Operational Capability Review (Commercial Lens)

A growth story only converts to cash if the company can make, move, and service what it sells—reliably, at the promised quality, and at a cost that sustains margin. The operational capability review is your bridge between the commercial thesis and the realities of supply, fulfillment, and service. It asks three pragmatic questions. First, can the business meet planned demand by segment and channel at target service levels (on‑time, in‑full, damage‑free)? Second, can it do so at a cost structure and working‑capital profile that preserves contribution margin and payback? Third, what shocks—supplier failures, logistics disruptions, quality issues, regulation—could bend revenue or compress margin inside the hold period?

This chapter is not an engineering audit; it is a commercial lens on operations. You will map how operational constraints show up in revenue, retention, realized price, and CM2. You will quantify revenue‑at‑risk and margin‑at‑risk for critical SKUs and services, and you will translate those risks into model ranges, terms (e.g., indexation, inventory covenants), and a 90‑day post‑close plan. The goal is a short list of must‑fix constraints and must‑protect advantages, each with an owner, a timeline, and a measurable impact on the P&L.

10.1 Supply Chain Resilience Checklist

Resilience is the ability to continue selling at planned service levels and margin despite shocks. In diligence, you need to (i) locate single points of failure, (ii) measure how fast the system fails when stressed, and (iii) specify the few actions that materially raise time‑to‑survive and reduce time‑to‑recover. The checklist below is designed for a two‑ to four‑week sprint. It yields three outputs the deal team can use immediately: a Revenue‑at‑Risk map, a Margin‑at‑Risk map, and a Readiness plan with costed actions and owners.

Define the unit of analysis and freeze segmentation

Keep the lens commercial. Analyze resilience where dollars sit: by product family/SKU class, customer segment, channel, and geography. For each view, write one sentence that fixes scope and time window (e.g., “OTIF and CM2 for Key Value Items (KVIs) sold via big‑box retail in North America, last 12 months”).

Assemble the minimum viable data pack

Request once; analyze many times. You are looking for artifacts that tie supply mechanics to revenue and margin:

Customer service: OTIF/OTD (on‑time delivery), fill rate, backorders, perfect order rate, chargebacks/compliance fines, returns and warranty rates, damage claims, NPS/CSAT tied to delivery issues.
Inventory & planning: on‑hand by SKU and location; safety stock policies and parameters; days of supply/cover (DOH/DOS); reorder points and lead times; shortage codes; expedite frequency and spend; forecast accuracy (MAPE) and bias.
Supply base: approved vendor list (AVL) with tier‑1 and critical tier‑2; dual‑source status; geographic concentration; MOQs; contractual terms (lead‑time commitments, liability caps, indexation, take‑or‑pay, exclusivity, step‑down pricing); supplier OTIF/quality scorecards.
Cost stack: BOM and route sheets; commodity exposure; freight/parcel mix; surcharges and accessorials; 3PL/forwarder contracts; hedging policies; scrap/yield.
Capacity & flow: demonstrated vs. rated capacity; OEE for in‑house operations; changeover times; constraint/bottleneck steps; subcontracting and alternate routings.
Compliance & quality: certifications (e.g., ISO, UL, CE; for med‑tech, QMS/ISO 13485, UDI/DMR/DHR), inspection/FAI, CAPA counts and cycle time, supplier non‑conformances (DPPM), recalls or field actions.
Logistics: lane‑level transit times; carrier performance; port/airport reliance; cross‑border steps (broker, bond, duties); 3PL SLAs; parcel audit data.
Governance: S&OP cadence and artifact quality; EDI/WMS/TMS integrations; business continuity and disaster recovery plans (BCP/DR); cyber posture for connected plants and 3PLs (high level).

Work from the customer backward: service first

Before you score suppliers, quantify what service looks like to customers and how deviations hit revenue and retention.

Define and compute core metrics by segment and channel:
- OTIF = Orders delivered on time and in full ÷ Total orders.
- Fill rate = Shipped quantity ÷ Ordered quantity (line and order level).
- Perfect order rate = On‑time × In‑full × Error‑free × Damage‑free.
- Backorder rate = Backordered lines ÷ Total lines; age of backorder.
- Chargebacks/compliance fines (retail): frequency, $ per incident, root causes.
Tie service misses to outcomes: lost sales (cancellations), markdowns/promo to clear late product, churn/NRR drag in B2B subscriptions tied to SLAs, delistings in retail/marketplaces.

Identify “critical‑to‑revenue” SKUs and components

Not all items are equal. Create a shortlist that drives value and risk:

KVIs and high‑runner SKUs: items that anchor price image or a large share of sales.
Regulated/approval‑bound SKUs: where alternate sourcing requires requalification.
Long‑lead/unique components: custom ASICs, specialty resins, firmware‑bound modules.
Service enablers: spares and consumables that protect annuity revenue.

For each, fill a Critical Item Card: revenue share; current on‑hand and days of cover; suppliers and geos; lead time and MOQ; dual‑source status; quality/defect trend; logistics path; recovery options and time‑to‑recover.

Compute Revenue‑at‑Risk (RaR) and Time‑to‑Survive/Recover

Make risk quantifiable and comparable.

Time‑to‑survive (TTS): how long demand can be met after a disruption before stockout, given current inventory, in‑transit, and confirmed supply.
Time‑to‑recover (TTR): time needed to restore supply to ≥90% of demand after the disruption.
Immediate RaR (monthly) ≈ Σ (SKU revenue × Indicator[TTS < TTR] × (TTR − TTS)/30).
Model at least three shocks: loss of top tier‑1 supplier; port closure on a critical lane; demand surge of +20–30% on KVIs. Where data is thin, bracket with ranges and state assumptions.

Quantify Margin‑at‑Risk (MaR)

Inflation, surcharges, and expedites can erase price gains. Build a margin bridge from BOM to deliver cost:

Commodity and index exposure: which inputs float (resin, steel, energy) and are any contracts indexed?
Freight/parcel and accessorials: fuel surcharges, peak season fees, residential and oversized charges; detention/demurrage exposure in ocean.
Expedite/spot premium: last‑quarter expedite spend as % of COGS and its trend.
Yield/scrap: first‑pass yield and scrap costs at constraint steps.
Quality cost of poor quality (COPQ): rework, returns, warranty, and chargebacks.
MaR (quarterly) ≈ ΔCOGS from price/volume variance + freight/accessorial drift + expedite premiums − price indexation pass‑through. Tie directly to CM2.

Score the supply base and remove single points of failure

Use a simple, auditable rubric for tier‑1 (and critical tier‑2):

Concentration: supplier share of spend for critical items; compute HHI and flag when >1,800.
Alternate source status: dual‑source qualified (Y/N); time/cost to qualify new source.
Geographic clustering: common shock exposure (same city/port/EPZ) for “dual” sources.
Financial and operational health: backlog, lead‑time trend, capacity headroom, capex plan, labor stability.
Contractual protections: lead‑time commitments, allocation priority, indexation, safety‑stock/VMI, audit rights, right of first allocation.
Quality & compliance: DPPM trend, CAPA responsiveness, certifications.

