Skip to content
Founder-led internal case — not client work.

Fetch Ads Control Room

How we mapped and built a paused-first operating system for website handoffs, Google Ads governance, lead evidence and controlled monitoring across two brands.

Verified internal deliveryEvidence is operational, not a client or ROI claim.
  • 426 automated tests
  • Paused publication
  • Zero unintended spend
Evidence boundary

Fetch is an oOMF! internal platform. This page documents our own workflow, safeguards and test evidence. It is not an independent client endorsement, a paid-media outcome or a promise of commercial performance.

Redacted public Opportunity Map

The problem was not one ad setting. It was the hand-off between systems.

Website releases, campaign configuration, conversion governance, lead records and monitoring evidence lived in different places. A safe launch needed one bounded path from approved offer to verified lead and an immediate way to stop serving.

01 · Trigger

A campaign is ready to publish

The signed plan names the brand, account, landing page, audience, budget, match types, exclusions and activation gate.

02 · Problem points

State could drift between tools

A campaign could look paused in one report while live evidence was stale, or a valid enquiry could lose the click context needed for later quality review.

03 · First improvement

Make paused publication the default

Publish the exact campaign package paused, verify every child resource, and require a separate activation token before spend is possible.

04 · Human decision

Quality and value stay reviewed

People classify valid, qualified, paid, proposal and won stages. Raw submissions remain secondary and collected gross profit is reconciled before value is used.

Implemented workflow

Plan, validate, publish paused, verify and monitor.

Plan

Signed brand package

Keep account, site, campaign, conversion and evidence paths separated by brand.

Check

Validate-only operations

Check policy, URLs, networks, match types, budgets, assets and conversion governance before publication.

Hold

Paused resources

Publish campaign, ad group, keywords and ads without allowing any serving or spend.

Watch

Repeatable checkpoints

Review live state, URL health, attribution, search terms and changes at bounded checkpoints.

Safeguards and fallback

The controls are designed to fail closed.

Search-only boundaries

No Display, Search Partners, broad match, Performance Max, AI expansion or final-URL expansion.

Evidence minimisation

Fetch evidence excludes contact details, free-text messages and raw click identifiers. Cross-brand transfer requires explicit authority.

Composite activation

If a child resource cannot be enabled and verified, the campaign is returned to paused instead of being left partially live.

Manual fallback

Campaigns can be paused directly in Google Ads, while lead handling and quality review continue without automated bidding.

Verified outcomes

Delivery evidence before performance claims.

At the recorded release checkpoint, 426 automated tests passed. A 128-operation Google Ads package was published in controlled paused state, with zero campaigns enabled, zero unintended spend and no offline conversion uploads. The lifecycle and monitor paths are repeatable. These are system-delivery outcomes, not evidence of lead volume or return on ad spend.

Use the same decision method

Map one workflow for A$1,250 ex GST.

An AI Opportunity Map covers one defined workflow, its problem points, practical options, safeguards and a written first-project boundary. Implementation and additional workflows are separate.