Delivery KPI dictionary
Purpose: Long-form reference for delivery KPIs. Each entry includes the metric definition, evidence source, owner, floor, excellence, and cadence.
The canonical Q2 scorecard is in the Delivery Performance One-Pager. Q2 CSO and SL bonus eligibility is gated on the 4 BAU KPIs per role defined there. This dictionary contains additional reference KPIs (e.g., team-level, hygiene, cadence metrics) used for operating diagnostics and coaching, but only the 4 KPIs per role from the one-pager are bonus-gating in Q2.
Use this dictionary when:
- you need a deeper definition or evidence-source detail for a KPI
- you are setting up tracking infrastructure
- you are calibrating a borderline miss with HoD
For bonus eligibility, payout, and dispute logic, see the bonus eligibility policy.
How to read this document
Each KPI entry contains:
- Definition — what the metric actually measures
- Evidence source — where the data lives (no self-reported numbers)
- Owner — the role accountable for the number
- Floor — minimum threshold
- Excellence — higher bar
- Cadence — how often it is reviewed
CSO KPIs
C1 — Plan defensibility rate
Definition: Percentage of Project Review Meetings where the plan is approved on first presentation (not sent back for revision).
Why it matters: Plan quality before the PRM is the CSO’s primary preventable failure mode. Repeated send-backs indicate weak CSO-SL prep or thin plan narrative.
Evidence source: PRM log in knowledge/delivery/03-project-lifecycle/ — approval vs send-back outcome per engagement.
Owner: CSO
| Threshold | Target |
|---|---|
| Floor | ≥ 75% of PRMs approved on first pass |
| Excellence | ≥ 90% of PRMs approved on first pass |
Cadence: Quarterly
C2 — No-surprise rate (client escalation to founder)
Definition: Number of times a client issue reaches founder level without the CSO having surfaced it internally first.
Why it matters: The CSO is the shield between client friction and founder involvement. Each uncontrolled escalation is a failure of the early-warning role.
Evidence source: Tracked in weekly leadership sync notes and founder log.
Owner: CSO
| Threshold | Target |
|---|---|
| Floor | ≤ 1 uncontrolled founder escalation per quarter per client |
| Excellence | 0 uncontrolled founder escalations per quarter |
Cadence: Quarterly
C3 — Client communication cadence and quality
Definition: Percentage of active accounts receiving a daily substantive client-relevant touchpoint, plus a weekly summary that includes all four required elements: status, risks, asks, and next steps. Daily touchpoints can be Slack, email, Loom, meeting recap, artifact share, or another client-visible update, but they must contain real delivery signal.
Why it matters: Legible momentum is the CSO’s main trust-building tool. Empty updates erode confidence regardless of actual delivery progress. The format requirement ensures updates are actionable, not just present.
Evidence source: Client comms log; touchpoint tracker in client vault or Linear; spot check by HoD.
Owner: CSO
| Threshold | Target |
|---|---|
| Floor | Daily substantive touchpoints completed for ≥ 90% of active account-days; weekly summaries include status, risks, asks, next steps |
| Excellence | Daily substantive touchpoints completed for 100% of active account-days; ≥ 80% include a concrete artifact, demo, or capability update |
Cadence: Monthly
C4 — CSO-SL alignment cadence
Definition: Whether the CSO-SL weekly or async alignment checkpoint is happening for every active account. Measured as: accounts where alignment is documented or CSO confirms it occurred vs accounts where it was skipped.
Why it matters: Most avoidable delivery pain is discovered too late because CSO and SL are not actively comparing notes. Alignment cadence is the preventative layer.
Evidence source: Meeting notes, Slack thread, or Linear comment confirming alignment per account per week.
Owner: CSO (scheduling and documentation responsibility)
| Threshold | Target |
|---|---|
| Floor | Alignment documented for ≥ 80% of active accounts per week |
| Excellence | 100% of active accounts; alignment note includes specific risk/status review |
Cadence: Monthly
C5 — Pre-build success criteria completeness
Definition: Percentage of active projects that have explicit, documented success criteria defined and approved before build work begins.
Why it matters: Without upfront success criteria, “done” is negotiable. Ambiguity at this stage causes scope drift, rework, and client disappointment at delivery.
Evidence source: SOW ↔ project plan §§2–4 in vault; confirmed in Linear milestone or project description before first IC ticket is created.