Inspect planning maturity and inventory policy

Forecast error is often the root cause of expedites and stockouts.

Compute MAPE and forecast bias by family and channel; compared to service levels.
Review safety stock math (service‑level targets, variability inputs, lead‑time assumptions); spot “frozen” parameters.
Check S&OP cadence: monthly cycle with cross‑functional agreement, or spreadsheet tennis?
Confirm planning horizons: lock windows and firm zones; supplier collaboration (PO acknowledgments, ASN usage).
Assess systems: WMS/TMS/ERP integration, EDI health, and data latency to carriers/3PLs.

Probe logistics and cross‑border risks

Freight can quietly dominate both service and margin.

Lane‑level transit time and reliability; carrier performance and capacity commitments.
Port and gateway reliance; free‑trade zone usage; broker performance and holds.
Cross‑border compliance: classification discipline, valuation, origin, sanctions/forced‑labor checks; error rates driving penalties or seizures.
3PL performance: SLA adherence, cut‑off compliance, pick/pack accuracy, inventory accuracy, cycle counts, shrink.

Check quality, warranty, and regulatory posture

Quality issues are revenue shocks wearing a different hat.

Incoming and in‑process defect rates; final test escape rates; field failure rates and MTBF.
Warranty reserve adequacy; root‑cause cycle time; CAPA effectiveness.
Regulated categories: evidence of compliant QMS; audit outcomes; change‑control discipline; documentation (DMR/DHR) completeness.
For consumer/retail: test reports for mandated standards; packaging compliance and transit tests.

Turn findings into scenarios that change decisions

Convert your evidence into two or three decision‑relevant scenarios with modeled impact and clear triggers:

Tier‑1 supplier outage (6–12 weeks): revenue lost by segment, expedite cost, and recovery path; post‑close action—qualify dual source and build buffer stock.
Port/lanes disruption (2–4 weeks): reroute feasibility, transit time delta, cost delta, and OTIF impact.
Commodity/fuel spike (+15–25%): margin compression vs. indexation pass‑through and timing; price/elasticity overlay from Chapter 6.4.
Retail compliance enforcement: OTIF fines increase; delist risk on KVIs; remedial logistics actions.
Demand surge (+20–30% on a KVI): capacity headroom and time to flex; price/mix levers.

Write the 90‑day Readiness Plan (post‑close)

List only actions that move TTS/TTR, OTIF, or CM2 in the hold period:

Qualify a named second source for the top three critical items; preload safety stock to span TTR.
Implement CPI/commodity/fuel indexation in new contracts; add allocation priority and minimum notice clauses.
Lock KVI service levels with specific buffers (days of cover) and lane diversifications.
Reduce expedite dependency: demand‑planning parameter refresh; MOQ/lead‑time renegotiation; alternate routings.
Logistics fix: secondary gateway and carrier adds; 3PL SLA reset; late cut‑off capacity.
Compliance hygiene: forced‑labor screening, classification audit, broker SLAs; QMS gaps closed where regulated.
S&OP hardening: monthly cadence with a single set of numbers; forecast and inventory KPIs on the PMO dashboard.

Early‑warning indicators to monitor monthly

Wire the leading signals into the operating rhythm:

OTIF, fill rate, perfect order rate; backorder age and trend.
Expedite spend as % of COGS; spot freight mix; accessorial $/order.
Supplier lead‑time trend; PO acknowledge lag; supplier OTD/quality score.
Inventory health: days of cover by KVI; safety‑stock breaches; slow‑/no‑move.
Forecast MAPE and bias; S&OP adherence.
3PL metrics: pick accuracy, on‑time ship, inventory accuracy, shrink.
Compliance/chargebacks: frequency and $; root causes closed.
Sanctions/forced‑labor and regulatory alerts on critical geos or materials.

Red flags—and how to respond

Treat these as valuation and term‑sheet items, not footnotes:

Single‑sourced KVI with TTR > 8 weeks and days of cover < 3 → haircut base revenue for that segment; require buffer stock covenant or escrowed capex for dual‑source.
Expedite dependence (>2% of COGS) to hit OTIF → reflect in CM2 downside and enforce planning parameter reset post‑close.
Retail OTIF fines and delisting warnings → widen price/mix ranges; add covenant on OTIF threshold and a 3PL remediation plan.
Commodity/fuel exposure without indexation → push price power to upside only; insert indexation clauses as a closing condition where feasible.
Weak QMS/compliance in regulated lines → move those revenues to upside pending remediation; require specific certifications or audit clears as conditions precedent.
3PL concentration (>60% of outbound volume) without alternate capacity → model a lane‑specific shock; negotiate secondary capacity.

Acceptance criteria for a decision‑grade resilience review

Critical Item Cards complete for top revenue drivers; TTS/TTR and RaR/MaR calculated with assumptions and confidence labels.
Supply‑base score with concentration, dual‑source status, geo clustering, and contract protections—per critical item.
Logistics, planning, and quality metrics produced by segment/channel with ties to revenue and CM2; hygiene issues disclosed.
Two to three shock scenarios modeled with toggles in the driver‑based model; early‑warning indicators and triggers defined.
A 90‑day Readiness Plan with owners, costs, and measurable targets (OTIF, days of cover, expedite spend, CM2).
Clean‑team protocols followed; supplier identities and commercial terms shared only in aggregated form outside the ring fence; all exhibits footnoted and replicable.

Copy‑ready templates (use as you work)

Critical Item Card
- Item/SKU, revenue share, KVI (Y/N), regulated (Y/N).
- Suppliers (tier‑1, critical tier‑2), geos, dual‑source status.
- Lead time (ordered/actual), MOQ, contract terms (indexation, allocation priority).
- Inventory on hand, in transit, days of cover; safety stock; backorder.
- Quality trend (DPPM/returns), warranty signals.
- Logistics path (lanes, carriers, gateways); alternate routings.
- TTS/TTR estimate; RaR/MaR; action and owner.
Supplier Risk Card
- Spend share and HHI contribution; financial/operational health cues.
- Lead‑time and OTD trend; capacity headroom.
- Quality (DPPM, CAPA responsiveness); certifications.
- Contract terms (indexation, allocation, penalties, VMI/consignment, audit rights).
- Geo/climate/political exposure; common‑mode failures with other suppliers.
- Risk rating (H/M/L), mitigation (dual‑source, safety stock, qualification plan), owner.
Logistics Lane Card
- Origin/destination, mode, carrier(s), 3PL/DC, average transit and p90, reliability.
- Volume share, peak constraints, cut‑off adherence.
- Cost per unit and accessorial mix; spot vs. contract share.
- Alternate lanes/carriers; activation time and cost.

