Master Services Agreements

How to prepare a clause-level risk assessment for Master Services Agreements

Prepare a customer-side MSA clause risk assessment with a copyable risk register, source checks, liability analysis, and a fictional worked example.

October 18, 2026  ·  10 min read
Book a Demo →

A customer-side MSA risk assessment connects each material clause to a service failure, business consequence, available remedy, and proposed response. Start with the contract package, establish which terms control each statement of work, and assess related provisions together rather than scoring clauses in isolation. Record evidence, uncertainty, severity, and decision ownership in a clause risk register. The deliverable should tell negotiating counsel what to change, the business what exposure remains, and the approver which unresolved risks require a decision.

When to use this workflow

Use this workflow when customer counsel receives a supplier’s MSA, reviews a renewal with changed services, or assesses an existing agreement before adding a material SOW. It suits commercial legal teams and outside counsel supporting implementation projects, outsourcing, and recurring managed services.

The task is clause-level assessment, not a complete negotiation or contract-management program. Review the MSA and documents affecting the proposed services, then deliver an evidence-linked register and decision summary.

This guide assumes US commercial contracting. Interpretation and enforceability depend on governing law, actual wording, the agreement hierarchy, and relevant facts. The assessment method below is a practical recommendation, not a legal requirement or advice for a particular transaction. Separate legal uncertainty from commercial exposure and operational feasibility.

What you need

Obtain the MSA, relevant SOWs, amendments, pricing and service-level schedules, security and data-processing addenda, and incorporated online terms. Distinguish signed documents from proposals. Preserve identifiable versions of online materials rather than relying only on live links.

Gather implementation milestones, acceptance tests, customer dependencies, data categories, supplier access, expected charges, replacement options, and transition requirements. Ask the service owner what failed implementation or prolonged disruption would interrupt.

Use a versioned customer negotiation checklist containing preferred positions, fallbacks, escalation triggers, and approval owners. This is the team’s work product, not an assumed software feature. Identify contacts for security, IP, finance, insurance, and service delivery. Missing facts should generate assigned questions rather than assumptions.

Use the template

The companion template is Master Services Agreements — Clause risk register. Copy this table into your working document or spreadsheet. Use one row per independently actionable risk; combine related provisions only where they create one exposure and decision.

ID / topic Source / version Evidence / status Scenario / impact Dependencies / remedy Severity / confidence Proposed action Owner / decision

In Source / version, identify the document ID, version, clause, and page. Cross-reference related register IDs. In Evidence / status, retain operative wording or a precise paraphrase, clearly distinguished from a quotation. Keep absent, ambiguous, redacted, and not reviewed separate. “Absent” means not found in the identified materials searched—not absent from unseen documents.

Scenario / impact states the customer consequence. Dependencies / remedy captures definitions, exceptions, notice conditions, exclusive remedies, and conflicting provisions. Separate severity from confidence: potentially critical exposure can remain uncertain because an incorporated schedule is missing.

Specify mitigation in Proposed action. Use Owner / decision for the accountable person, status, and dated resolution or acceptance. Sending a comment does not close a finding.

Step by step

1. Establish the package and document hierarchy

Assign document IDs and versions. Identify incorporation language, missing attachments, amendment coverage, and execution requirements. Determine whether an SOW can override the MSA and whether an exception affects only that undertaking.

The Access Worldwide–E*TRADE MSA, §§1.3–1.4 gives master terms priority unless the Scope of Work explicitly provides otherwise and conditions SOW effectiveness on authorized signatures. Those are agreement-specific provisions, not universal rules. (sec.gov)

By contrast, ProQuest–IBM, Exhibit B §1.2 gives conflicting Service Agreement terms priority for the particular undertaking without changing other Service Agreements. (sec.gov)

Record: controlling-document conclusions, affected SOW IDs, and missing-document or conflicting-precedence risks. Leave the conclusion open if the hierarchy cannot yet be resolved.

2. Translate the services into failure scenarios

Interview the service owner before assigning severity. For implementation, examine delayed milestones, failed acceptance, unusable deliverables, and dependency disputes. For managed services, examine repeated outages, security incidents, subcontractor failure, and inability to transition.

Map scenarios to acceptance, fees, service levels, IP, confidentiality, security, subcontracting, indemnity, audit, termination, and transition provisions. Check customer cooperation duties: access delays or missed approvals may qualify supplier obligations or trigger extra charges under the wording reviewed.

