Saas Agent Sla Cogs Treatment

SaaS Agent SLA-COGS Treatment Skill

Overview

SLA-related cost lines hit the income statement in four places — COGS, contra-revenue, S&M, or G&A. The classification is not a stylistic choice; it is an accounting policy under ASC 606 / IFRS 15 (for the contra-revenue items) and under standard cost-classification guidance (for the opex items). Getting it wrong:

• Overstates revenue (if SLA credits are misclassified as opex)
• Understates COGS (if SLA-defence HITL labour is misclassified as G&A)
• Distorts gross margin (the metric investors lean on hardest)
• Triggers audit adjustments and DD findings

This skill installs the classification discipline and the disclosure policy.

Use When

• An agent product has contractual SLA commitments and credits, refunds, or related costs
• The income statement must reconcile to gross margin and contribution margin investors will lean on
• An audit firm or DD team has asked for the cost-classification policy
• Cross-loaded with saas-agent-revenue-recognition, saas-agent-deferred-revenue-and-credit-reserves, and saas-agent-unit-economics-and-cogs

Do Not Use When

• The agent product has no SLA commitments and no refunds
• The plan is pre-revenue and the cost shape is not yet committed (use directional treatment)

Required Inputs

• Pricing primitive(s) (from saas-agent-pricing-strategy)
• SLA terms (from contract templates)
• Cost-line inventory (HITL labour, retraining cost, observability cost, eval cost, CS cost, legal cost, refund-processing cost)
• Functional org chart (who does SLA defence, who manages disputes, who delivers customer success)
• Tooling cost breakdown (LLM, observability, audit-log retention, SLA-monitoring)

Workflow

Apply the ordered stages below; stop and recover when a stage lacks its required evidence.

1. Apply the classification policy

Per references/saas-agent-sla-cogs-policy.md, classify each SLA-related cost line:

COGS (Cost of Revenue):

• HITL labour deployed specifically to defend SLA (humans intervene to restore service quality)
• HITL labour deployed on eval-flagged tickets to maintain accuracy SLA
• Retraining cost amortisation tied to SLA-relevant quality metrics
• SLA-monitoring infrastructure (telemetry, alerting, dashboards)
• Eval cost specifically on SLA-relevant metrics (accuracy, latency, refusal rate)
• Audit-log retention infrastructure for SLA-evidence purposes
• LLM cost retried because SLA quality threshold not met on first attempt
• Refund-processing cost (mobile-money fees, payment-gateway fees on refund issuance)

Contra-revenue (reduction of revenue, NOT opex):

• SLA credits issued (because of breach)
• Outcome refunds (because of failed verification)
• Volume rebates (variable consideration)

Sales & Marketing (S&M):

• Customer-success management of SLA conversations
• Account-management time on SLA-tier upsell / downsell
• SLA-as-differentiator marketing content

General & Administrative (G&A):

• Legal cost defending SLA disputes
• Compliance cost responding to regulator-mandated SLA standards
• Senior management time on SLA escalations
• Insurance premium for SLA-related liability coverage

2. Build the policy memo

Per references/saas-agent-sla-cogs-policy.md — produce a policy memo covering:

• Cost-line inventory with classification per line
• Allocation rules for shared cost lines (e.g. an eval engineer who works on both SLA-relevant evals and product-improvement evals — split by time)
• Disclosure language for the audited financial statements

3. Wire to the COGS waterfall

Cross-load with saas-agent-unit-economics-and-cogs. Each COGS line in the waterfall must map to a classification decision documented here.

4. Wire to the contra-revenue presentation

Cross-load with saas-agent-revenue-recognition:

• Gross agent revenue
• Less: SLA credits
• Less: refunds
• Less: volume rebates
• Net agent revenue (this is the line that flows to gross margin)

5. Reconcile to gross margin

Gross margin = (Net agent revenue - COGS) / Net agent revenue

The SLA-COGS items are part of the COGS subtraction. The SLA-credit and refund items have already reduced revenue (above the GM line) — they do NOT also reduce COGS.