72‑hour sprint plan (from blank page to decision‑grade view)

Day 0: Freeze scope and segments; issue the data request; confirm clean‑team boundaries.
Day 1: Build the service view (OTIF, fill, backorder, chargebacks) and inventory health by KVI; draft the first five Critical Item Cards.
Day 2: Score top suppliers; compute TTS/TTR and first‑pass RaR/MaR; map two logistics lanes with performance and cost.
Day 3: Model two disruption scenarios and one inflation/fuel scenario; publish early‑warning indicators and the 90‑day Readiness Plan; push RaR/MaR toggles into the financial model and propose term‑sheet levers (indexation, inventory covenants, dual‑source milestones).

Run this checklist with discipline and you will replace vague comfort about “strong operations” with a quantified resilience view: what threatens service and margin, how big the exposure really is, and the handful of actions that raise confidence in the growth and price story—before you close and while you still have leverage to protect value.

10.2 Capacity and Scalability Analysis – Step-by-Step Guide

Capacity and scalability analysis connects the commercial plan to the real world of plants, partners, people, and platforms. The core question is simple: can the company meet the volumes, mix, service levels, and geographies in the plan—at the promised margin—without relying on heroics? The answer requires a disciplined walk from demand to the bottleneck, translating engineering and staffing metrics into revenue‑at‑risk and CM2 impact. Treat capacity as a system: physical assets (make/move), human capacity (implement, service, support), and digital capacity (cloud, data, integration limits). Then size headroom, time‑to‑add, and cost‑to‑add by segment and route‑to‑market.

Principles to anchor the work

Capacity is effective, not nameplate. Adjust for uptime, changeovers, learning curves, quality yield, staffing, and supplier reliability.
Bottlenecks drive throughput. Non‑constraints don’t set output; measure and manage the constraint first.
Utilization beyond ~80–85% at the constraint causes non‑linear wait times; backlog and cycle times balloon before output does.
Step‑fixed costs and lead times matter. Most capacity comes in steps (new shift, tooling set, cluster, or partner) with real commissioning time.
Commercial consequences come first: when capacity is short, service levels drop, realized price slips (discounts, chargebacks), and retention suffers.

Step 1 — Freeze demand scenarios and service levels

Start with the plan the deal depends on. Translate growth into concrete “demand objects” by segment: units per SKU family, installs per month, implementation hours, tickets per active customer, API calls per second, orders per lane. Fix the service promise (OTIF, lead‑time quote, SLA/uptime). Define three demand cases—base, surge (+20–30% on key value items or peak weeks), and mix shift (toward the products/segments with the highest observed growth). These scenarios will “drive” your model through the capacity stack.

Step 2 — Map value streams and name the constraint

Create a fast, visual map from order to cash for each major product/service family. Mark process steps, resources, and queues: sourcing → fabrication/assembly → test/pack → ship, or sign‑up → implementation → go‑live → support. Identify candidates for the constraint (machine center, testing cell, skilled role, specific partner tier, database or third‑party rate limit). If you cannot state “the constraint is X in stream Y,” you are not ready to do math.

Step 3 — Assemble the capacity data pack (minimum viable but decision‑grade)

Physical operations: rated speeds, shift patterns, planned maintenance, changeover times, scrap/rework, first‑pass yield, OEE history, labor rosters and cross‑training, overtime history and policy, subcontract options, test cell throughput, pack/ship cut‑offs, and 3PL wave capacity.
Supply inputs: critical component lead times and MOQs, supplier OTIF, dual‑source status, and time‑to‑qualify alternates (tie to 10.1).
People‑based services: implementation/project templates with hours by role, active pipeline of projects, FTE counts by skill, utilization and shrinkage (vacation, training, admin), attrition, time‑to‑hire and time‑to‑productivity by role, partner bench capacity.
Digital platform: current and p95 CPU/memory/IO utilization, request concurrency and RPS, database QPS and locks, cache hit rates, autoscaling policies and warm‑up times, third‑party API limits (payments, messaging, data), multi‑tenant “noisy neighbor” controls, and SLO breach rates.
Cost and timing: capex for adds (tooling, lines, environments), install/commission lead times, contract terms with partners/cloud (reserved vs. on‑demand), and any regulatory or certification gates.

Step 4 — Compute effective capacity where it really lives

Convert raw capability into continuous, comparable units for each stream.

Manufacturing/fulfillment: Effective hourly capacity = rated speed × availability × performance × quality yield (OEE). Adjust for changeovers by using an effective cycle time at the planned mix. Convert to weekly/monthly capacity by multiplying by staffed hours and planned shifts; show headroom at 80–85% utilization.
Service/implementation: Effective monthly capacity (by role) = FTE × available hours × utilization × productivity factor. Deduct shrinkage (holidays, training, meetings). For multi‑skill tasks, compute capacity in “standard hours” and reflect the critical role as the constraint.
Support/tickets: Effective capacity = FTE × available hours × productivity; cross‑check with historical tickets per active account and seasonality.
Digital/cloud: Effective capacity is defined by the first SLO‑limiting resource at p95 (database, message broker, backend). Use current p95 loads and autoscaling policies to estimate safe headroom without violating latency/uptime targets. Compute cost per incremental request or tenant at that headroom.

Step 5 — Compare demand to capacity with simple, defensible math

Takt time vs. cycle time: Takt = available production time ÷ demand. If the bottleneck cycle time > takt, you will miss.
Little’s Law (sanity check): WIP/Queue = Throughput × Flow time. If flow time or WIP explodes at current throughput, the constraint is saturated.
Headroom: Headroom% = (Effective capacity − Required demand) ÷ Effective capacity. Keep a cushion at the constraint; running >85% invites reliability issues.
Mix sensitivity: Recompute effective capacity at the “surge” and “mix shift” cases; many lines break at pack/test or at the one role the staffing model undercounts.

Step 6 — Build a capacity ladder (what you can add, how fast, and at what cost)

Separate the “no‑regrets” operational levers from step‑changes, and tie each rung to time‑to‑impact and $.

Zero‑to‑low cost (0–12 weeks): schedule smoothing, sequence optimization to reduce changeovers, cross‑training, modest overtime caps, preventive maintenance catch‑up, debottleneck test or pack, second carrier/gateway activation, enable autoscaling with pre‑warm pools, raise API concurrency with caching.
Step‑fixed (1–2 quarters): add a shift, buy duplicate tooling/fixtures, hire and ramp 3–10 key roles, qualify a subcontractor for a specific step, add DB read replicas, shard a hot dataset, add edge cache/CDN, extend 3PL wave windows.
Strategic (2–6 quarters): new line or cell, major automation, new facility or 3PL, full second‑source qualification for critical components, replatform or multi‑region cloud, partner program overhaul.

For each rung, complete a Capacity Add Card: bottleneck addressed, capacity delta, time‑to‑benefit (calendar), one‑time and run‑rate cost, quality or SLA risk, and dependencies (e.g., supplier tooling, regulatory validation, security review).