Record: each material scenario and customer impact. “Supplier-friendly” is not an impact; “customer must pay before testing functionality” is. Identify essential protections checked even when no risk row is needed, so silence is not mistaken for completed review.

3. Extract operative language and qualifications

Read definitions, exceptions, cross-references, and schedules with the main clause. For acceptance, identify tests, the review period, rejection method, deemed-acceptance trigger, cure process, and payment dependencies. For IP, distinguish pre-existing materials, project deliverables, embedded components, and rights needed to use or maintain the output.

Preserve enough evidence to support the finding without reproducing unnecessary text. Where language is missing, identify the documents searched. Where wording is ambiguous, explain competing readings rather than selecting the favorable interpretation.

Record: clause/page citations, operative evidence, review status, and confidence. Link qualifications to affected findings; do not bury a remedy-limiting exception in an unrelated row.

4. Test remedies and liability together

Follow each failure through the remedies. Determine whether credits are exclusive, termination requires repeated failures, indemnity covers the relevant claim, or transition assistance requires additional payment. Test notice and cure conditions against the customer’s operational capabilities.

Model liability using fee scope, lookback period, triggering event, aggregation, exclusions, exceptions, separate caps, and erosion rules. Test an early-term incident and concurrent claims with stated assumptions. A cap is a ceiling under the relevant wording, not a prediction of recoverable damages.

For illustration, Invacare–Birlasoft §19.2(b) uses the greater of $30 million or specified charges over an 18-month lookback, subject to exceptions. This is negotiated wording, not a benchmark. (sec.gov)

Section 19.2(c) separately caps a defined security-breach category, prevents payments under either cap from eroding the other, and disapplies §19.2(a)’s damages exclusion for that category, subject to exceptions. (sec.gov)

Record: remedy gaps, calculation assumptions, and interacting clauses. Assign uncertain governing-law treatment to counsel rather than presenting it as arithmetic.

5. Assign severity using customer consequences

Use a consistent working scale. Critical issues block an essential requirement or require senior approval; high issues materially impair delivery or recovery; medium issues need negotiation or operational mitigation; low issues have limited consequences for the identified services.

Explain the rating through service criticality, credible impact, alternatives, and available remedies. These are proposed assessment categories, not legal classifications. Do not reduce severity because wording is familiar or the supplier calls it standard.

Record: severity, confidence, and the escalation reason. Assign missing-fact questions to a business or specialist owner. Where uncertainty prevents a reliable rating, use a provisional rating and state what evidence could change it.

6. Specify the preferred change and fallback

Propose targeted changes: revise deemed acceptance, narrow supplier relief, preserve remedies beyond credits, clarify deliverable rights, or define transition assistance. Check consequential edits to definitions, fees, and cross-references.

Offer a fallback where useful, explaining its different risk profile. An extended testing period may address the operational problem better than unrestricted rejection rights. An operational safeguard can reduce exposure but should not be treated as equivalent to a contractual remedy without analysis.

Record: preferred change, fallback, approval requirement, and residual exposure. Keep an unapproved fallback open. Distinguish recommendations to negotiate from questions requiring evidence or specialist advice.

7. Verify and prepare the decision handoff

Recheck material findings against their identified source versions. Confirm that amendments or SOW provisions have not resolved them. Have the service owner validate operational assumptions and specialists address unresolved security, IP, or governing-law questions.

Prepare a short summary separating blocking issues, proposed negotiating positions, and risks recommended for acceptance. Link every conclusion to register IDs. Preserve the earlier rationale when revised wording changes the recommendation.

Record: verification status, outstanding action, accountable owner, and dated decision. Close a row only when the resolution or authorized acceptance is documented, including any residual exposure.

Worked example

Fictional scenario: Juniper Quay Foods proposes buying implementation and managed support from Cobalt Finch Systems. All parties, documents, amounts, and clauses in this example are fictional. First-year fees total $240,000, including $60,000 payable before implementation testing. Disruption would delay warehouse dispatch planning; operations needs eight business days for testing.

The package contains draft MSA M-01 v3 and SOW S-01 v2. MSA §2 permits explicit SOW overrides. The SOW expressly replaces the MSA’s ten-day acceptance period. It also incorporates unavailable Security Schedule D. For the liability scenario only, assume an incident occurs after payment of $60,000 but before any other fees are paid; no prior claim has reduced the cap.

