Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Customer leadership advisor on CX strategy, retention and expansion, and voice-of-customer programs. Use when defining a CX strategy, scoring CX maturity, planning churn interventions, or designing a VoC program.
.claude/skills/borghei-chief-customer-officer-advisor/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | 129% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 68% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 98% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 123% | 0% |
The agent acts as a fractional Chief Customer Officer, providing customer strategy, retention/expansion, and voice-of-customer guidance grounded in SaaS retention benchmarks, modern CX program patterns, and the operational realities of post-sale teams.
voice, operations, talent).
cx_maturity_scorer.py against the populated JSON.bashpython3 chief-customer-officer-advisor/scripts/cx_maturity_scorer.py \ --input cx_state.json --format markdown
churn_intervention_planner.py to prioritize and assign interventionsmatched to risk type and tier.
bashpython3 chief-customer-officer-advisor/scripts/churn_intervention_planner.py \ --input at_risk_accounts.json --format markdown
voc_program_designer.py to recommend a target VoC architecture and a12-month rollout sequence.
bashpython3 chief-customer-officer-advisor/scripts/voc_program_designer.py \ --input voc_state.json --format markdown
Pick clearly. Most CCO scope debates stem from ambiguous ownership.
| Function | Default ownership | |----------|-------------------| | Customer Success | CCO | | Support / Customer Support | CCO (or VP Support reporting in) | | Onboarding / Services | CCO (or separate Services GM in larger orgs) | | Renewals | Usually CCO; sometimes CRO | | Expansion (cross-sell / upsell) | Split: CCO on usage-driven; CRO on net-new product lines | | VoC program | CCO | | Customer marketing (advocacy, references, community) | Often CCO; sometimes CMO | | Customer Education / Training | CCO |
When the CRO and CCO both report to CEO, the renewals + expansion question is the friction point. Resolve it explicitly; don't leave it to a quarterly food fight.
A useful segmentation is one your motion actually differentiates on:
If you've defined "Enterprise" but you treat all customers identically, your segmentation is theater. Tie segments to:
A rough guide (highly company-dependent):
| Segment | ARR per CSM (USD) | Accounts per CSM | |---------|-------------------|------------------| | Enterprise high-touch | $4M–$10M | 10–25 | | Mid-Market | $2M–$5M | 30–80 | | SMB / Pooled | $1M–$2M | 200–500 | | PLG / Tech-touch | $5M+ | 1000+ |
If your ratio is far above the band, expect churn to creep up; far below, your CS unit economics will hurt margin. Either way, name the choice explicitly.
NRR is the single most predictive metric of long-term outcomes. Drivers:
Things often credited for NRR that don't move the needle:
references/customer-experience-strategy.md — CX strategy framing, segmentation, scorecardsreferences/retention-and-expansion-frameworks.md — NRR thesis, save programs, expansion motionsreferences/voice-of-customer-program.md — VoC architecture, action loops, instrumentsbusiness-growth/customer-success-manager — operational CSM tacticsbusiness-growth/churn-prevention — execution of save programsc-level-advisor/cmo-advisor — customer marketing alignmentc-level-advisor/cro-advisor — renewals + expansion boundaryc-level-advisor/cpo-advisor — feedback loop to productproduct-team/user-research — interview frameworks for churn / expansion| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | fail→fail | 14,530 | 13,499 | -7% | 1 | 1 | 0% | 2,483 | 4,293 | +73% | 0 | 0 | — |
case-09 | fail→pass | 7,490 | 2,261 | -70% | 1 | 1 | 0% | 1,052 | 2,413 | +129% | 0 | 0 | — |
case-01 | fail→pass | 23,833 | 26,610 | +12% | 1 | 1 | 0% | 3,757 | 6,316 | +68% | 0 | 0 | — |
case-02 | fail→fail | 38,116 | 19,730 | -48% | 1 | 1 | 0% | 6,075 | 5,116 | -16% | 0 | 0 | — |
case-03 | pass→pass | 18,750 | 14,428 | -23% | 1 | 1 | 0% | 2,653 | 4,138 | +56% | 0 | 0 | — |
case-04 | fail→pass | 19,668 | 15,642 | -20% | 1 | 1 | 0% | 3,010 | 4,207 | +40% | 0 | 0 | — |
case-05 | pass→pass | 18,566 | 14,805 | -20% | 1 | 1 | 0% | 2,689 | 4,148 | +54% | 0 | 0 | — |
case-06 | pass→pass | 17,236 | 17,058 | -1% | 1 | 1 | 0% | 2,541 | 4,421 | +74% | 0 | 0 | — |
case-07 | fail→fail | 20,720 | 17,305 | -16% | 1 | 1 | 0% | 3,037 | 4,447 | +46% | 0 | 0 | — |
case-08 | fail→fail | 16,294 | 12,280 | -25% | 1 | 1 | 0% | 2,751 | 3,978 | +45% | 0 | 0 | — |
case-10 | fail→pass | 7,304 | 2,824 | -61% | 1 | 1 | 0% | 1,264 | 2,503 | +98% | 0 | 0 | — |
case-11 | fail→pass | 13,952 | 1,912 | -86% | 1 | 1 | 0% | 1,066 | 2,376 | +123% | 0 | 0 | — |
case-12 | pass→fail | 14,469 | 12,952 | -10% | 1 | 1 | 0% | 2,538 | 4,241 | +67% | 0 | 0 | — |
case-13 | fail→fail | 14,396 | 12,335 | -14% | 1 | 1 | 0% | 2,428 | 4,074 | +68% | 0 | 0 | — |
case-14 | pass→pass | 16,152 | 15,517 | -4% | 1 | 1 | 0% | 2,584 | 4,575 | +77% | 0 | 0 | — |
case-16 | pass→pass | 12,899 | 12,494 | -3% | 1 | 1 | 0% | 2,121 | 4,117 | +94% | 0 | 0 | — |
case-17 | pass→pass | 13,863 | 14,298 | +3% | 1 | 1 | 0% | 2,329 | 4,446 | +91% | 0 | 0 | — |
case-18 | pass→pass | 15,043 | 19,627 | +30% | 1 | 1 | 0% | 2,080 | 4,605 | +121% | 0 | 0 | — |
case-19 | pass→pass | 15,311 | 14,652 | -4% | 1 | 1 | 0% | 2,250 | 4,030 | +79% | 0 | 0 | — |
case-20 | pass→pass | 13,308 | 18,238 | +37% | 1 | 1 | 0% | 2,363 | 4,924 | +108% | 0 | 0 | — |
case-21 | pass→pass | 13,216 | 11,456 | -13% | 1 | 1 | 0% | 2,354 | 4,108 | +75% | 0 | 0 | — |
case-22 | pass→pass | 12,705 | 13,548 | +7% | 1 | 1 | 0% | 2,725 | 5,005 | +84% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +18 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.