Step 7 — Price capacity realistically: ramp curves and learning

Capacity is not a light switch. Apply conservative ramps:

Hiring ramp: time‑to‑productive by role (e.g., 60–120 days for implementation consultants; 30–60 days for L1 support). Include attrition backfill.
Equipment ramp: delivery → install → qualification → yield stabilization (often 8–16 weeks even for “simple” adds).
Learning curve: unit time or scrap improves with cumulative output; use a progress ratio (e.g., 85–95%) if historicals support it.
Cloud ramp: scaling features ship fast but stabilize slower; allocate sprints for load testing, failure injection, and rollback plans.

Step 8 — Translate constraints into commercial impacts within 24 hours

Push three numbers into the deal model by segment/channel:

Revenue‑at‑risk (capacity‑limited): units that cannot be shipped/installed × realized price; add the retention or delist risk if the item is a KVI or SLA‑anchored.
Margin‑at‑risk: expedite premiums, overtime, yield loss, spot freight, cloud overages; show CM2 impact at required volumes.
Lead‑time and SLA changes: service penalties, OTIF fines, and price concessions needed to sustain orders when lead times extend.

Where uncertainty remains, widen ranges and propose structure (e.g., inventory or dual‑source covenants, capex escrows, earnouts keyed to activation capacity or NRR in capacity‑constrained segments).

Step 9 — Run three stress tests that actually change decisions

Demand surge on KVIs (+20–30%) for two months: does the constraint hold without SLA breaches or ruinous expedites?
Up‑mix to complex variants or premium tiers: do changeovers and test time collapse headroom?
Partner/platform shock: marketplace or 3PL cuts capacity; cloud region outage or third‑party API throttling; what’s the time‑to‑recover and the revenue at stake?

Step 10 — Service and implementation capacity (often the hidden constraint)

For subscription and enterprise businesses, the gating factor is frequently people and calendar slots, not machines.

Implementation: build a role‑by‑role hours model from the last ten projects; compute throughput (projects/month) at current staffing and utilization; check backlog age and slippage. Add partners as a parallel path with realistic productivity and quality multipliers.
Customer success and support: ensure coverage ratios (accounts per CSM, tickets per agent), queue time, and resolution SLOs do not degrade with planned cohort adds; rising ticket backlog is an early churn signal.
Training and certification: count certified personnel for regulated installs; quantify how many are needed to hit the plan by geography.

Step 11 — Digital scalability (how software really breaks)

Throughput and latency: runbooks should define safe p95 limits for core services; assess headroom at those limits.
Stateful bottlenecks: databases and queues throttle more than stateless computers; if QPS at the primary nears limits, sharding/partitioning work belongs in the capacity ladder.
Third‑party limits: payment gateways, messaging providers, and data vendors impose RPS caps and rate limiting; treat these as external bottlenecks with their own add cards and SLAs.
Cost curves: autoscaling hides limits but raises unit cost; model the step from on‑demand to reserved commitments and the savings ramp.

Step 12 — Governance and cadence (so capacity doesn’t become a quarterly surprise)

Establish a lightweight operating rhythm that your sponsor can keep post‑close:

Monthly S&OP with one set of numbers: demand, supply, capacity, and inventory or backlog.
A constraint dashboard for each value stream: utilization at the constraint, queue length, cycle time, yield/defects, and service/SLA hits.
A change‑control gate: any promotion, pricing move, or big deal that changes mix/volume must pass a capacity check before launch.

Early‑warning indicators to put on the PMO wall

Constraint utilization >85% three weeks running; rising queue length or WIP; expedite spend trending up.
Rising changeover hours per week; climbing scrap/rework at the constraint; test/inspection backlog.
Implementation backlog age >45–60 days; certified headcount below plan; partner activation slippage.
Ticket backlog and first‑response times drifting; SLA credits issued; p95 latency creeping toward SLO.
Supplier lead times stretching; 3PL wave cut‑off misses; cloud bill spikes without matching revenue.

Red flags—and immediate responses

Single‑point constraint with no ladder: pay for upside only in that stream; require a funded capacity add as a closing condition.
Plan counts on overtime as strategy: cap at sustainable levels; price in quality and attrition costs; move capacity adds forward.
Implementation capacity underestimates ramp: push bookings to upside or phase roll‑outs; add partner capacity with guardrails.
Digital limits ignored: database hot shards or third‑party API caps threaten SLOs; add sharding, caches, or provider tiers to Day‑1 priorities.
Changeover‑heavy mix: if premium variants explode changeover time, shift pricing/pack architecture or sequence rules; otherwise margin will erode.

Acceptance criteria for a decision‑grade capacity analysis

Named constraints by value stream with effective capacity calculated (not nameplate) and headroom shown at the constraint.
Three demand scenarios run through the stack, with revenue‑ and margin‑at‑risk quantified by segment/channel.
A capacity ladder with time‑to‑impact and cost‑to‑add for the top two constraints; owners and dependencies clear.
Implementation/support and digital capacity assessed alongside physical operations; partner and third‑party limits included.
Model updated within 24 hours with capacity‑gated revenue, CM2 effects, and lead‑time/SLA implications; residual uncertainty translated into terms or covenants.
Early‑warning indicators and S&OP cadence defined; exhibits footnoted and replicable under clean‑team rules.

Copy‑ready templates you can lift into your workbench

Constraint Card
Constraint resource; value stream; current effective capacity and utilization; headroom; quality/yield; cost to run at surge; revenue/CM2 at risk; next rung on ladder; owner.
Capacity Add Card
Bottleneck addressed; capacity delta (units/hour or projects/month); one‑time cost; run‑rate cost; time‑to‑benefit; enabling actions (tooling, hires, qualification); risks; decision gate.
Implementation Capacity Sheet
Standard hours per project by role; active FTEs, utilization, productivity; pipeline of projects by start month; throughput; backlog age; partner contribution; ramp plan.
Digital Capacity Sheet
p95 load vs. limit for compute, DB, messaging, search, and third‑party APIs; autoscale policies; SLOs; safe headroom; cost per incremental request; mitigation items with effort.

72‑hour sprint plan (from blank page to a defendable view)

Day 0: Lock demand scenarios and service/SLA targets; draw value‑stream maps; nominate likely constraints.
Day 1: Land the capacity data pack; compute effective capacity and headroom at each candidate constraint; validate with leads on the floor and platform runbooks.
Day 2: Run base/surge/mix scenarios; quantify revenue‑ and margin‑at‑risk; draft Constraint and Capacity Add Cards; cost and time the top two rungs.
Day 3: Update the financial model; publish the capacity ladder and early‑warning dashboard; propose term‑sheet protections (e.g., indexation, dual‑source milestones, capex escrow) and the 90‑day post‑close plan.

Run this guide and you will convert “we can handle the growth” into a precise, time‑phased view of what can ship, what could break, what it costs to prevent, and how quickly you can add headroom—by stream, by segment, and with numbers that hold up in Investment Committee.

