Analyze competitor advantages and prior round outcomes with a frontier model to generate operator directives:
Enter ideas, specific angles, or content constraints for the Ideator model in the next 10 rounds:
Fork run
All fields are pre-filled from the source run. Edit any to fork. The source run is unchanged.
Snippet Optimization Mode (Default)
Manual content entry on suspected scrape fails. (Default)
Advanced Options
BYO Content
Paste target page text to override the cached body. Useful for paywalled, anti-bot, or heavy-JS pages.
Force refresh
Re-fetch from source instead of using the property cache. Each scope is independent.
Target scrape failed — paste page content to resume
The target URL couldn't be reached by either DataForSEO or the native scraper (paywall, anti-bot, heavy JS, etc.). Paste the page's text below and click Resume to retry from the start with this content. Existing partial state (items, claims, tests) will be wiped.
Some pages couldn't be scraped — paste content or skip
The run is paused because one or more competitor URLs couldn't be scraped. Paste each page's text below to include it in the optimisation, or leave a textarea blank to skip that URL. Click Resume when ready — the run continues with whatever you provide.
Manual edit — you are the ideator
Pick a Rank Factor (suggested or any other), edit the page below, optionally name the hypothesis, then Save & Test. The system substitutes your edit into the target's slot and ranks against the existing competitive set. Each test creates a new claim row in the ledger tagged with the chosen factor.
Loading suggestions…
Leading factors for this property + ranker
Descriptive aggregation across all classified claims for this property and ranker. Not a Bayesian pool — sums and bests only.
Loading…
Rank trajectory
Rank movement (all results)
Full-field ranking wasn't recorded for this run.
Calibration (stated prior vs realised mean)
Hover a point to see its claim here.
Claim ledger (sorted by posterior mean)
Competitive set
URLs in the SERP this run optimised against, in baseline-rank order. The target row is highlighted.
Ranker rationales
Per-round, per-item rationale the ranker emitted, grouped by domain. Every row from every round is shown — r0 is the unedited baseline — and target rows are included.
Competitor analysis
Press Analyze to compile a per-domain narrative from this run's ranker rationale. Costs one LLM call.
Page content
Full scraped body of each page (target + competitors where stored).
Loading…
Generated snippets
Per-URL extractive summary fed into the ranker. Target highlighted.
Claims (selectable)
All hypotheses tested in this run, sorted by posterior mean. Click a cell to select; text is fully copyable.
Last round detail
All rounds
Content brief
Rank Factors
This run's evidence split into positive vs negative. By rank = SERP-position movement vs baseline. By posterior = Bayesian belief shift (final μ vs initial prior). A mixed-direction claim appears under both tabs in different sub-tables — never duplicated within the same sub-table.
Run narrative
Edit Run Label
network_intelligence
Auxy
Add credits to continue
This runs on pay-as-you-go credits. Add credits to unlock it.
Billing for this project is handled by someone else
