IT services / MSPs Operations Questions

The questions that recur in managed IT share one shape: the service sells protection and gets graded on response. Churn concentrates in clients who never had an incident while disaster-rescued clients stay for years, and project revenue spikes right before managed-service cancellation. The answers below measure the absence of disaster, which is the actual product.

• Why does tools-stack cost per tech rise as we add clients, which integration tax are we not counting?

  The per-client onboarding and exception-handling tax: every non-standard client environment creates one-off scripts, alert tuning, and documentation in your RMM/PSA. Tools are priced per endpoint but operated per exception. Enforce stack standardization or price the exceptions. The tax is real either way.

  §12 platform layer, §13 standard work, §1.3

• Why do we win on security posture, then get reviewed quarterly on response time?

  Because the pitch set one winner and the QBR measures another: post-sale, the client's daily experience is tickets, not posture. Align the quarterly review to include the security outcomes you sold (incidents prevented, patch compliance), or the measured metric becomes the de facto product.

  §13 Goodhart, §1.1 OrderWinner migration, §6.3

• Why does churn concentrate among clients who never had a major incident, while disaster-rescued clients stay forever?

  Invisible value: no incident means no proof of your prevention. The quiet client concludes "nothing ever happens, why pay?" The rescued client saw the alternative. Counter with value reporting (blocked threats, patched CVEs, downtime avoided). Make prevention legible or it reads as absence.

  §6.1 perceived quality of prevention, §8.2 recovery paradox, §2.3

• Does per-seat pricing survive client headcount downturns, or is every layoff an unbudgeted renegotiation?

  It does not: per-seat couples your revenue to their HR decisions. A demand risk you have priced as zero. Options: minimum-commit tiers, per-device + floor pricing, or explicit annual true-up terms. Choose the structure before the downturn chooses the renegotiation for you.

  §11.4 demand risk, §2.3 pricing structure, §0.3 governedBy

• When the client's office manager is our champion, why does the account wobble when they go on leave?

  Single-thread dependency: your institutional knowledge of the client lives in one person who routes everything. When they are out, requests stall, tickets age, and the client experiences your service as degraded. Multi-thread every account: at least two contacts who know the escalation paths.

  §13 relationship capital, §11.4 single-point failure, §8.2 blueprint

• Are QBRs changing client behavior or billing ourselves for theater, which clients act on them?

  Audit: which clients implemented last quarter's recommendations? If <20%, QBRs are theater for most accounts. Restructure: only run QBRs where an executive owner attends and decisions get minuted. For the rest, a quarterly report suffices. Match the ceremony to the audience's authority.

  D3 NVA, §6.3, §1.1 segment

• Clients resisting stack standardization become our highest-margin accounts. What does resistance predict?

  Deep entanglement: resistant clients have complex environments where you are embedded (custom integrations, institutional knowledge). High switching costs and premium tolerance. Resistance predicts lock-in. Profitable, but note the risk mirror: you are also locked into their complexity.

  §11.2 switching costs, §11.4, §1.3

• Why does project revenue spike right before managed-service churn. Are projects a symptom of disengagement?

  Yes, often: a client planning their exit commissions final projects (migration to in-house, new platform) before canceling the agreement. The project pipeline is a leading churn indicator nobody watches. Flag accounts with unusual project activity for a relationship review.

  §2.2 leading indicators, §13, §11.2

• Fully-managed clients generate more tickets per seat than break-fix converts. Engagement or noise, which are we staffed for?

  Mostly engagement (they use the service they pay for) plus expectation inflation (everything is "included"). Staffing assumes break-fix ticket rates and drowns in managed volume. Normalize: ticket deflection (self-service, automation) and per-seat ticket budgets baked into pricing tiers.

  §8.1 arrival rates, §12 automation, §2.3

• Escalations cluster by time of day, not complexity. Who is on shift, or what do we call complex?

  Both possibilities resolve with data: if escalations cluster at shift boundaries or junior-coverage windows, it is coverage skill, not ticket nature. If truly time-correlated regardless of staff, it is client-side patterns (their batch jobs, their EOD panic). Stratify by shift first. The cheaper fix.

  §6.3 stratification, §7 scheduling, §4.4 skill matrices

How these answers work

Each answer names the operational mechanism the question is circling, then states the directive that follows from the ontology in Part One of the book. Bracketed citations point to the ontology sections and axioms that produced the answer.

Related industries

• Management consulting operations questions
• Marketing / advertising agencies operations questions
• Staffing / recruiting agencies operations questions
• Accounting / bookkeeping / tax prep operations questions
• Web design / development shops operations questions

Keep reading

• All 100 industries and the full book
• Download the PDF
• Free diagnostic tool
• Methodology