200 variants. 5 test slots. Rank them with evidence, not vibes.
Updated
Example shape: each variant scored against the category's survival record and the audience's real language.
The short answer
When agents hand you 200 variants, you want them ranked against category evidence, so the test budget starts from the top of the deck. Adveron supplies the evidence to rank against: which angles have survived in the category's paid record, and which language the audience actually uses for the pain each variant names.
Every variant scored against which angles have survived in the category's paid history.
Variants that name the pain the way the audience does rise; the rest sink before a dollar moves.
The budget starts on the strongest candidates instead of a random draw of 200.
Generation stopped being the bottleneck; selection is. When creative volume comes from agents, the test slots are the scarce resource, and filling them by gut reading two hundred variants is where the leverage leaks out. The category's ad record and the audience's own language are both queryable, which makes ranking a scoring pass instead of a taste debate.
The pass works variant by variant: does the angle match territory with a survival record in the category, does the hook use phrasing the audience actually produces, and is the claim one the tracked record shows to be crowded or open? Variants that score on evidence go to the test slots first; the rest wait their turn instead of stealing budget. The same scoring runs inside an agent loop over MCP, so ranking scales with generation.
Example request and response
{"category": "your-category","angles_scored": 3,"top_angle": "identity-aspiration","top_angle_avg_days_running": 61}
Early access. 5,000 free credits, no card.
Get early access/v1/brands/{brand_id}/paid-media/adsThe category's survival record for each angle and territory./v1/audiences/{audience_id}/insightsThe audience's real language to score hooks against.MCP serverRun the scoring pass inside the agent loop that generates.No. The live test still decides; ranking decides what gets tested. Starting the budget on variants with category evidence behind them means the winners tend to surface in fewer rounds, and the slots stop being spent disproving obvious losers.
Two main reads: whether the variant's angle matches territory with a survival record in the tracked category ads, and whether its language matches phrasing the audience actually uses. Both are queries against sourced data, so the ranking is auditable when someone challenges it.
Adveron is opening to a first wave of teams. Ask for a key and we will open your workspace with one credential for REST and MCP. Usage-priced credits, no seat licenses.
Get early access