Glossary
The formulas behind account health, retention risk, and voice-of-customer signals, the same inputs BaseDynamics' Autonomous Scoring and AI agents use, defined for post-sales revenue teams.
The average time tickets stay open or take to resolve, across both currently open and already-resolved tickets. A slipping number usually means support responsiveness is degrading before customers say so directly. Avg resolution time (resolved + open tickets aging).
Computed automatically from connected helpdesk data.
A risk rule that fires when an account both talks about alternatives and is close to renewal, two individually minor signals that become urgent together.
Rule: ≥2 competitor mentions (negative sentiment) AND renewal within the next 120 days.
Part of: Retention Likelihood Score inputs.
Detected automatically through competitor mention scanning cross-referenced with CRM renewal dates.
How much core-feature usage volume each active user is generating on average, a depth-of-engagement signal beyond simple login counts.
Core Events per Active User = (total core events, last 30d ÷ MAU) x 100One of the derived scoring inputs computed automatically.
What share of core feature usage comes from just your top 10% of users. A high number means the account's real value delivery is concentrated in a few people, a champion-risk signal.
Core Feature Usage Concentration Ratio = (core feature usage by top 10% of users, last 30d / total core feature usage, last 30d) × 100.Used as a portfolio distribution guardrail inside health scoring.
Whether core feature usage is declining compared to the prior period, one of the clearest early indicators that an account's engagement is backsliding.
Core Feature Usage Regression Ratio = avg core feature usage, last 30d ÷ avg core feature usage, last 90d.Part of: Retention Likelihood Score composite.
A direct input into the Retention Likelihood Score.
An urgency rule that flags when an account's stickiness suddenly falls well below its recent baseline, a faster-firing signal than the standard Stickiness Drop metric.
Rule: Stickiness, last 30d ≤ 70% of prior 60d.
Part of: Urgency Signal system (separate real-time layer).
An urgency rule that flags a sudden fall in how many features an account is actively using. Distinct from a raw usage volume drop, this specifically tracks breadth of adoption collapsing.
Rule: Feature usage, last 14d ≤ 70% of prior 30d.
Part of: Urgency Signal system (separate real-time layer).
Whether the number of core features an account touches is shrinking over time. A narrowing footprint often precedes a broader engagement collapse.
Feature Churn Ratio = (core Features Used, last 30d ÷ core Features Used, last 90d).One of the regression and risk trailing indicators computed automatically.
An urgency rule that flags a sudden fall in raw usage volume. Separate from Feature Adoption Drop, this catches an account using the same features less intensely, not just fewer features.
Rule: (Usage volume, last 14d ≤ 70% of prior 30d).
Part of: Urgency Signal system (separate real-time layer).
A score reflecting how much of an account's recent support activity carries negative sentiment, designed to surface mounting frustration before it becomes a churn event.
Frustration Index = (negative-sentiment tickets, last 90d / total tickets, last 90d).Part of: Retention Likelihood Score composite.
A direct input into the Retention Likelihood Score.
A risk rule based on an account's own past behavior: if they've downgraded before a prior renewal, that history itself becomes a standing risk factor.
Rule: Previously downgraded before renewal.
Part of: Retention Likelihood Score composite.
Pulled automatically from CRM contract history.
Whether "how-to" questions (a signal of active but confused usage, not necessarily unhappy usage) are accelerating in the most recent window compared to the prior one.
How-to Intent Trend = (total "How-to" intents, last 15d / total "How-to" intents, last 30d) × 100.Computed automatically from intent-classified customer voice data.
A risk rule that treats disconnecting an integration as a meaningful signal. Removing a connected tool is often an early, quiet sign that an account is disengaging.
Rule: ≥ 1 integration removed
Part of: Health Score inputs.
Tracked automatically wherever integration status changes are visible in your connected stack.
A weighted blend of what customers are talking about, issues, feature asks, and how-to questions, giving more weight to problems than requests or confusion.
Intent Ratio = ((issue Intents, last 30d × 0.5) + (ask Intents, last 30d × 0.3) + (howTo Intents, last 30d × 0.2)) ÷ total Detected Intents, last 30d × 100.Computed automatically from AI-enriched customer voice intent classification.
An urgency rule that flags a sudden increase in the weighted intent mix, typically driven by a burst of issue-heavy conversations in a short window.
