Discovery and qualification before provisioning
Every order begins with a Telegram or WhatsApp discovery covering vertical declaration, target geography, currency of spend, intended monthly volume, conversion event the buyer will optimise toward, and landing page URL. The reason we ask is not gatekeeping. It is operational. A hosted account configured for the wrong currency cannot be reconfigured later without resetting the trust signal. A hosted account pointed at a landing page that fails domain reputation review will damage the container on first impression. Five minutes of discovery prevents weeks of remediation.
Landing page and creative pre-flight review
Before the account is provisioned, our verification team runs the buyer's landing page and any planned creative through a pre-flight checklist. The checklist covers HTTPS configuration, privacy policy presence, terms of service presence, cookie consent compliance for the geography, content match between ad and landing page, claims that require local financial regulator disclosure, claims that require local healthcare regulator disclosure, and the absence of redirect chains that platforms read as cloaking. If anything fails, we surface it during the discovery call rather than after the first disapproval. Buyers occasionally find this annoying. The container survives because we do it.
Provisioning and white glove handover
Provisioning happens inside our container. The buyer's brand name is set on the account, the billing currency is configured, the timezone is set to the buyer's primary operating timezone, vertical whitelisting is applied where required, and conversion tracking is configured to point at the buyer's first party server or measurement endpoint. The handover call covers admin access, billing instrument attachment, the first campaign build, and a walkthrough of the support escalation path. Median handover call duration: 30 minutes. The buyer leaves the call ready to launch the same day.
Ongoing health monitoring
Every hosted account is enrolled in our health monitoring loop. The loop checks daily for billing settlement health, policy review queue depth, sudden delivery throttle (a leading indicator of trust score change), conversion tracking drift, and creative disapproval rate. Alerts are sent to the responsible operator on the buyer side and to our duty engineer simultaneously. The buyer is rarely the first to know that something has changed in the platform. We are.
Suspension response and replacement workflow
When a suspension occurs, the workflow is fixed and time-bounded. The account is read from the platform side within minutes of the suspension event because our duty engineer monitors the queue. The buyer is notified within the hour with the suspension reason as the platform stated it, our internal classification of whether the suspension is in-scope for replacement, and the estimated replacement timeline. In-scope suspensions are replaced operationally within 24 hours on standard inventory and within 48 to 72 hours on custom configurations. Out-of-scope suspensions (for example, a buyer running creative the pre-flight check rejected and the buyer overrode) are explained in writing with a path to remediation if the buyer wants to continue.
Quarterly container hygiene
The container itself is audited quarterly. Aged accounts that have not been used in 90 days are recycled to fresh buyers. Accounts that have accumulated policy strikes are retired. New accounts are added at a steady cadence to maintain age distribution across the container. The buyer never sees this work, but the buyer benefits from it: the container the buyer joins today carries the same trust signal as the container the buyer would have joined two years ago, because we have been maintaining it the entire time.
Why the container model wins over the account farm model
The competing model in the agency hosted account market is what operators call the account farm: thousands of disposable accounts created on rented identities, run hot for a few weeks, and discarded. The farm model is cheaper to provision and faster to scale, but it carries no trust signal because no individual account has any meaningful operating history, and it carries no whitelisting because no platform issues vertical approvals to disposable identities. The buyer who orders from a farm receives an account that looks identical to a hosted account on the surface and behaves nothing like one once it tries to spend at scale. The container model is more expensive to maintain because every account is real, every container has a real partner relationship, and every quarter of audit work is real labour. Buyers who have run on both models call us back. The reason is simple. Day one performance looks similar on both. Day thirty performance does not, and day ninety is the difference between a working business and a stalled one.