Delivery Leadership Operating Guide
For: CSOs, SLs, and anyone moving into a delivery leadership role. Purpose: Behavior standards, weekly self-checks, and coaching examples for hitting the role.
For the actual scorecard, KPIs, excellence, and bonus math, go to:
This guide does not own KPI math. It owns the behaviors that produce the numbers.
What excellent looks like — by role
CSO: excellent looks like a client who never has to guess
- The client always knows where things stand, what is at risk, what is needed from them, and what happens next — without asking.
- Founder involvement is intentional, not reactive. Founders join client conversations because the CSO brought them in strategically.
- Plans are defensible before they are presented because the CSO and SL did the alignment work before the meeting, not in it.
- Every client ask is in Linear within 24 hours. Nothing lives in Slack threads or someone’s memory.
- Success criteria exist in writing before build work begins, so “done” is never up for debate.
The failure mode of a CSO who hits the floor but misses excellence: reactive updates, reactive escalation, plans that need revision, asks tracked informally. Nothing is broken — but nothing feels controlled either.
SL: excellent looks like a team that never has to guess what good work means
- Playbooks exist, are complete, and reflect how delivery actually runs.
- Quality gates are not a formality. Every milestone close has a real sign-off, and reopens are low because the sign-off catches real issues.
- Estimation is trustworthy. When the SL says a date is possible, it is — because assumptions were pressure-tested before commit.
- ICs are unblocked fast. Documented blockers get a real decision, not an acknowledgement, within a day.
- The roster reflects reality. No shadow staffing. No IC running 20 hours on a client that shows 5 in Operating.
The failure mode of an SL who hits the floor but misses excellence: playbooks exist but no one uses them, quality gates pass but reopens are high, estimates are usually correct but misses are rarely flagged early.
Your weekly self-check
Don’t wait for the operating review to find out where you stand. Each week, before the leadership sync, run this 10-minute check yourself.
CSO self-check
- Did every active account get a substantive client touchpoint this week?
- Does every weekly summary I sent include status, risks, asks, and next steps — not just status?
- Did I document CSO-SL alignment for every active account this week?
- Are there any unlogged client asks sitting in Slack, email, or memory right now?
- Is there any project where build work has started without documented success criteria?
- Is there any client conversation happening that a founder could be surprised by?
SL self-check
- Is every active service component covered by a real, non-placeholder playbook?
- Did I complete a documented technical sign-off before every milestone that closed this week?
- Did I update at least one playbook entry this sprint, traceable to a real delivery event?
- Are there any orphaned tickets or issues with zero updates in the last 30 days?
- Are there any IC-documented blockers that have been sitting unresolved for more than a day?
- Is there any date I committed to that I no longer believe is technically achievable?
If the answer to any of these is “no” or “unsure” — that is the week’s work.
Worked example — SL (Strategy & Intelligence)
Scenario: Jasmin is an SL on two active accounts in Q2.
| KPI | Account A | Account B | Assessment |
|---|---|---|---|
| QA/sign-off before milestone close | 100% sign-off documented | One milestone closed before SL sign-off | Below 90% across portfolio → floor miss |
| Reopen rate on done tickets | 8% | 14% | Portfolio reopen rate above 10% → floor miss |
| Brainforge-caused estimation misses | <5% | 25% (2 of 8 misses) | Portfolio above 10% → floor miss |
| Playbook entry per sprint | 1 update | 0 updates | Did not hit ≥ 1 per sprint → floor miss |
Result: Jasmin missed the floor on all 4 SL KPIs. She is not Q3 bonus-eligible. HoD opens a corrective conversation. Q3 bonus access depends on Q2 floor being clean before assessment.
What needs to change:
- Treat SL sign-off as a hard gate; no milestone closes without it.
- Run a sprint-level playbook discipline check — at least 1 update per sprint, even when delivery is calm.
- Pressure-test estimates with the SL before client commit, especially on Account B’s complexity.
Worked example — CSO
Scenario: Greg is a CSO on two active accounts in Q2.
| KPI | Account A | Account B | Assessment |
|---|---|---|---|
| Weekly client updates with status/risks/asks/next | 100% complete | 1 weekly summary missing explicit asks | Floor met (just); excellence missed on consistency |
| Success criteria before build | Defined and measurable | Defined but not fully measurable | Floor met; excellence missed on measurability |
| Asks triaged to Linear within 24h | 100% | 100%, with written rationale | Floor and excellence met |
| Active project plan linked to Linear | Linked, current | Linked but plan not refreshed in 2 weeks | Floor met; excellence missed on freshness |
Result: Greg cleared the Q2 floor gate on all 4 CSO KPIs and is bonus-eligible for Q3. He hit excellence on 1 of 4 KPIs in this period, which would put him in the 25% band unless he closes the gaps on KPIs 1, 2, and 4 in the scoring window. Reaching the 50% band requires excellence on all 4.
What needs to change to reach 50%:
- Every weekly summary structurally complete — status, risks, asks, next steps — on every active account, every week.
- Every project’s success criteria measurable and testable before build, not descriptive.
- Project plans refreshed weekly with the latest milestone and risk view from CSO-SL alignment.
Related
- Delivery Performance One-Pager
- Bonus eligibility policy
- Delivery KPI dictionary
- IC RACI
- Weekly delivery operating review
Last updated: 2026-04-28