{
  "version": "1.3.0",
  "releasedAt": "2026-05-20",
  "kpi": {
    "rogueId": "sales.downgrades",
    "slug": "downgrades",
    "domain": "sales",
    "defaultLabel": "Downgrade ARR",
    "description": "Annualized recurring revenue lost from existing customers who reduced spend mid-term or at renewal (seat reductions, tier downgrades, removed modules) — without leaving entirely. The \"contraction\" line of the ARR waterfall, distinct from full churn. Often a more sensitive leading indicator than churn because customers tend to contract before they cancel. Common pitfall: lumping downgrades into churn obscures the early-warning signal — boards looking only at logo churn miss the slow-bleed pattern. Surfaces in the KpiVarianceTable widget alongside expansion and churn so the net-retention math is auditable.",
    "fieldType": "currency",
    "unit": null,
    "maturity": "general",
    "suggestedForStages": [
      "seriesA",
      "seriesB",
      "seriesC",
      "public"
    ],
    "defaultOwningFunctions": [
      "Sales"
    ],
    "stageRelevance": {
      "seriesA": "recommended",
      "seriesB": "recommended",
      "seriesC": "recommended",
      "public": "recommended"
    },
    "definitionSource": {
      "tier": "editorial",
      "sourceName": "imboard Editorial",
      "sourceUrl": null,
      "sectionRef": null,
      "publicationDate": "2026-04-01",
      "attributionNotice": null
    },
    "formula": "Downgrade ARR = Sum across existing customers of (ARR at period start − ARR at period close) for the subset where the delta is positive AND the customer's ARR did not drop to zero. Customers whose ARR drops to zero count in Churned ARR instead.",
    "whyItMatters": "Earliest leading indicator of retention risk — customers usually contract before they cancel, so a rising downgrade line predicts churn 1–2 quarters out. Inputs NRR (subtracts from expansion) and CCO/CS comp models that gate on Net Retention.",
    "interpretationGuidance": "Downgrade ARR rising as a share of expansion ARR over 2+ quarters is the canonical leading signal of a renewal-cycle problem. There is no widely-published cross-company benchmark for downgrade rates as a standalone — read it in context of NRR (industry folk-wisdom: a healthy SaaS company with NRR ≥ 110% typically has downgrade ARR ≤ 3% of starting ARR per period).",
    "relatedKpiIds": [
      "sales.expansion",
      "sales.churn_arr",
      "sales.arr",
      "customers.net_revenue_retention",
      "customers.gross_revenue_retention"
    ],
    "calculationPolicy": {
      "inclusionRules": [
        "Sum across existing customers of (period-start ARR − period-end ARR) where the delta is positive AND the customer's ARR did not drop to zero.",
        "Includes seat reductions, tier downgrades, removed modules/products, and renewal-time contractions."
      ],
      "exclusionRules": [
        "Customers whose ARR dropped to zero — those are full churn (`sales.churn_arr`), not downgrades.",
        "New-logo deals and expansions — positive-delta customers do not appear here.",
        "Contractual price decreases that were pre-agreed in the original contract (rare, but disclose if present)."
      ],
      "requiredInputs": [
        "Customer-level ARR snapshot at period start.",
        "Customer-level ARR snapshot at period end.",
        "Per-customer churn flag (to separate full churn from contraction)."
      ],
      "dataSourcePriority": [
        "Customer-level ARR ledger (same source as expansion / churn / NRR).",
        "CRM contract changelog as a fallback."
      ],
      "edgeCases": [
        "Customer downgrades one product and expands another in the same period: net the per-customer delta; if negative, it counts here; if positive, it counts in expansion.",
        "Renewal at a lower price after a discount expires: count the reduction as a downgrade.",
        "Customer contracts to zero: that is churn, not a downgrade — even though it \"feels\" like the extreme downgrade."
      ],
      "validationChecks": [
        "Downgrade ARR ≥ 0 always.",
        "NRR = (Starting ARR + Expansion − Downgrades − Churn) / Starting ARR — downgrades and churn together must reconcile the waterfall.",
        "Per-customer downgrade ≤ that customer's period-start ARR."
      ],
      "commonMiscomputations": [
        "Lumping downgrades into churn — destroys the leading-indicator signal; customers usually contract 1-2 quarters before they cancel.",
        "Counting a contract-to-zero as a downgrade — it is churn; mixing the two corrupts both GRR and the churn line.",
        "Netting downgrades against same-customer expansion and reporting only the net — both gross lines are needed for the waterfall.",
        "Missing renewal-time downgrades because the system only captures mid-term amendments — renewal contractions are the larger share for most SaaS."
      ]
    },
    "metricBasis": {
      "timeBasis": "period_flow",
      "moneyBasis": "contracted_arr",
      "dateBasis": "period_close",
      "production": "primary"
    }
  }
}