Owner: CSO
| Threshold | Target |
|---|---|
| Floor | 100% of projects have documented success criteria before build begins |
| Excellence | Criteria are measurable (not “working dashboard” but “dashboard loads in < 3s, shows three KPIs, approved by client sponsor”) |
Cadence: Per project at PRM; reviewed monthly
C6 — Ask-to-ticket triage rate
Definition: Percentage of client-facing asks (requests, questions, scope items received from the client) that are logged in Linear and tied to a milestone or backlog within 24 hours of receipt.
Why it matters: Unlogged asks are invisible to the SL and ICs, which creates shadow scope, missed work, and accountability gaps. 24-hour triage prevents backlog contamination.
Evidence source: Linear ticket creation timestamp vs documented receipt time in client comms log or Slack.
Owner: CSO
| Threshold | Target |
|---|---|
| Floor | 100% of client-facing asks logged in Linear within 24 hours |
| Excellence | 100% logged and correctly triaged (scope vs backlog vs declined) with written rationale |
Cadence: Monthly
SL KPIs
S1 — Playbook coverage and maintenance cadence
Definition: Two components measured together:
- (a) Coverage: Percentage of active service components (actively billed work) that have a complete, non-placeholder playbook.
- (b) Maintenance: At least one playbook entry created or updated per sprint, directly traceable to an actual delivery failure or recurring friction pattern.
Why it matters: Playbooks turn individual execution into repeatable delivery. Coverage gaps mean ICs are improvising. The maintenance cadence ensures the playbooks reflect how work actually runs — not how it was ideally designed at first pass.
Evidence source: (a) PR audit against active service component list. (b) PR or vault commit with description referencing the triggering delivery event.
Owner: SL
| Threshold | Target |
|---|---|
| Floor | 100% of active service components have a playbook |
| Excellence | Playbooks used on ≥ 2 engagements with CSO-confirmed quality |
Cadence: Quarterly
S2 — Allocation drift
Definition: Percentage of active ICs whose actual Operating roster allocation is within ±10% of planned hours.
Why it matters: Roster drift is the leading indicator of shadow staffing, hidden capacity gaps, and planning failures.
Evidence source: Operating dashboard snapshot.
Owner: SL
| Threshold | Target |
|---|---|
| Floor | All active ICs ≤ 10% off Operating roster |
| Excellence | All allocations match roster; 0 shadow staffing instances |
Cadence: Monthly
S3 — Linear hygiene
Definition: Combination of (a) zero orphaned tickets (unlinked to a Project), (b) no open issues with zero updates for more than 30 days, and (c) a weekly hygiene pass owned by an explicitly assigned PM-focused IC wherever the account has that role available.
Why it matters: A dirty board means the system of record is not trustworthy. ICs and CSOs cannot make good decisions from a board that does not reflect reality. This work should be operationally adopted by a PM-focused IC when possible, with SL accountable for technical board truth.
Evidence source: Linear report; automated hygiene view; weekly hygiene pass note.
Owner: PM-focused IC when assigned; SL remains accountable for their client accounts.
| Threshold | Target |
|---|---|
| Floor | Zero orphaned tickets; zero open issues stale > 30 days; PM-focused IC assigned for hygiene where available |
| Excellence | ≥ 95% of tickets linked to Projects; weekly hygiene pass documented and reviewed by SL |
Cadence: Monthly
S4 — Quality gate compliance and defect rate
Two related measurements tracked together:
(a) QA/sign-off gate compliance
Definition: Percentage of milestone closes where the SL completed a documented technical sign-off before the milestone was marked done and before any client-facing delivery occurred.
Why it matters: Quality gates only work if they are consistently applied. An SL who signs off on nine milestones but misses one creates a false sense of safety. The gate must be 100% or it is not a gate.
Evidence source: Linear milestone close timestamp vs SL sign-off comment/approval timestamp.
| Threshold | Target |
|---|---|
| Floor | 100% of milestone closes have SL technical sign-off documented |
| Excellence | Sign-off includes explicit checklist items (e.g., tested, reviewed, acceptance criteria met) |
(b) Reopen rate on completed tickets
Definition: Percentage of tickets marked “Done” that are subsequently reopened within 14 days due to quality issues.
