How MagikBull works for your team

MagikBull is built around workspaces: campaigns, assets, audit trails, and integrations are scoped to the workspace you select. Team members only see workspaces they are invited to — there is no cross-tenant browsing of customer ad accounts.

OAuth connections to Meta and Google are used only to perform the actions you enable in product (read structure, publish where you approve, sync metrics you connect). Tokens are stored with platform-grade controls; rotate or disconnect anytime from workspace settings.

Owners and admins can remove a member from the workspace immediately. Removed users lose access to the workspace on next request; pending approvals they owned can be reassigned according to your approval templates.

Change your password and revoke active sessions from your identity provider or MagikBull security settings. Disconnect ad platforms temporarily, review the audit feed for unexpected AI actions, and contact support if you need a full workspace lockdown.

Operational data lives in our managed cloud stack with encryption in transit and at rest. Exact regions may evolve with enterprise agreements; your order form or DPA is the contract source of truth for regulated customers.

Subscriptions are billed in advance per billing period (monthly or yearly where available). AI usage draws from your included credit pool; optional usage-on-demand can extend capacity within caps you configure. Checkout (e.g. Razorpay) is the source of truth for tax, currency, and final totals.

You can move between catalog tiers when your billing flow supports it. Upgrades typically take effect quickly so you get a higher credit allowance; downgrades may apply at the next renewal depending on processor rules. When in doubt, check the confirmation screen at checkout.

When a coupon campaign is active for your workspace, eligible checkouts apply it automatically per campaign rules (for example first-time subscriber offers). Not every tier or region may be eligible — the cart will tell you before you pay.

Payment confirmations and tax documents are issued by your payment provider and mirrored in workspace billing views where wired. Keep billing contacts up to date so finance receives every renewal notice.

Cancel from workspace billing settings or through your subscription management link. You retain access through the end of the paid period unless otherwise stated. Export anything you need before offboarding — support can help with timelines on enterprise agreements.

High-impact AI proposals (copy changes, publish steps, bulk edits) land in a review queue instead of going live immediately. Approvers you designate can approve, reject, or request revisions with a clear audit trail.

Workspace admins configure who holds approver roles and can layer approval hierarchy templates for larger teams. Default-safe setups start with manual approval until you explicitly widen autopilot scope.

Nothing ships. The proposal is marked rejected, authors see your reason when provided, and you can regenerate or edit manually. Rejections never debit publish-side actions that did not occur.

Yes — by design when you want governance. Set clear approver coverage, use templates for recurring campaign types, and tune autopilot rules for low-risk automations so humans only review edge cases.

Autopilot lets MagikBull execute approved classes of AI actions on your behalf within boundaries you set — for example refreshing creative variants or adjusting bids only after policy checks. It is not a black box: scope, caps, and kill switches stay visible in workspace AI control settings.

Use workspace-wide pause controls in AI control to stop pending autopilot runs and reject queued actions. For severe incidents, owners can trigger emergency stop flows so autopilot and risky publishes halt until you explicitly resume.

Manual mode requires explicit human approval per sensitive step. Autopilot automates the mechanical steps you have pre-authorized while still respecting credit balances, platform policies, and any hard caps you configure.

Yes. AI action history and audit feeds list who or what initiated each step, timestamps, and outcomes. Use these exports for internal compliance reviews or client reporting.

Each AI action (research, copy, optimization proposals, etc.) debits from your subscription balance via the ledger. When usage-on-demand is off, actions stop if credits are exhausted.

You can allow limited overages beyond included credits, bounded by your on-demand cap and a simple enable flag. Ongoing overage stays visible on the subscription so finance can reconcile.

Yearly prepay reflects roughly two months free versus paying month-to-month. Credit renewal targets stay aligned to the catalog; confirm annual SKUs in checkout when enabled.

When a coupon is active for your workspace, checkout applies it according to campaign rules (e.g., first-time subscriber offers). Not all tiers or regions may be eligible.

Each billable AI operation debits from your workspace subscription balance. The ledger shows running balance, renewals, and on-demand overage when enabled. If usage-on-demand is off, actions that require credits will stop when the pool hits zero.

Rollover behavior follows your plan and any promotional terms shown at purchase. The billing summary in-product reflects the current period allowance — if rollover applies, it will be stated there; otherwise assume use-it-or-lose-it within the renewal window.

Premium intelligence routes may consume credits faster than baseline routes because they use larger context windows or higher-quality model tiers. MagikBull surfaces routing choices so teams can balance quality vs. burn rate.

Use separate child workspaces per client or brand, each with its own integrations and approvals. Agency plans can pool AI credits across those workspaces where enabled so finance sees one consolidated burn while client data stays isolated.

Yes — brand knowledge and creative libraries can be partitioned inside a workspace when your workflow demands it. For strict client separation, prefer separate workspaces so reporting and access control stay clean.

Reach out via the contact page with your seat count, regions, and compliance needs. Enterprise agreements can include custom coupon programs, negotiated credit pools, and tailored governance defaults.

You connect via Meta’s OAuth flow, pick the ad accounts MagikBull may access, and map them to campaigns inside MagikBull. Publishing and sync respect Meta’s API limits; if a token expires, reconnect from settings to restore sync.

Google OAuth grants access to the customers and accounts you authorize. Structural reads and publishes flow through MagikBull’s integration layer; Google Ads remains the billing and policy authority for your ads.

Yes — map only the accounts and entities you want managed. Narrow scopes reduce risk and make approvals easier because reviewers see a bounded set of entities.

Platforms batch metrics on their own cadence. MagikBull refreshes on a schedule and on demand after major edits; sub-hour discrepancies versus native UIs are normal during peak API windows.