10.3 Service-Level Performance Template

Service levels are where promises meet cash. Miss them and you pay—in chargebacks, service credits, price concessions, delistings, churn, or reputational drag that raises CAC. Hit them reliably and you unlock price power, distribution, and expansion. This template gives you a fast, defensible way to measure service performance the way buyers do, link it to unit economics, and set a short list of fixes you can execute inside the hold period. Keep it segmented by product family, customer tier, channel, and geography; averages hide the truth.

Start by separating three ideas that often get conflated. Service Level Indicators (SLIs) are the measurements (e.g., “p95 API latency,” “OTIF by order line,” “First Response Time”). Service Level Objectives (SLOs) are the internal targets (e.g., “p95 latency ≤ 300 ms, monthly”). Service Level Agreements (SLAs) are the contractual promises, with penalties, credits, or delisting risk if breached. In diligence, you will compute SLIs, compare them to SLOs, and price the SLA exposure.

Define your vocabulary and freeze the scope

Choose the buyer‑visible metrics that actually decide renewals, shelf space, or deal wins. Write one sentence per view—for example: “OTIF, perfect order rate, and chargebacks for big‑box retail in North America, LTM,” or “Availability, p95 latency, and case response/resolution SLAs for enterprise SaaS in EMEA, last 8 quarters.” Lock the measurement window (rolling 12–24 months), period (weekly/monthly), and the segmentation you established in Chapter 6.

Assemble a minimum viable data pack (once, use everywhere)

Contracts and policies: SLAs, service credit schedules, OTIF and compliance guides (retail/marketplace), SOWs for implementations, support plans by tier, penalty ladders, delisting rules.
Operational systems:
- Fulfillment: WMS/TMS/ERP order lines with commit date, ship date, delivered date, quantity ordered/shipped, damage/return codes, chargebacks.
- Support/Success: ticketing (created/first touch/resolve timestamps, severity, channel), case ownership, backlog age, CSAT.
- SaaS/platform: telemetry (availability by region, p95/p99 latency, error rates), incident logs, deployment calendar, third‑party API rate‑limit and outage records.
- Field/service/implementation: work orders, arrival windows, time‑to‑repair, first‑visit fix flags, project plans vs. actual, milestone acceptance.
- Finance: credits/waivers, SLA penalties, OTIF fines, markdowns tied to late deliveries.
Context: promotions, price changes, peak calendars, inventory constraints, carrier/3PL SLAs, change freezes, product releases.

Metric taxonomy that buyers recognize (with formulas)

Select the subset that fits your model; do not boil the ocean.

Fulfillment & logistics (retail, hardware, CPG, marketplaces)
- On‑Time In‑Full (OTIF) = (# orders delivered by promised date AND in full) ÷ (total orders).
- Fill rate (line‑level) = shipped qty ÷ ordered qty.
- Perfect order rate = on‑time × in‑full × error‑free × damage‑free.
- Backorder rate = backordered lines ÷ total lines; track backorder age.
- Chargeback rate = chargeback count ÷ shipments; $/order for fines.
- Return rate and damage rate by reason code (defect vs. carrier).
SaaS / platforms / APIs
- Availability = 1 − (total downtime ÷ total time) over the SLA window. Report by region/AZ if relevant.
- Latency: p95 and p99 request latency by endpoint and region.
- Error rate: 5xx/4xx shared by endpoint; SLOs typically exclude user errors (4xx) but include 5xx.
- Change failure rate and MTTR (mean time to recover) from incidents.
- Third‑party dependency health: % time under partner API rate‑limit or outage.
Support / Customer success
- First Response Time (FRT) and Time to Resolution (TTR) by severity.
- SLA attainment = (# cases meeting SLA ÷ # cases with SLA).
- First Contact Resolution (FCR) and ticket backlog (open > X days).
- Contact rate = tickets ÷ active customers; a leading indicator of churn.
Field service / Implementation / Professional services
- Appointment on‑time rate, First‑visit fix rate, Mean Time To Repair (MTTR).
- Time to go‑live vs. statement of work; schedule variance and rework hours.
- Billable utilization where it constrains throughput (connect to 10.2).

Measurement rules that keep numbers defensible

Clock start/stop: define “on‑time” explicitly (carrier scan vs. customer delivery, order commit vs. requested date). For support, define first response (bot vs. human) and resolution (workaround vs. permanent fix).
Partial shipments and split orders: choose to count as “in‑full” only if all lines meet promised dates; otherwise tag partials separately.
Percentiles, not averages: report p50/p95 for latency and response times; averages hide tail pain.
Severe vs. non‑severe: weight breaches by severity (P1 vs. P3) when pricing credits or churn risk.
Attribution discipline: tie chargebacks and credits to root cause (your warehouse, carrier, customer ASN errors) to avoid over‑correcting.

Build the baseline and link it to money

Compute monthly SLIs for the last 8–12 quarters by segment and channel; show mean and dispersion.
Compare SLIs to SLOs (internal) and SLAs (contractual). Quantify breach frequency and duration.
Translate breaches to dollars:
- Service credits / penalties: sum actuals; model exposure from breach bands (“if availability falls to 99.5–99.8%, credits = 5% of MRR for affected tenants”).
- Chargebacks/compliance fines: $ per breach × incidents; include delisting risk in marketplaces/retail.
- Margin drag: expedites, overtime, dual‑ship, extra inventory handling.
- Revenue & retention: cancellations, lost orders, downgrade rates, and observed churn hazard deltas after breaches (use 6.3 hazard methods).
Produce a Service‑to‑Economics Bridge for each top segment showing how a 1‑point change in OTIF, availability, or response time moves CM2 and NRR.

Tiering and promises—match the offer to the engine

If the target sells Gold/Platinum tiers with different SLAs, verify that capacity and process actually support them (10.2). Confirm price fences, eligibility, and enforcement (8.2). For marketplaces, confirm that seller tier benefits (search rank, paid placement, fees) are earned and at risk if SLIs slip.

Design a simple, auditable scorecard

Use this one‑pager per segment or customer tier.

Service‑Level Performance Card (copy and fill)
- Scope: [segment, channel, geo, period].
- Contracted SLAs: [metrics, thresholds, credit ladder].
- Observed SLIs (LTM): Availability __%, p95 latency __ ms; OTIF __%; perfect order __%; FRT/TTR by severity; appointment on‑time __%; first‑visit fix __%.
- Breach summary: [count, duration, severity mix]; credit/penalty $; delisting warnings Y/N.
- Root‑cause mix: [inventory, capacity, carrier, third‑party API, change failure, customer EDI].
- Economic link: CM2 drag $; NRR/GRR impact (hazard + pts) with evidence; price concessions $__.
- Risk scenarios: [e.g., peak weeks, carrier change, API throttle].
- Remediations & owners: [top 3, time‑to‑impact, cost].
- Confidence & sources: H/M/L; system logs, WMS/TMS extracts, ticketing IDs, APM traces.

