CHATTERgo's reports answer the question that matters: what did the AI actually contribute? Everything below lives in your dashboard.
Conversion attribution
The overall report lets you choose a conversion attribution scope:
- Chat-linked — conversions from shoppers who engaged in an AI conversation, linked across sessions and devices where possible.
- Recommendation-linked — conversions credited to a product recommendation click.
- All tracked — every store order the tracking pixel captured, a baseline comparable with Google Analytics.
The headline conversion tile always shows its arithmetic — for example, "attributed orders ÷ chat sessions" — and a "How this is calculated" tooltip spells out numerator, denominator, and the rule for the selected scope. Each scope divides by its own population (chat sessions for chat-linked, sessions that clicked a recommendation for recommendation-linked, tracked sessions for all tracked), so rates stay comparable and honest. Checkout funnels follow the selected scope.
Agent Performance report
Per-team-member performance for human handovers, described in Inbox and human handover: pickup SLA against your business-minutes target, business-hours-aware average response times, and AI summary, sentiment, and key topics per conversation.
Recommendation analytics
- Recommendation Performance — impressions, clicks, add-to-carts, purchases, and revenue per recommendation row, with fast date-range filtering (7 / 30 / 90 days or a custom range).
- Strategy Funnel — the same funnel aggregated per strategy, labelled by how each strategy picks products. See Recommendations overview.
Keeping numbers honest
- Admins can exclude individual conversations (tests, training) from KPI reporting.
- Background processing never reshuffles conversation history; nightly jobs surface failures in the dashboard instead of failing silently.