Adding acquisition controls after a static release
Publishing was only the first boundary. I added site-level controls for acquisition tactics, budgets, experiments, delivery records, and crawl/index observations so each release can be operated and compared instead of merely launched.
24 versioned techniques · immutable experiment records · IndexNow and XML-RPC automated
A successful build is not a traffic result
The renderer can produce a valid static release, but that says nothing about whether search engines discover it, whether a distribution attempt ran, or whether either action changed the outcome. Those states need separate records.
I moved post-release work into the same operator system while keeping it outside the publishing gate. A site can be live, paused for acquisition, or running one bounded experiment without those states being confused.
Controls come before automation
Each site opts into a versioned technique definition with its own configuration, budget, pause state, and kill history. Estimates, model advice, experiment assignments, exposures, outcomes, delivery attempts, placements, and URL observations are retained as separate evidence.
- Per-site opt-in, configuration, budget, pause, and kill controls
- Immutable estimates and controlled experiment assignments
- Separate artifacts, executions, placements, exposures, and outcomes
- URL-level crawl and index observations alongside traffic comparisons
Only two transports are automatic
The catalogue currently contains 24 versioned techniques. Only IndexNow and XML-RPC delivery to Ping-O-Matic and Twingly have automated transports. The other 22 remain manual or proposal-only until they have a real execution path and evidence.
That limit matters. A catalogue of possible tactics is not an autonomous growth engine, and an automated submission is not evidence of indexing, traffic, or commercial value.
What the implementation proves
- Eight selectable public renderer architectures
- Validated 50,000-page streaming renderer
- Versioned technique, estimate, experiment, exposure, outcome, and observation records
- Dedicated Rust distribution worker for implemented transports
Limits: The 50,000-page result proves renderer capacity, not 50,000 AI-authored pages. · Twenty-two techniques are not automated. · No ranking, traffic, lead, or revenue result is attached to the acquisition layer.