Governance and cadence

Introduce a monthly SLO review that pairs product/ops with sales/CS and finance.
Use an error‑budget concept for SaaS: if availability or latency burns the monthly budget, freeze high‑risk changes and prioritize reliability work.
For retail/marketplaces, run a weekly OTIF/chargeback clinic with 3PLs, carriers, and top accounts; close root causes within a defined SLA.
Tie incentives: managers own SLOs; deal desks honor price floors only when SLOs stay green.

Early‑warning indicators (watch weekly)

OTIF < target by ≥2 pts, or backorder age rising.
p95 latency drifting toward SLO; rising 5xx error rate; incident MTTR lengthening.
Case backlog age > threshold; FRT/TTR slips on P1; rising contact rate per account.
Appointment on‑time slippage; first‑visit fix rate downtrend.
Chargeback frequency, SLA credit accruals, or waiver requests trending up.

Red flags—and what to do now

Chronic SLA credits with no roadmap item: move price power to upside only; require a reliability plan as a closing condition.
OTIF fines on KVIs or delisting warnings: widen revenue ranges; add inventory/3PL covenants and a lane‑diversification action.
Availability claims without third‑party monitoring: lower confidence; require independent uptime verification post‑close.
Support “green” while churn rises: compute contact rate and backlog age; many dashboards hide tail pain—cut by severity and p95.
Implementation overruns burning capacity: tie bookings to implementation slots; add partner capacity with guardrails (10.2).

Push results into the model within 24 hours

Add SLA credit and chargeback lines to CM2 by segment/channel.
Adjust price realization corridors where service misses force concessions (8.2).
Introduce NRR/GRR drag in segments with proven hazard after breaches (6.3).
For SaaS, reduce new ARR recognition in periods with change freezes or backlog; for retail, reduce sell‑through where delist risk is flagged.
Encode scenario toggles: peak‑week OTIF −3 pts, availability −0.2 pts, or carrier switch; show CM2/NRR sensitivity.

Quality controls you must enforce

Two‑source rules: pair system logs or scans with finance credits/chargebacks for each breach band.
Time‑zone and clock alignment across WMS/TMS/APM/ticketing.
Partial shipment and severity rules stated and applied consistently.
Incident and ticket deduplication; suppress bot auto‑responses as “first touch” unless contracted.
Verifier re‑computation from raw extracts before any exhibit leaves the workbench; clean‑team protocols for customer‑specific data.

Industry nuances to handle explicitly

Healthcare and regulated: TAT (turnaround time) for specimens or prior auths, adverse‑event reporting timeliness, and QMS audit trails are service levels with regulatory teeth—treat breaches as valuation and closing‑condition issues.
Marketplaces: buyer cancel rate, late shipment rate, refund/return rate, and seller response time determine search rank and fee tiers; delisting risk is a step‑function, not a slope.
Usage‑based SaaS: measure active‑rate and intensity alongside latency/availability; a reliability dip often shows up as intensity decay before logo churn.
Hardware + service: protect the annuity; spares fill rate and MTTR drive consumable/service ARPU more than new device OTIF.

Copy‑ready checklists and mini‑templates

Service Data Hygiene Checklist

Clock definitions documented (on‑time, first response, resolution).
Partial shipment policy applied consistently.
Percentiles (p95/p99) used for latency/response; averages flagged as supplemental.
Root‑cause taxonomy implemented; carrier vs. warehouse vs. customer EDI separated.
Ticket deduplication and bot filters applied; severity mapping aligned to contracts.
Finance tie‑out for penalties/credits complete; verifier sign‑off logged.

SLA Credit Exposure Card (per contract/tier)

SLA metric and threshold; credit ladder.
Observed breaches (count, duration, severity).
Credits taken (LTM) and modeled exposure (+/− bands).
Root cause; mitigation; owner; due date.

Peak Readiness Checklist (retail/marketplaces)

KVI safety stock days; secondary carrier/gateway activated; late cutoff capacity.
ASN accuracy and label compliance checks complete.
OTIF dashboard on hourly cadence; chargeback dispute playbook ready.

Reliability Sprint Template (SaaS/platform)

Top 3 incident patterns; error budget burn; rollback coverage; load test schedule.
Dependency SLAs (payments, messaging, data) verified; rate‑limit alarms added.
Release freeze criteria; on‑call staffing; comms templates.

72‑hour sprint plan (from blank page to decision‑grade)

Day 0: Freeze scope and segmentation; issue data request (contracts, operations, finance logs); document measurement rules (clock starts, partials, severity).
Day 1: Compute SLIs for top segments; compare to SLOs/SLAs; quantify credits/fines and margin drag; build first two Service‑Level Performance Cards.
Day 2: Link breaches to revenue and retention (hazard deltas); run one peak and one dependency scenario; draft remediation actions with time‑to‑impact and cost.
Day 3: Update the financial model (SLA credits, chargebacks, CM2 drag, NRR/GRR); publish early‑warning indicators and a 90‑day go‑to‑green plan; propose term‑sheet levers (indexation, inventory/3PL covenants, uptime verification, milestone‑based earnouts).

Acceptance criteria for a decision‑grade service review

Contracted SLAs, internal SLOs, and observed SLIs reconciled by segment/channel with definitions on the page.
Breach frequency and financial impact (credits, fines, concessions) quantified and tied to CM2 and NRR/GRR.
Root‑cause mix established; at least two decision‑relevant scenarios modeled with toggles in the plan.
A 90‑day remediation plan with owners, time‑to‑impact, and cost; capacity dependencies aligned with 10.2.
Early‑warning indicators defined and wired into the operating cadence.
Verifier tie‑out to raw extracts complete; clean‑team protocols observed.

Use this template and “great service” stops being a slogan. You’ll have a buyer‑true picture of performance, the dollar impact of misses, and a practical path to raise reliability where it moves price, shelf space, win rate, and retention—inside the horizon that matters for the deal.

10.4 Cost-to-Serve Benchmark Checklist

Cost‑to‑serve (CTS) is the money you spend after you win the order to fulfill, implement, support, and retain that revenue—by segment and route‑to‑market. In diligence, CTS is the hinge between a beautiful gross margin and the CM2 that actually lands. Done right, a CTS view explains price corridors (8.2), clarifies unit economics (8.3), exposes channel realities (9.2), and quantifies how service promises (10.3) translate into cash. The objective in a sprint is not an academic activity‑based costing time; it is a defendable, segmented CTS build you can plug into the model and use to set immediate margin and policy moves post‑close.

