Link Pilot
Link Pilot is a standalone Shopify app for internal linking—separate from the EShopSet bundle. These guides cover scanning the store, how AI embeddings and rules generate suggestions, how Overview Score and suggestion Confidence differ, working the Suggestions queue, applying links, Opportunities, Settings, Analytics, and billing.
Start here: How AI and rules generate suggestions · How suggestion confidence is scored · How the Overview linking score works.
Product overview: Link Pilot app page. Playbooks: Internal linking use cases.
What is Link Pilot?
Link Pilot is a standalone Shopify app for internal linking: audit orphans and broken links, approve AI suggestions, and apply links with credits or DIY export.
Scan the store and generate suggestions
Run Link Pilot onboarding scan to index Shopify products, collections, pages, and blogs, then audit and generate internal link suggestions.
Read Overview link health
Use Link Pilot Overview to review orphans, broken links, underlinked hubs, and the in/out content graph for Shopify body HTML.
Review, edit, and approve suggestions
Work the Link Pilot Suggestions queue: filter by type and confidence, edit anchors and destinations, approve or dismiss internal link recommendations.
Apply links with credits or DIY export
Apply approved Link Pilot suggestions with implement credits, or DIY copy/CSV export. Safe wrap of existing phrases in Shopify body HTML.
Map Search Console keywords as opportunities
Import Google Search Console Queries.csv or add keywords manually in Link Pilot Opportunities to generate keyword-target internal link suggestions.
Configure limits, filters, notifications, and automation
Set Link Pilot max links per page, ignore handles, notification digests, and Pro auto-implement confidence thresholds.
Analytics, credits, and Link Automation billing
Understand Link Pilot implement credits, one-time packs, Link Automation Starter Growth Pro, and Analytics CTR for applied internal links.
How AI and rules generate suggestions
Link Pilot combines rule-based audits with AI embeddings: cosine retrieval, weighted confidence, then an AI-written anchor and Why explanation for merchant review.
How suggestion confidence is scored
Link Pilot Confidence is a 0–100% weighted score: 35% semantic cosine, plus context, hub importance, commercial value, scarcity, and title mention—not an AI guess.
How the Overview linking score works
Link Pilot Overview Score is 0–100 graph health: 35% orphans, 30% broken links, 20% underlinked hubs, 15% dead-end articles and pages. Body HTML only.
Suggestion types and when to use them
Link Pilot suggestion types: broken target, orphan inbound, underlinked hub, missing commercial link, semantic match, keyword target, and content gap.
