denial rate)
Custom approval webhook

See also