This is the policy that prevents double-counting.

6. Worked example

7. Disclosure language

Include in the audited financial statements a note covering:

• SLA-credit policy (contra-revenue, estimated under variable-consideration constraint)
• Refund policy (contra-revenue, estimated under variable-consideration constraint)
• HITL labour classification (COGS where deployed on customer service delivery)
• Retraining amortisation policy (COGS if quality-related; R&D if capability-related)
• Eval cost allocation policy
• Insurance classification

8. Wire to bankability and DD

Bankability scorecards (especially DFI / institutional) check:

• Is gross margin reported on net revenue (after SLA credits + refunds)?
• Is HITL labour in COGS?
• Is SLA-credit visible separately?

The classification policy is the artefact that satisfies the check.

9. Wire to living-plan governance

Per cadence table below.

Quality Bar

• Every SLA-related cost line is classified
• Shared cost lines have a documented allocation rule
• SLA credits are contra-revenue, not opex (the #1 misclassification)
• Refunds are contra-revenue, not COGS
• HITL labour for SLA defence is COGS, not S&M
• Retraining classification documented (COGS vs R&D)
• Policy memo defensible in an audit
• Disclosure language drafted
• Gross margin computed on net revenue
• Cross-referenced to rev-rec, reserves, unit economics

Anti-Patterns

• SLA credits as marketing expense — wrong
• Refunds as COGS — wrong
• HITL labour as S&M — wrong if for service delivery
• All retraining as R&D — wrong if quality-related
• Mixing SLA-monitoring and product observability without allocation
• "We classify intuitively" — audit will require documentation
• Reporting gross margin on gross revenue — overstates
• No disclosure language drafted — DD finding

Outputs

• Cost-line inventory with classification
• Allocation rules for shared cost lines
• SLA-COGS policy memo
• Disclosure language draft
• Worked income-statement example
• Cross-reference to rev-rec, reserves, unit economics
• Living-plan cadence assignment

Living-Plan Cadence Defaults

References

• references/saas-agent-sla-cogs-policy.md — full classification policy and disclosure language
• skills/10-financial-projections/saas-agent-revenue-recognition/SKILL.md — contra-revenue side
• skills/10-financial-projections/saas-agent-deferred-revenue-and-credit-reserves/SKILL.md — reserve side
• skills/10-financial-projections/saas-agent-unit-economics-and-cogs/SKILL.md — COGS waterfall consumer
• skills/meta-agent-sla-financial-controls/SKILL.md — controls
• skills/meta-accounting-finance-review/SKILL.md — accounting review gate
• book-extractions/agent-sla-commercial-business-plan-audit-2026.md — audit
• book-extractions/accounting-bookkeeping-finance-controls-extraction.md — controls reference

Africa / Uganda Application Notes

• HITL labour in Africa is materially cheaper than US benchmarks (UGX 4,000-8,000/hour fully loaded vs USD 30-60/hour in the US); the SLA-defence-HITL line is smaller in absolute USD terms but the classification rule is the same
• Mobile-money refund-processing fees (1-2.5%) — clearly COGS (cost of delivering the refund)
• VAT classification of SLA credits — Uganda VAT (18%), Kenya (16%), Nigeria (7.5%), South Africa (15%), Rwanda (18%) — SLA credits as transaction-price reductions reduce VAT-output proportionally; document the policy
• Public-sector contracts in Africa — SLA-defence cost may include in-country support staff cost (data-residency requirement); classify as COGS
• Sovereign-AI compute cost premium — in-region GPU is 1.5-3x US/EU; the cost is COGS regardless of where SLA defence happens
• Insurance premium classification — AI E&O coverage where available; G&A; document allocation between SLA-related and general
• Local audit firm coaching — provide the worked example; clarify the contra-revenue concept which mid-tier firms may not have seen before

July 2026 Portable Contract

Required Inputs

Outputs

Evidence Produced

Capability and Permission Boundaries

For saas agent sla cogs treatment, the controlling focus is SLA-credit classification, service delivery cost, contra-revenue risk, and policy consistency. This skill may inspect records and calculate planning scenarios in read-only mode; it may not post entries, change ledgers, set accounting policy, certify IFRS treatment, or release statutory values without authorised professional review. Its normal mode is read-only analysis and drafting. Any mutation, external communication, spending, certification, or professional conclusion outside that boundary requires explicit authority and must remain traceable to the approving role.

