Skip to main content

Documentation Index

Fetch the complete documentation index at: https://hc.pillargtm.com/llms.txt

Use this file to discover all available pages before exploring further.

​
Renewal Forecasting

The Renewal Forecasting engine computes a probability of renewal for every upcoming renewal in your portfolio, then aggregates those probabilities into org-level GRR and NRR projections. Each probability is decomposed into explicit risk factors and positive signals so teams can see exactly why a renewal is confident or at risk.
Renewal Forecasting builds on data from the Scoring Engine (health scores, risk scores) and the Signals taxonomy (critical signal detection). It does not replace those systems — it consumes their outputs to produce forward-looking retention projections.

​
How It Works

Scoring Engine (entity-level)
  → Health score, risk score, engagement score

Renewal Forecasting Engine
  → Per-renewal probability (0.01–0.99)
  → Risk factor decomposition (what's dragging probability down)
  → Positive signal decomposition (what's boosting probability)
  → Org-level GRR / NRR projection (ARR-weighted)
  → Forecast snapshot (persisted for trend analysis)

​
Probability Model

Each renewal receives a probability score between 0.01 and 0.99 computed from multiple input signals including:
  • Health Score — Current account health level
  • Risk Score — Renewal risk assessment
  • Engagement Score — Activity and usage patterns
  • Disposition — Current renewal lifecycle state
  • Days to Renewal — Proximity to renewal date and urgency
  • Active Playbooks — Whether intervention plays are in progress
  • Critical Signals — Recent critical signals on the account
  • NPS Score — Customer satisfaction level
  • Contract Tenure — Length of customer relationship
  • Deal Size — ARR-based scrutiny adjustment
Each factor adjusts the probability up or down from a calibrated baseline. The final value is clamped to a valid probability range.
Exact model weights, base probability, adjustment formulas, and clamp values are available in the PILLAR Implementation Guide provided to active customers.

​
Confidence Tiers

Each renewal is classified into a confidence tier based on its computed probability:
TierMeaning
highRenewal is highly likely
mediumSome risk factors present
at_riskSignificant risk — intervention recommended

​
Org-Level Projections

The engine aggregates per-renewal probabilities into portfolio-level metrics:
MetricDescription
Projected GRRARR-weighted average of renewal probabilities
Projected NRRGRR adjusted for expansion rate
Weighted Renewal ARRProbability-weighted sum across all renewals
High Confidence ARRARR from high-confidence renewals
At-Risk ARRARR from at-risk renewals

​
Data Model

PILLAR stores org-level forecast snapshots for trend tracking and per-renewal probability computations with full risk/signal decomposition. Each forecast is versioned and keyed by organization and period.
Detailed data model schemas are available in the PILLAR Implementation Guide provided to active customers.

​
Period Formats

The forecast engine accepts multiple period formats:
FormatExampleRange
Quarter2026-Q2Apr 1 - Jun 30
Half2026-H2Jul 1 - Dec 31
Year2026Jan 1 - Dec 31
If no period is specified, the engine defaults to the next quarter.

​
API Endpoint

GET  /api/forecasting/renewals
POST /api/forecasting/renewals
See the Renewal Forecast API reference for full endpoint documentation.

​
Access

Available to: CRO/CEO, VP Sales, VP CS, RevOps