What CTS covers (anchor this definition up front)

Think “everything variable or semi‑variable required to fulfill, deliver, and keep the customer productive,” excluding product COGS already in CM1. Typical CTS buckets:

Fulfillment & logistics: pick/pack labor, carton/dunnage, labels, parcel/line‑haul/last‑mile, accessorials (fuel, oversized, residential), sortation, cross‑dock, brokerage, duties, storage for fast turns, outbound 3PL fees.
Returns & reverse logistics: return shipping, inspection/screening, refurbish/scrap, re‑stock, refund/chargeback handling, re‑pack, secondary market markdowns.
Payments & risk: processor/assessment fees, chargebacks/disputes, fraud losses, KYC/AML checks where applicable.
Customer care & success: contact center minutes, tiers/queues, success/CSM time, onboarding aids and training consumables.
Field/implementation/service: travel time & miles, on‑site labor, spares, test instruments, third‑party contractors, project management and cut‑over nights.
Platform/compute & third‑party APIs (for digital): cloud usage (compute/storage/network/DB), observability, CDN, per‑message/per‑transaction fees to providers (payments, SMS, data).
Compliance & penalties: OTIF/chargebacks (retail), SLA credits (SaaS), warranty claims processing.
Channel programs: co‑op/MDF you treat as post‑sale support, paid placement tied to service, marketplace support fees.

Everything here is below your price waterfall’s vendor‑realized revenue (8.2) and above corporate overhead. Put overhead in a separate analysis; do not bury it in CTS.

Step‑by‑step: build a defendable CTS in a diligence sprint

1) Freeze your scope and segmentation
Write a one‑line definition per view: “CTS per order for DTC US, AOV $80–$120, LTM,” “CTS per enterprise tenant per month, EMEA hosted, Gold support,” or “CTS per reseller order in NA distribution.” Apply the segmentation from Chapter 6 (customer/job, size band, channel, region, product family). Use the same period and currency as your pricing and CM2 work.

2) Assemble the minimum viable data pack (request once)

Operations systems: WMS/TMS pick/pack timestamps and units handled; parcel and LTL invoices with accessorials; 3PL bills; returns reasons and routes; ACD/call center minutes; ticketing/CS workflow timestamps; field service/PS time logs; cloud provider invoices by service; third‑party API usage and caps.
Finance: GL/ledger extracts for shipping, payment fees, fraud/chargebacks, warranty, SLA credits, OTIF fines, refunds; accrual vs. cash timing.
Contracts: 3PL/carrier rate cards, rebate ladders and surcharges, marketplace take‑rate and paid placement terms, payment processor schedules, warranty policies, SLA credit ladders.
Drivers: order lines per order, items per carton, DIM weight mix, distance/zones, service levels, return rates by reason, contact rate per account, implementation standard hours by role, active user intensity, API calls per tenant.

3) Choose a method—and keep it simple and auditable
Use time‑driven ABC (TDABC) for people/process steps and invoice‑based actuals for bought services.

TDABC rate: Cost per minute = fully loaded hourly labor cost ÷ (paid minutes × utilization). CTS for an activity = minutes per transaction × cost per minute × volume.
Invoice‑anchored cost: use carrier, processor, cloud, and 3PL invoices as the source of truth; reconcile to GL. Avoid model‑only rates where actuals exist.

4) Define the cost equation per segment (write it down explicitly)
For a transactional order:
CTS_per_order = PickPack + Materials + Outbound Freight + Accessorials + Payment Fees + Fraud/Chargebacks + Customer Care Minutes × Rate + Returns (probability × cost) + SLA/Compliance Penalties (probability × cost).
For subscription/usage:
CTS_per_account_month = Hosting/Cloud + Support Minutes × Rate + Success Hours × Rate + Third‑Party API Fees + Payment Fees + SLA Credits (expected) + Trust & Safety.
For enterprise project work:
CTS_per_project = Implementation Hours × Role Rates + Travel & Expenses + Cut‑over Overtime + Spares/Tools Consumed + Rework (expected).

5) Compute corridors, not single points
Produce medians and interquartile ranges for CTS per order/account/project and as % of vendor‑realized revenue (8.2), by segment and channel. Corridors reveal mix and routing noise; single numbers get weaponized.

6) Tie CTS to service and price—make the economics visible
Overlay OTIF/latency/FRT from 10.3 to show where higher CTS actually protects revenue (e.g., premium shipping on KVIs) versus where it is pure waste (expedites caused by planning errors). Map returns and contact rate to churn hazard for recurring models (6.3). Then connect to price ladders: if Gold support includes 24/7 and a 1‑hour SLA, price fences must reflect its CTS delta.

7) Reconcile to CM2 and GL (the credibility pass)

Ensure numerator/denominator alignment: CTS per order must divide total CTS dollars by the correct order count for that segment and period.
Run a GL tie‑out: the sum of CTS buckets equals the shipping/handling, support, warranty, SLA credit, payment fee, fraud, and cloud lines in the ledger (after pulling COGS out). Explain variances (timing, accruals).
Apply ratio‑of‑sums vs. sum‑of‑ratios discipline: roll‑ups should be ratio‑of‑sums to avoid small‑order bias.

8) Benchmark with guardrails—directional ranges to frame debate
Use these corridors to sanity‑check your build. Tighten them with your data by segment; do not treat them as absolutes.

DTC e‑commerce (mid‑weight, AOV $50–$150): outbound freight and accessorials typically 6–12% of revenue; pick/pack/materials 2–5%; payment processing 2–4%; returns and reverse logistics 3–8% (can spike well higher in apparel); customer care 0.5–2% depending on contact rate.
B2B distribution: outbound freight 3–6%; warehouse handling 2–5%; order entry and care $2–$6 per line; returns 1–3% of revenue.
SaaS (mid‑market): hosting/third‑party APIs 5–12% of revenue; support & success 5–10% for standard tiers; implementation is non‑recurring and should be treated outside run‑rate CTS (but include for payback).
Usage/API platforms: cloud and third‑party fees move with intensity; total platform CTS often 10–25% of revenue with sharp step‑downs when reserved capacity replaces on‑demand.
Marketplaces: payment fees 2–3% GMV; trust & safety 0.5–1.5% GMV; customer support 1–3% GMV; returns/refunds net 2–6% GMV in physical goods categories.
Hardware + service: field service CTS 8–20% of service revenue (travel is the swing factor); spares logistics 3–6%; warranty claims 1–3% of product revenue for stable lines—outliers demand root‑cause action.

If your computed CTS sits well outside these bands, either you’ve found a lever—or a data issue.

9) Translate CTS into levers with dated, costed actions

Policy & pricing: set free‑shipping thresholds by AOV and DIM weight; add return fees for buyer’s remorse in low‑margin categories; monetize premium SLAs; install CPI/fuel indexation where allowed.
Network & routing: second DC or carrier to reduce zones and accessorials; zone‑skip; consolidation rules; late cut‑off capacity in peak.
Process & tech: slim packaging to reduce DIM; auto‑deflect contacts with self‑serve flows; triage to asynchronous channels; returns triage and refurbishment playbooks; trust & safety automation; reserve cloud capacity to cap unit cost.
Partner terms: tighten 3PL SLAs, paid placement caps, and audit rights; payment processor schedule optimization; marketplace fee tiering quid‑pro‑quo (rank vs. returns).