Degraded Mode

For saas agent sla cogs treatment, loss of evidence about SLA-credit classification, service delivery cost, contra-revenue risk, and policy consistency activates degraded mode. If the controlling saas agent sla cogs treatment evidence is unavailable, the same boundary applies. When contract terms, usage evidence, framework, or cost drivers are unavailable, isolate the affected schedule, label it unassessed, and do not force the model to balance with a plug. Return the verified subset, label the affected decision qualified or not assessed, explain the downstream consequence, and state the smallest evidence request or authorised action that permits recovery. Do not convert the missing check into a pass.

Decision Rules

Workflow

1. Define the exact saas agent sla cogs treatment decision, intended reader, jurisdiction, business stage, and permission boundary.
2. Collect approved commercial assumptions, contracts, usage/cost evidence, accounting framework, opening position, and projection horizon and map each material conclusion to its source; stop the affected conclusion when an input could change it.
3. Apply the specialist methods and directly linked references already contained in this skill, retaining its domain thresholds, calculations, and Uganda or East Africa context where applicable.
4. Compare the credible alternatives, test the countercase and failure path, and apply the decision table rather than selecting a template default.
5. Produce sla-credit accounting classification memo and projection-line mapping with the evidence, exception, and handoff records; reconcile every shared assumption with its owning section.
6. Run the section quality checks, applicable finance or professional review, and anti-slop gate. If a gate fails, correct the evidence or decision and return to the responsible step.

Quality Standards

• SLA-credit accounting classification memo and projection-line mapping must answer a real decision for the named bank, investor, DFI, grant, board, or strategic-partner reader.
• Formula trace, source/assumption register, three-statement or schedule reconciliation, and finance-gate record must be source-linked, dated where facts can change, and sufficient for another reviewer to reproduce the conclusion.
• The section exposes its countercase, stop condition, recovery action, and effect on neighbouring sections.
• No unavailable source, calculation, tool, or professional review is reported as passed; finance and statutory judgements follow the governing doctrine.
• Language remains specific to saas agent sla cogs treatment, uses British English naturally, and passes the repository anti-slop gate without promotional filler.

Anti-Patterns

• In saas agent sla cogs treatment, treating an unavailable approved commercial assumptions, contracts, usage/cost evidence, accounting framework, opening position, and projection horizon as confirmed. Correction: qualify the affected conclusion and issue the named evidence request.
• Producing sla-credit accounting classification memo and projection-line mapping that restates the brief but makes no choice. Correction: record the choice, rejected alternative, rationale, countercase, and implication.
• Ignoring a conflicting upstream assumption. Correction: return it to its owning section and resume only from a reconciled version.
• Reporting an unavailable check as passed. Correction: mark it not assessed and narrow the release state.
• Claiming compliance, assurance, bankability, or investor readiness from narrative quality. Correction: run the applicable gate and retain its evidence.
• Copying the worked example into a client plan. Correction: use the method only and replace every fact with verified engagement evidence.

Worked Example

Monthly service credits compensate customers for failed delivery rather than purchasing a distinct service. Document whether they reduce revenue or represent fulfilment cost, and obtain finance-policy approval before modelling COGS.

References

• Use the verified project evidence register and the owning upstream pipeline section for saas agent sla cogs treatment; no local deep-dive reference is declared.
• For saas agent sla cogs treatment claims involving money, tax, grants, reserves, revenue, cost, valuation, or financial statements, apply the Chwezi finance doctrine and record the required professional-review state; illustrative figures never become client facts.