ID / topic Source / version Evidence / status Scenario / impact Dependencies / remedy Severity / confidence Proposed action Owner / decision
R1 Acceptance S-01 v2 §4, p.3; M-01 v3 §§2–3, pp.2–4 Express override: acceptance follows three business days unless rejected in writing Eight-day testing cannot finish before deemed acceptance Fee already prepaid; post-acceptance correction rights unclear High / high on timing; medium on remedies Extend period to accommodate testing; clarify rejection and correction rights Counsel and operations / open
R2 Liability M-01 v3 §12, p.11 Aggregate cap: fees paid under affected SOW before event; lost-profit exclusion Early failure may leave insufficient recovery for replacement Assumed formula yields $60,000; exceptions and recoverable costs need analysis High / medium Seek fixed floor and defined fee scope; verify exceptions and cost treatment Counsel and finance / open
R3 Security S-01 v2 §8, p.6 Schedule D incorporated but unavailable Controls and incident-response commitments cannot be assessed Relationship to confidentiality and liability remains unreviewed Potentially high / low Obtain schedule and assess before recommending acceptance Security lead and counsel / blocked

The register changes negotiation priorities. R1 concerns premature acceptance and uncertain correction rights, not a payment triggered by acceptance: the implementation fee is already prepaid. R2 exposes the difference between expected annual spend and fees paid before an early incident, without treating $60,000 as assured recovery. R3 remains an evidence gap, not proof that security protections do not exist. Request the schedule while negotiating the evidenced issues.

Common mistakes

  • Scoring only the MSA. Check SOW overrides and service-specific definitions before deciding which protection controls.
  • Treating a cap multiple as the liability position. Test timing, fee scope, exclusions, aggregation, exceptions, and separate caps together.
  • Listing deviations without consequences. Identify the failure scenario and explain the customer’s remedy gap.
  • Calling missing evidence low risk. Separate uncertainty from severity and assign the document request.
  • Recommending unusable remedies. Check testing capacity, notice mechanics, cure dependencies, and evidence the customer can actually produce.

Final quality check

  • Every material finding identifies a document version and clause/page location.
  • Controlling terms are mapped for each relevant SOW.
  • Operative evidence and cross-references match the source.
  • Missing, ambiguous, redacted, and unreviewed material remain distinct.
  • Liability scenarios identify assumptions, fee scope, and unresolved exceptions.
  • Recommendations address consequential definitions and related remedies.
  • Open issues have accountable owners; accepted risks have dated decisions and residual exposure.

Complete the workflow with DecoverAI

DecoverAI is the Legal Intelligence Platform for litigation and transactions. In its contract review workflow in Word, choose the customer as the represented party and provide service-specific instructions. Findings quote clauses, rank severity, locate text, and identify missing language as not present. Lawyers can edit or insert suggestions as ordinary tracked changes in the signed-in Word user’s name, or dismiss findings with reasons. Counsel retains responsibility for verifying hierarchy, consequences, and decisions. Book a demo to discuss reviewing your MSA package. (decover.ai)

Frequently asked questions

Should every clause receive a risk row?

No. Track review coverage separately and reserve substantive rows for actionable exposure or uncertainty. Otherwise routine provisions can obscure acceptance, liability, security, and exit decisions.

Can one row cover several clauses?

Yes, where they create one exposure, such as deemed acceptance combined with limited correction rights. Cite each operative provision. Split rows when remedies, owners, or decisions differ.

How should an unresolved enforceability question be scored?

Describe commercial exposure separately and assign the governing-law question to counsel. Use a provisional rating where needed; do not assume the provision either fails or operates exactly as drafted.

When should the assessment be updated?

Update affected rows when wording, services, pricing, or incorporated documents change. Recheck linked provisions and confidence levels. Preserve prior reasoning so approvers can trace changes in the recommendation.

Sources

These SEC-filed agreements illustrate specific wording, not general law or market standards.

This guide is general information about legal workflows, not legal advice for any matter. It was drafted and reviewed with AI assistance. The worked example is fictional. Contract rights and procedure depend on the governing law, the actual wording, and the facts, so check the cited sources and the rules that apply in your jurisdiction before relying on it.

Run this workflow in DecoverAI

Clause-quoted findings, missing-language flags, and tracked changes in your name — with your team deciding every call. Plans from $299 per seat per month, SOC 2 Type II certified.

Book a demo
to learn more.
Book a Demo