Why it matters: Reopen rate is the clearest lagging indicator of quality gate failure. High reopen rates mean ICs are marking work done before it is actually done, or the SL’s sign-off is not catching real issues.
Evidence source: Linear — tickets closed then reopened within 14 days; label reopen or status transition log.
| Threshold | Target |
|---|---|
| Floor | Reopen rate < 10% of closed tickets |
| Excellence | Reopen rate < 5%; zero reopens on client-visible milestones |
Bug definition: A bug is a defect in delivered work where the implementation fails to meet agreed acceptance criteria, expected behavior, or previously working functionality. New client requests, scope changes, external outages, and client-data issues not caused by Brainforge do not count as bugs.
Owner: SL (both components)
Cadence: Monthly
S5 — Estimation accuracy (Brainforge-caused timeline misses)
Definition: Of all timeline misses in the period, the percentage attributable to Brainforge-caused technical underestimation (as opposed to client delays, external dependencies, or approved scope changes).
Why it matters: Timeline misses happen. What matters is whether the SL’s estimates are the cause. Repeated estimation failures erode trust in planning, force CSO into reactive client communication, and signal that the SL is not pressure-testing commitments.
Evidence source: Per-miss root cause noted in Linear or vault; HoD and CSO confirm classification. Classes: (1) Brainforge estimation error, (2) client delay, (3) external dependency, (4) approved scope change.
Owner: SL
| Threshold | Target |
|---|---|
| Floor | < 10% of timeline misses are Brainforge-caused estimation errors |
| Excellence | < 5%; SL proactively flags estimation risk before the miss, not after |
Cadence: Monthly
S6 — IC unblock velocity
Definition: Average time from an IC-documented technical blocker to SL-confirmed resolution, redirect, escalation, or de-scope. Acknowledgement alone does not stop the clock; the SL must provide a decision, path, or owner.
Why it matters: Slow unblock velocity stalls delivery silently. ICs spinning on blockers without SL action is one of the primary delivery drag patterns.
Evidence source: Linear blocker label + resolution comment; Slack thread when relevant.
Owner: SL
| Threshold | Target |
|---|---|
| Floor | ≤ 2 business days average resolution for SL-owned blockers |
| Excellence | Same-day response; ≤ 1 business day average resolution |
Cadence: Monthly
IC KPIs
I1 — Execution reliability
Definition: Percentage of committed tickets (tickets assigned to this IC with a target date in the sprint window) delivered within the sprint window, or proactively rescheduled with a documented reason before the miss.
Sprint window = the agreed sprint boundary (Linear cycle end date, or HoD-set period if cycles are not in use).
Why it matters: Reliable execution — not heroics — is the IC’s core contribution. Unannounced misses are more costly than early-flagged ones. The sprint window framing ties individual reliability to team-level committed delivery.
Evidence source: Linear — issue completion timestamp vs cycle end date; proactive reschedule comment = not a miss.
Owner: IC (tracked by SL)
| Threshold | Target |
|---|---|
| Floor | ≥ 90% of committed tickets delivered in sprint window or proactively rescheduled |
| Excellence | ≥ 95% in window; no silent misses; reschedules include impact assessment |
Cadence: Monthly
I2 — Issue state honesty (ticket update SLA)
Definition: Percentage of open issues assigned to this IC that are updated within the 48-hour SLA — meaning: a substantive status comment, a progress note, a blocker flag, or a completion. Issues with zero updates for > 48 hours are an SLA miss.
Update SLA defined: 48 business hours from the last update (or from assignment if never updated). Weekends excluded.
Why it matters: The board is only as trustworthy as the ICs who maintain it. The OKR target is 95% compliance — this means nearly every open ticket must have a recent update. A single IC keeping false state creates planning failures for the whole team.
Evidence source: Linear audit; SL weekly hygiene check; automated stale-ticket report if available.
Owner: IC (enforced by SL)
| Threshold | Target |
|---|---|
| Floor | ≥ 95% of open issues updated within 48-hour SLA |
| Excellence | 100% compliance; updates are substantive (not just “working on it”) |
Cadence: Monthly
I3 — Escalation quality
Definition: Whether escalations from this IC include: (a) what is blocked, (b) what is at risk, (c) what was already tried, and (d) a proposed next step. Measured as percentage of IC escalations that meet all four criteria.
Why it matters: Low-quality escalations create work for SL and CSO instead of accelerating resolution. High-quality escalations compress the unblock cycle.
