Competitor Displacement Intelligence

Status: Apify Store Actor — REST + MCP

Find accounts becoming winnable because a competitor’s technology is disappearing or changing. This single Actor exposes REST and MCP surfaces over the same engine. It separates observable public-web evidence from inference; it never claims confirmed purchase intent, a confirmed buyer, or internal software usage.

Surfaces

REST endpoints:

• POST /v1/scan — observe one current stack without changing the baseline.
• POST /v1/change — compare and persist one observation.
• POST /v1/displacement/search — scan up to 1,000 domains and filter competitor/replacement matches.
• GET /v1/history/{domain} — return the latest 20 stored observations.

MCP uses the same engine over Streamable HTTP at /mcp with stateless_http=true: scan_technology_stack, detect_technology_change, find_competitor_displacement, and get_technology_history.

Example displacement request:

{"competitors":["Salesforce"],"replacement_candidates":["HubSpot","Pipedrive"],"domains":["company-a.com"],"min_evidence_score":70}

V1 interpretation

Signals: NO_CHANGE, OBSERVED_ADDED, OBSERVED_REMOVED, LIKELY_REPLACEMENT, INSUFFICIENT_EVIDENCE.

Reason codes are machine-readable: competitor_removed, competitor_added, same_category_replacement_detected, scan_quality_comparable, scan_quality_insufficient, change_confirmed, insufficient_evidence, and no_material_change.

LIKELY_REPLACEMENT requires reliable removal and addition, shared category, adjacent observations, and comparable scan quality. A degraded scan exposes provisional_removed and cannot produce a removal or replacement claim. Evidence score is not purchase probability.

The detector runs Opentechalyzer built-in MIT fingerprints only, with --no-external; no enrichment, intent data, webhooks, Wappalyzer database, or HTTP Archive is used. Snapshots and bounded history use the named Apify Key-Value Store technology-change-buying-signals-v1.

Distribution and pricing

This is one Apify Actor with REST and stateless Streamable HTTP MCP. PAY_PER_EVENT pricing is declared in .actor/pay_per_event.json: technology scan $0.01/domain, technology change $0.02/domain, and displacement analysis $0.03/domain, with displacement analysis as the primary event.

Limitations and release gate

Results describe observable public-web technology evidence only. 403/blocked pages can have low evidence quality; a timeout is a failed scan. An evidence score is not purchase probability, and no result confirms purchase intent or that a company bought a product. No RapidAPI or ApyHub publication is part of this product.