com-a-002record 1
{
"rank": 1,
"id": "com-a-002",
"identity": "Cloobot X",
"url": "https://www.cloobot.ai/",
"segment": "ai-discovery",
"observed": "2026-08-27",
"evidence_class": "D",
"verified": "fetched",
"denominator": 55,
"selection_reason": "Coverage-gap analysis + requirement-to-source traceability: the single most transferable discovery mechanism found",
"mechanism": "Ingests multiple unstructured sources (workshop notes, SoWs, stakeholder conversations, project docs) with live intent capture, and links every derived requirement back to its originating source context.",
"limitations": "The '80-90% accuracy' converting SoWs and workshop notes into requirements is a customer testimonial on the vendor's own page, not a measured benchmark. No public detail on how contradictions between two sources are surfaced or adjudicated, nor on what happens when coverage analysis finds a gap.",
"license_claimed": null,
"disposition": "top10_promoted"
}
com-a-003record 2
{
"rank": 2,
"id": "com-a-003",
"identity": "Fin for Sales (Intercom)",
"url": "https://fin.ai/sales/",
"segment": "ai-discovery",
"observed": "2026-08-27",
"evidence_class": "D",
"verified": "fetched",
"denominator": 55,
"selection_reason": "Only surface in 55 with explicit configurable terminal states — directly answers P01's stopping-rule question",
"mechanism": "Conversational agent on the website asks the same discovery questions a human SDR would, about needs, budget and timeline; simultaneously enriches the conversation with CRM data and real-time research on the visitor's company.",
"limitations": "Public docs do not state how many questions it asks, in what order, or whether ordering adapts to answers. No published accuracy for qualification decisions or false-disqualification rate. Confirmed gap: no automated email follow-up sequences after qualification.",
"license_claimed": null,
"disposition": "top10_promoted"
}
com-a-013record 3
{
"rank": 3,
"id": "com-a-013",
"identity": "Demandbase",
"url": "https://www.demandbase.com/products/data/",
"segment": "enrichment",
"observed": "2026-08-27",
"evidence_class": "D",
"verified": "fetched",
"denominator": 55,
"selection_reason": "Sharpest precompute/ask boundary: Toggle Fields shown only on enrichment failure (per-field fallback as confidence mechanism)",
"mechanism": "Anonymous visitor de-anonymisation into known accounts, plus Forms Enrichment which supplements a form submission with known company data — populating hidden fields so firmographics are captured without displaying them, and Toggle Fields that stay hidden unless enrichment fails.",
"limitations": "The product page makes no first-party claim about forms enrichment and does not name specific appended fields or the identification methodology — the Toggle Fields and hidden-field mechanics come from Demandbase's support documentation surfaced in search, not from the fetched page. Fill rates are not published.",
"license_claimed": null,
"disposition": "top10_promoted"
}
com-a-001record 4
{
"rank": 4,
"id": "com-a-001",
"identity": "Auctor",
"url": "https://www.getauctor.com/",
"segment": "ai-discovery",
"observed": "2026-08-27",
"evidence_class": "D",
"verified": "fetched",
"denominator": 55,
"selection_reason": "Closest structural analogue to Actionist: agent sits inside the human discovery call rather than replacing it",
"mechanism": "Agent attends live discovery/refinement calls and ingests institutional knowledge and workshop inputs; unstructured conversation is transformed into structured requirements and kept synced across downstream tools as requirements change.",
"limitations": "getauctor.com returned 403 to fetch; verification is via YC's first-party company profile, not the vendor site. Nothing public establishes extraction accuracy, how contradictions between sessions are resolved, or whether the agent ever asks a clarifying question. The '70% of enterprise implementations experience significant delays' figure is a vendor launch-post claim, unsourced.",
"license_claimed": null,
"disposition": "top10_promoted"
}
com-a-004record 5
{
"rank": 5,
"id": "com-a-004",
"identity": "Avoma",
"url": "https://www.avoma.com/",
"segment": "conversation-intelligence",
"observed": "2026-08-27",
"evidence_class": "D",
"verified": "fetched",
"denominator": 55,
"selection_reason": "Named schema (MEDDIC/BANT) populated by inference from conversation — the exact shape P01 needs; publishes no citations, which is the gap",
"mechanism": "Auto-records and transcribes across conferencing tools with speaker identification in 75+ languages; custom Smart Trackers detect keywords and topics by semantic matching; AI-generated Smart Chapters segment the call by topic.",