Rule: Intent Ratio, last 14d ≥ 130% of prior 30d
Part of: Urgency Signal system (separate real-time layer).
A capped, longer-window measure of how much of an account's customer voice volume is specifically issue-related, distinct from the shorter-window, multi-intent blend used in Intent Ratio.
Issue Intent Ratio = Issue intents, last 90d / total intents detected, last 90d.Part of: Retention Likelihood Score composite.
A direct input into the Retention Likelihood Score.
What share of started core workflows an account actually finishes, a direct measure of whether your product's key flows are usable in practice, not just entered.
Key Workflow Completion Rate = (completed Workflows, last 30d ÷ started Workflows, last 30d) × 100.A simple, direct comparison of recent monthly active usage against the trailing 90-day baseline, the most fundamental usage-regression signal in the health scoring model.
MAU Decline Ratio = MAU, last 30d ÷ MAU, last 90d.Part of: Retention Likelihood Score composite.
A direct input into the Retention Likelihood Score.
Whether monthly active usage is currently running above or below its own recent average, a momentum read rather than a point-in-time snapshot.
MAU Momentum = (MAU, last 30d / avg MAU, last 90d) × 100.Computed automatically alongside MAU Decline Ratio.
What share of all customer feedback in the trailing 30 days carries negative sentiment, a blunt but fast-moving temperature check on account mood.
Negative Emotion Ratio = (negative-sentiment feedback rows, last 30d / total feedback rows, last 30d) × 100.Computed automatically from AI-enriched sentiment classification.
Whether the volume of negative-sentiment tickets is easing or worsening compared to the prior period, framed so a higher score means things are improving.
Negative Ticket Trend = 100 − ((negative open tickets, last 30d / negative open tickets, last 90d) × 100).Computed automatically from connected helpdesk data.
An urgency rule that flags a sudden decline in average sentiment score, catching a mood shift in near real time rather than waiting for a monthly rollup.
Rule: Sentiment, last 14d ≤ 70% of prior 30d.
Part of: Urgency Signal system (separate real-time layer).
The average sentiment across all customer feedback in the trailing 30 days, the baseline qualitative health read that other sentiment-derived metrics build on.
Sentiment Score = avg(feedback_sentiment_score, last 30d).Computed automatically from AI-enriched customer voice sentiment scoring.
A time-weighted version of Sentiment Score where more recent feedback counts more heavily than older feedback, so a sharp recent dip shows up faster than it would in a flat 30-day average. Weighted sentiment score with a decay function favoring recent feedback.
Computed as a decay-weighted variant of Sentiment Score.
Whether support ticket volume is rising or falling relative to the prior period, a volume-based companion to the sentiment-based trend metrics.
Ticket Count Trend = (ticket count, last 30d / ticket count, last 90d) × 100.Computed automatically from connected helpdesk data.
A raw count of how-to intents detected in the trailing 30 days, the base input behind How-to Intent Trend. Count of total "How-to" intent in the last 30 days.
Captured automatically from AI intent classification.
A raw count of issue-related intents detected in the trailing 30 days, the base input behind several issue-focused ratio metrics on this page. Count of total "Issue" intent in the last 30 days.
Captured automatically from AI intent classification.
A raw count of all feedback items received in the trailing 90 days, across every connected voice channel, the base denominator for several longer-window ratio metrics. Count of total feedback in the last 90 days.
Captured automatically from every connected customer voice channel.
An urgency rule named for what it catches: a sudden, sharp fall-off in monthly active users, sharper and faster-firing than the standard MAU Decline Ratio.
Rule: MAU, last 14d ≤ 70% of prior 30d.
Part of: Urgency Signal system (separate real-time layer).
What share of active users are engaging with two or more core features, not just one, a breadth-of-engagement signal at the individual user level.
User Engagement Depth Ratio = (users using ≥2 core features, last 30d / MAU) × 100.One of the portfolio distribution guardrails computed automatically.
What share of an account's total usage comes from just its top 10% (or single most active) users. A high concentration means the account's success depends heavily on one or two people, a churn risk if that person leaves.
User Usage Concentration Ratio = (usage by top 10% of users, last 30d / total usage, last 30d) × 100.Used as a portfolio distribution guardrail inside health scoring.
BaseDynamics computes these health, voice, and urgency signals automatically from connected product, support, and CRM data.