10) Wire CTS into valuation, terms, and the model within 24 hours

Update CM2 lines with CTS corridors by segment and channel.
Add scenario toggles: +200 bps fuel surcharge, return rate +300 bps, payment fee +50 bps, call minutes +20%, cloud overage at on‑demand.
Where uncertainty is material, reflect it in structure: covenants on promo/take‑rate intensity, indexation clauses, caps on expedite spend, 3PL remediation milestones, earnouts tied to NRR or gross margin after shipping.

Activity driver library (pick a small set that actually explains 80% of CTS)

Order complexity: lines per order, items per carton, hazardous/temperature control, custom pack.
Distance & speed: zones/miles, premium service share, delivery window precision.
DIM & weight: billed weight vs. physical; oversize/long‑length penalties.
Customer behavior: return probability by reason; contact rate per 100 orders; abuse/fraud propensity.
Digital intensity: API calls per tenant, data processed per session, concurrent users.
Implementation difficulty: sites/users per project, integrations count, required certifications.
Channel: marketplace vs. direct vs. distributor (policy/fee differences); retail compliance.

Quick worked mini‑examples (for calibration)

DTC apparel, AOV $90: outbound freight $7.80 (8.7%); pick/pack/materials $2.70 (3.0%); payment fees $2.43 (2.7%); returns 22% with $5.00 handling average → $11.00 expected reverse cost per sold order (12.2%); care $0.60 (0.7%). CTS ≈ 27.3% of revenue before SLA/chargebacks. Free‑shipping threshold increase and fit tools reduce returns by 400 bps → ~4 pts CM2 recovery.
Mid‑market SaaS, Gold tier: cloud and APIs $2.50 per $25 ARPU (10%); support 15 minutes/month at $1.20 per minute fully loaded ($18.00) → 7.2% of revenue; success 0.25 hours/month at $60/hour ($15.00) → 6.0%. CTS ≈ 23% of revenue. Without a Gold premium or price fence, price power is being left on the table.
Enterprise implementation: 180 standard hours @ blended $110, travel $2,400, spares $600, rework expected $1,000. Project CTS $23,400; if billed services are discounted or “thrown in,” payback stretches—move to productized packages or minimum commits.

Data hygiene and QA checklist (use this verbatim)

CTS scope and segmentation written on page; period and currency aligned with price and CM2.
GL tie‑out performed; timing differences explained; accruals vs. cash labeled.
Carrier/processor/cloud invoices used where available; not modeled rates.
Labor rates built from fully loaded costs and realistic utilization; shrinkage included.
Partial shipments, split orders, and reships counted once with clear rules.
Returns reason taxonomy applied; “buyer’s remorse” separated from defects.
Fraud/chargebacks reconciled to processor statements; disputes resolved lag handled.
Percentiles (p50/p75/p90) reported where tails matter (contact time, latency).
Ratio‑of‑sums used for roll‑ups; sum‑of‑ratios shown only for diagnostics.
Verifier replication complete; clean‑team and privacy rules observed for order/customer‑level data.

Early‑warning indicators (put these on the monthly dashboard)

Outbound freight + accessorials as % of revenue drifting upward by >100 bps; spike in fuel/oversize fees.
Return rate or “no fault found” fraction rising; refurbish capacity backlog.
Payment fee schedule changes; chargeback rate > threshold for card network programs.
Contact rate per 100 orders/accounts increasing; backlog age > SLA.
Cloud unit cost (per API call/account) spiking; on‑demand share rising.
3PL/marketplace policy changes: paid placement creep, storage fees, compliance fines.
Expedite spend > target; OTIF fines or SLA credits escalating.

Red flags—and how to respond now

CTS > 30% of revenue in core DTC segments with no premium price lever → reset free‑shipping thresholds, introduce return fees for remorse, and redesign packaging; reweight mix to replenishment SKUs.
Gold/Platinum support tiers priced like Standard while CTS is 2–3× higher → enforce price fences or de‑scope entitlements.
Carrier/3PL concentration >60% with rising accessorials → run a lane/zone rebid and activate a secondary gateway.
Cloud overage dependence → move to reserved or savings plans; prioritize the two highest‑burn endpoints for caching and query optimization.
Chargeback or fraud spikes → tighten SCA/3DS, address item‑not‑received root causes (scan chain), and recalibrate cancellation windows.

Copy‑ready templates

Cost‑to‑Serve Canvas (one per segment × channel)
- Scope & period; revenue metric (vendor‑realized).
- CTS equation with buckets; drivers selected.
- Corridors: CTS per unit and CTS as % of revenue (median/IQR).
- Service overlay: OTIF/latency/FRT; return/contact rates.
- Levers & scenarios: top three with $/timelines; owners.
- GL tie‑out notes; confidence label; verifier initials.
Driver & Rate Sheet (time‑driven ABC)
- Roles, loaded hourly cost, utilization assumptions, cost/minute.
- Standard minutes per activity (pick, pack, care tier, success task, field visit step).
- Evidence source (time study, WMS timestamps, ACD logs).
Invoice Reconciliation Log
- Carrier/3PL/processor/cloud invoices matched to GL lines; variance reasons; accruals reversed.

72‑hour sprint plan (from blank page to decision‑grade CTS)

Day 0: Freeze segments and CTS scope; issue a single data request (ops, finance, contracts); publish measurement rules (what counts where).
Day 1: Build first CTS equations and invoice‑anchored buckets for two top segments; compute corridors; start GL tie‑out; draft Canvas pages.
Day 2: Add TDABC for people steps; overlay service metrics; run one cost‑shock scenario (fuel/returns/chargebacks) and one service‑policy scenario (free‑shipping threshold, support tiering).
Day 3: Push CTS into CM2 by segment; set early‑warning indicators; publish a 90‑day margin plan (3–5 levers with owners and dates); propose term‑sheet protections (indexation, expedite caps, 3PL remediation milestones, earnouts tied to gross margin after shipping).

Acceptance criteria for a decision‑grade CTS benchmark

Clear CTS definition and segmentation; corridors computed per segment/channel; no averages that hide truth.
Invoice‑anchored buckets reconciled to GL; TDABC rates documented; verifier tie‑out complete.
Service overlay and return/contact rates linked to CTS and to price/retention.
Two or more decision‑relevant scenarios modeled with toggles in the plan; early‑warning indicators set.
Model updated within 24 hours; residual uncertainty translated into valuation ranges and terms; 90‑day action plan issued with owners and timelines.

Work through this checklist and cost‑to‑serve will stop being a vague “shipping and support line.” It will become a segmented, lever‑ready map of where margin is lost or protected—and exactly what to do, at what cost and speed, to move CM2 in the direction your deal requires.

Commercial Due Diligence Playbook >