Competitor content intelligence

Business value

• Finds pages where competitors are vulnerable to higher-specificity content.
• Helps prioritize content investments by weakness, not just keyword volume.
• Creates a cheap quality map of a niche.

Agent job to be done

Act as a competitive analyst. Score competitor pages, extract evidence of genericness or weak provenance, and recommend content angles that would beat them on specificity.

format: articleintended_use: otherdomain: competitor content intelligence

When to call VeracityAPI

Run during keyword/content gap research, before assigning new pages or refreshes.

What text to submit

Competitor page main body, headings, comparison tables, recommendation copy, and source/citation sections. Respect robots/terms and do not submit private or paywalled content unless permitted.

Decision policy

• low competitor risk: page is probably strong; compete with unique data, angle, or authority.
• medium competitor risk: target specific sections with more concrete examples.
• high competitor risk: prioritize this URL/keyword as a gap if business value is high.
• store evidence spans as the why-beat-this-page rationale.

Automation recipe

• Crawler collects competitor URLs for target keywords.
• Extractor isolates main content.
• VeracityAPI scores each page.
• Agent clusters high-slop URLs by topic and missing detail type.
• Planner creates briefs emphasizing the competitor’s gaps.

Evidence spans agents should inspect

• vague advice without named examples
• thin product comparisons
• unsupported best/safest claims
• generic intros/conclusions that consume word count

Policy pseudocode

if (result.recommended_action === "allow") continueWorkflow();
if (result.recommended_action === "revise") rewriteWith(result.evidence, result.recommended_fixes);
if (result.recommended_action === "human_review") queueForHumanReview(result);
if (result.recommended_action === "reject") discardOrRebuild();

KPIs to track

• competitor URLs scored per dollar
• high-opportunity pages discovered
• content briefs generated
• rank gains against high-slop competitors
• editorial research hours saved

What can go wrong

• High slop risk does not guarantee a competitor ranks poorly today.
• Respect legal/ethical scraping boundaries.
• Use as prioritization signal alongside authority, backlinks, SERP intent, and business value.

Cost and latency notes

Analyze only is $0.005 per 1,000 characters; Analyze + revise with auto_revise=true is $0.010 per 1,000 characters. Both round up to the nearest 1,000 characters. Short captions/emails usually cost $0.005; longer pages or chapters scale linearly by length. Current v0.1 latency is LLM-bound, so batch/concurrent orchestration is recommended for high-volume pipelines.