Funnel measurement breaks when it's done piecemeal — marketing measures MQLs, sales measures pipeline, finance measures revenue, and nobody sees the connection. The agent builds a unified funnel model that traces every opportunity from first touch through every stage to closed-won or closed-lost, regardless of how many channels, campaigns, or touches were involved. Conversion rates are calculated at every stage transition — not just lead-to-opportunity or opportunity-to-close, but every micro-conversion in between. Stage velocity is tracked alongside conversion rates, so a bottleneck that's dragging deal length is visible even when conversion rate looks healthy.

Last-touch attribution is easy to implement and almost always wrong. A deal that took six months and a dozen touches doesn't owe its existence to the last LinkedIn ad the champion clicked. The agent supports configurable multi-touch attribution models — linear, time-decay, W-shaped, and custom weighting — applied consistently across all channel data connected to Lantern. Revenue credit is distributed across every tracked touchpoint in the path, giving a defensible answer to which channels and campaigns contributed to closed deals rather than which ones happened to touch them last. Attribution data feeds directly into the budget allocation and channel prioritization views.

Aggregate funnel metrics hide the heterogeneity that makes GTM analysis actionable. A 28% win rate across all opportunities is a very different story if mid-market deals close at 42% and enterprise deals close at 14%. The agent slices every funnel metric by configurable dimensions — industry vertical, company size band, ACV tier, geographic region, GTM motion (inbound vs outbound vs PLG), and sales rep. Segment views reveal which ICP segments are the most efficient to acquire, which motions are working by segment, and where resources are concentrated versus where they generate the most revenue. Insights are surfaced automatically as anomalies rather than requiring manual cross-tab analysis.

A performance analysis tool that requires a weekly manual review meeting to produce insights isn't a system — it's a reporting layer. The agent monitors GTM metrics continuously and surfaces anomalies, trend breaks, and optimization opportunities as they emerge. When a channel's CAC spikes 40% in a two-week window, the relevant team gets an alert with context — which campaigns, which segments, and what changed. When a deal stage conversion rate drops below the historical baseline, the alert includes the segment breakdown so the owner can diagnose whether it's a rep performance issue, a product-market fit gap, or a campaign targeting problem. Insights arrive as actionable observations, not raw numbers.