Evidence source: Slack escalation threads, Linear blocker comments; reviewed by SL.
Owner: IC (reviewed by SL)
| Threshold | Target |
|---|---|
| Floor | ≥ 80% of escalations include impact + proposed path |
| Excellence | ≥ 95% of escalations include all four criteria; no founder-routed escalations that should have gone to SL first |
Cadence: Quarterly
Team-layer KPIs
T1 — Committed ticket delivery rate (team-level, exec-intervention-free)
Definition: Percentage of committed team-level tickets (all ICs across all accounts) delivered within the sprint window, without requiring executive intervention to unblock or reprioritize.
“Exec intervention” = a founder or HoD having to directly unblock, reprioritize, or make a scope decision that should have been handled at CSO/SL/IC level.
Why it matters: This is the portfolio-level view of I1. It captures whether the CSO/SL system is functioning well enough to protect delivery from executive bottlenecks. A team consistently needing exec intervention means the leadership layer is not working.
Evidence source: Linear cycle completion report + HoD intervention log from weekly operating review.
Owner: HoD
| Threshold | Target |
|---|---|
| Floor | ≥ 90% of committed tickets delivered in sprint window without exec intervention |
| Excellence | ≥ 95%; zero exec interventions for issues that should have been handled by CSO/SL |
Cadence: Monthly
T2 — Checkpoint compliance rate
Definition: Percentage of scheduled checkpoints (PRM, weekly leadership sync, CSO-SL alignment, client touchpoints) that occurred as intended in the period.
Why it matters: Checkpoints are how the standard is enforced. A team with low checkpoint compliance is running on informal norms, not the system.
Evidence source: Meeting log in Linear / vault; confirmed by HoD.
Owner: HoD (portfolio-level); CSO/SL (account-level)
| Threshold | Target |
|---|---|
| Floor | ≥ 90% of scheduled checkpoints completed |
| Excellence | 100%; checkpoint notes filed per template within 24 hours |
Cadence: Monthly
T3 — First-touch routing efficiency
Definition: Percentage of inbound client requests (scope questions, timeline questions, technical-quality questions) that are correctly routed and resolved without requiring a follow-up call to clarify ownership. CSO handles scope/timeline; SL handles technical-quality.
Why it matters: Requests that require a second call to establish ownership signal unclear role seams and erode client confidence. First-touch resolution is a direct measure of how well the CSO/SL partnership is functioning in practice.
Evidence source: Client comms log; CSO notes in vault; HoD spot check. A request is “resolved” if the client received a clear answer and owner on first contact.
Owner: CSO (routing discipline); SL (technical response accuracy)
| Threshold | Target |
|---|---|
| Floor | > 90% of client requests handled with clear routing on first touch |
| Excellence | 100%; zero requests where ownership was unclear or required a follow-up call to establish |
Cadence: Monthly
T4 — Escalation aging
Definition: Average number of days open escalations have been sitting without a named owner or resolution decision.
Why it matters: Escalations that sit unresolved degrade team trust and allow risks to compound. Escalation aging is a proxy for leadership responsiveness.
Evidence source: Linear escalation label; HoD review.
Owner: HoD
| Threshold | Target |
|---|---|
| Floor | No escalation open > 5 business days without named owner and active resolution |
| Excellence | Average escalation resolution ≤ 3 business days; no escalations older than 7 days on weekly review |
Cadence: Weekly review; monthly reported
T5 — Re-gate compliance
Definition: Percentage of material scope or technical changes that go back through the re-gate process (updated plan + HoD review) rather than being absorbed silently.
Why it matters: Silent scope absorption is one of the most common sources of delivery failure and margin erosion. Re-gate compliance is the control mechanism.
Evidence source: PRM / re-gate log; HoD tracking.
Owner: HoD (enforcement); CSO (trigger)
| Threshold | Target |
|---|---|
| Floor | 100% of material changes re-gated (no exceptions) |
| Excellence | Re-gates initiated proactively by CSO or SL before drift is visible to client |
Cadence: Per-event; reviewed quarterly
Related
- Bonus eligibility policy
- IC RACI
- SL success metrics
- Consolidated delivery standards
- Weekly delivery operating review
Last updated: 2026-04-26 (reconciled with Q2 Delivery OKRs)