"limitations": "The vendor page markets outcomes (4+ hrs/week saved, 40% win-rate lift) without describing the underlying models or, critically, how an extracted MEDDIC field ties back to specific transcript evidence. Whether a filled field carries a citation or a confidence score is not established publicly.",
"license_claimed": null,
"disposition": "top10_promoted"
}
com-a-007record 6
{
"rank": 6,
"id": "com-a-007",
"identity": "Loopio",
"url": "https://loopio.com/",
"segment": "intake",
"observed": "2026-08-27",
"evidence_class": "D",
"verified": "search-only",
"denominator": 55,
"selection_reason": "Per-answer confidence scoring plus an explicit model of content decay over time (search-only; needs first-party pass)",
"mechanism": "Inverted discovery: the counterparty asks the questions, and Loopio matches each incoming question against a library of pre-approved answers, importing questionnaires directly from procurement portals.",
"limitations": "Fetch of loopio.com was not completed (model rate limit), so confidence-indicator specifics rest on Loopio's own blog content surfaced via search rather than a fetched product page. The scoring scale, thresholds, and what a low score actually blocks are not established. '1,700 companies' and '61% achieve ROI within a year' are vendor claims from Loopio's own benchmark report.",
"license_claimed": null,
"disposition": "top10_promoted"
}
com-a-047record 7
{
"rank": 7,
"id": "com-a-047",
"identity": "SellScale / Selix",
"url": "https://www.sellscale.com/",
"segment": "ai-discovery",
"observed": "2026-08-27",
"evidence_class": "D",
"verified": "search-only",
"denominator": 55,
"selection_reason": "Narrated reasoning in a channel humans already read, plus pre-execution human approval gate",
"mechanism": "Agentic research with a narrated, inspectable trace surfaced into Slack, and a human approval gate before campaigns launch.",
"limitations": "Not fetched; the Slack narration and approval-gate behaviour is described in SellScale's own published field guide, so it is a first-party claim reached through search rather than a fetched product page. No evidence on how often proposals are rejected.",
"license_claimed": null,
"disposition": "top10_promoted"
}
com-a-005record 8
{
"rank": 8,
"id": "com-a-005",
"identity": "Gong",
"url": "https://www.gong.io/",
"segment": "conversation-intelligence",
"observed": "2026-08-27",
"evidence_class": "D",
"verified": "fetched",
"denominator": 55,
"selection_reason": "Category leader retained as a methodological caution: most famous, least mechanism-bearing first-party evidence",
"mechanism": "Automatic capture of calls, emails and meetings across integrated channels; AI analysis over the accumulated interaction corpus, with keyword/topic trackers and methodology scoring layered on top.",
"limitations": "The main product page is notably thin on mechanism: it does not name trackers, next-step extraction, or evidence-structured discovery, and the help-centre tracker doc 404'd. Pricing is unpublished; the $100k-200k/yr for 20-50 seats figure is third-party. Gong's own page establishes marketing positioning far more than it establishes how discovery is structured into evidence.",
"license_claimed": null,
"disposition": "top10_promoted"
}
com-a-031record 9
{
"rank": 9,
"id": "com-a-031",
"identity": "Financial Cents",
"url": "https://financial-cents.com/",
"segment": "intake",
"observed": "2026-08-27",
"evidence_class": "D",
"verified": "search-only",
"denominator": 55,
"selection_reason": "Accounting-vertical intake with prior-year pre-population then confirmation — industry-specific precompute precedent",
"mechanism": "Workflow templates emitting client requests, with automated follow-up on outstanding documents; integrates with Ignition so a signed proposal auto-creates and assigns the project.",
"limitations": "Not fetched directly; the Ignition integration mechanics are documented first-party by Ignition and by trade press (CPA Practice Advisor, Aug 2025) and are US/Canada only. Karbon's comparison page calling it feature-shallow is a competitor's assessment and should not be treated as fact.",
"license_claimed": null,
"disposition": "top10_promoted"
}
com-a-055record 10
{
"rank": 10,
"id": "com-a-055",
"identity": "Outfindo",
"url": "https://outfindo.com/",
"segment": "guided-selling",
"observed": "2026-08-27",
"evidence_class": "I",
"verified": "search-only",
"denominator": 55,
"selection_reason": "Guided-selling needs-analysis retained as the strongest of a thin 6-row segment",
"mechanism": "Question-based product guides that translate buyer language into catalogue filters.",
"limitations": "Not fetched. Pricing (99 EUR/mo Essential, 599 EUR/mo Plus) is third-party. Small vendor; question-policy depth unestablished.",
"license_claimed": null,
"disposition": "top10_promoted"
}