Case Study 15-01: Elena's Contract Analysis

Claude and the 80-Page Agreement

Persona: Elena Rodriguez, independent strategy consultant. Serves mid-market companies on strategic initiatives across technology, financial services, and business services sectors. Operates as a solo practitioner with occasional subcontractors. Deliverables are primarily Word documents and PowerPoint decks.

The Situation: Elena's client is a $45M revenue technology services company in the final stages of a strategic partnership negotiation with a large enterprise software vendor. The vendor's legal team has produced an 82-page Services Agreement and Partnership Framework document. Elena's client does not have in-house legal counsel; they rely on outside counsel who charges $450/hour.

Elena's task: review the agreement before it goes to outside counsel, identify the most significant issues, and produce a structured brief so that the legal review time is targeted rather than comprehensive — meaning the attorney can focus specifically on the issues Elena has flagged rather than reading the entire agreement from scratch.

Without AI assistance, this work would take Elena two full working days: careful reading with extensive notes, drafting a structure for the brief, writing the analysis. With Claude, Elena aims to complete the same quality of work in one day.

The Setup

Elena's preparation before engaging Claude:

Tool selection rationale: Elena chose Claude over ChatGPT for this task specifically because of the long context window and what colleagues had described as more careful reading of dense text. An 82-page agreement in PDF was approximately 65,000 tokens — well within Claude's context window, and at a length where middle-document content can be lost by less capable models.

Document preparation: Elena converted the PDF to plain text (more reliable for upload than PDF) and verified the character count was within limits. She saved the document in sections (Introduction, Commercial Terms, Liability and IP, Data Provisions, Miscellaneous) to allow for section-by-section analysis if needed, though she planned to upload the full document first.

Prompt preparation: Elena drafted her analysis prompt before the session began rather than improvising it. She knew from experience that complex analytical tasks benefit from a well-structured prompt, and she had learned XML tagging from a colleague.

Phase 1: Document Confirmation

Elena's opening message after uploading the full document:

I've just uploaded an 82-page Services Agreement and Partnership Framework.
Before we do any analysis, please confirm you have the complete document by:
1. Telling me how many major parts or sections the document contains
2. Quoting the first sentence of the final section (whatever comes before signatures)

Claude's response confirmed: the document contained seven parts plus exhibits, and quoted the opening of the final "General Provisions" section accurately. Elena verified this against her own copy. The document was received in full.

This step took two minutes and caught something important: Elena discovered that one of the exhibits was not included in the text conversion. She uploaded it separately and confirmed Claude had received both documents before proceeding.

Phase 2: Initial Structural Analysis

Before jumping to issue identification, Elena asked for a structural overview:

<background>
My client is a $45M technology services company negotiating a strategic partnership
with a large enterprise software vendor. My client has limited in-house legal resources.
I am a strategy consultant, not an attorney. My analysis will feed into a focused
legal review by outside counsel.
</background>

<instructions>
Provide a structural overview of this agreement:
1. What type of agreement is this and what relationship does it create?
2. What are the primary obligations of each party?
3. What is the commercial structure (payment terms, revenue share, fees)?
4. What is the term and termination structure?
5. What are the renewal and exit provisions?

For each area, summarize what the agreement says in 2-4 sentences.
Flag anything that seems incomplete, undefined, or that references another
document that was not included.
</instructions>

<output_format>
Use headers for each area. Plain text, no markdown formatting.
Professional language accessible to a business audience.
</output_format>

Claude's structural overview confirmed the basic commercial framework and flagged two items: a reference to a "Product Schedule" that was not in either document, and a term in the revenue share calculation that referenced "Adjusted Gross Revenue" which was defined in an appendix that Elena did not have.

Elena contacted her client to obtain the missing documents. They arrived within an hour. She uploaded them and continued.

Why this step mattered: Gaps in reference documents are common in contract review and can make analysis of dependent clauses misleading. Identifying them at the structural phase prevented wasted analysis time on provisions whose meaning could not be determined without the missing material.

Phase 3: Systematic Issue Analysis

Elena's core analysis prompt, using the six-area framework she had developed for partnership agreements:

<document>
[full agreement text]
</document>

<supplementary_documents>
[product schedule and appendix text]
</supplementary_documents>

<background>
Client context:
- $45M revenue technology services company
- This is their first partnership with a large enterprise vendor
- They have moderate negotiating leverage (vendor needs regional partners, but
  client is not the only option)
- Client priority: protect their customer relationships — they are worried that
  the vendor relationship could allow the vendor to go direct to the client's
  customers over time
- Second priority: protect their IP — they have a proprietary service delivery
  methodology they do not want the vendor to acquire
- Timeline: client wants to sign within 30 days

I am a strategy consultant providing pre-legal analysis to focus outside counsel's
review time. My client has a $40,000 legal budget for this review.
</background>

<instructions>
Analyze this agreement across six areas. For each area:
a) Summarize what the agreement says (2-4 sentences)
b) Flag any terms that appear asymmetric, unusual, or potentially problematic
   for my client given the context above
c) Indicate negotiating posture: is this typically negotiable language, or
   is this standard boilerplate that is unlikely to change?
d) Priority level: HIGH (must discuss with counsel), MEDIUM (worth discussing),
   LOW (standard, likely fine)

Six analysis areas:
1. Customer Relationship Protection — who can the vendor contact directly?
   What restrictions apply to the vendor approaching the client's customers?
2. Intellectual Property — who owns what? What license does the vendor get
   to the client's methodology and service delivery approaches?
3. Revenue Share and Commercial Terms — how is the revenue split? What
   deductions or adjustments are allowed? Are the definitions clear?
4. Exclusivity and Competitive Restrictions — what can the client not do
   during and after the partnership?
5. Liability and Indemnification — who is liable for what? Is it asymmetric?
6. Termination and Exit — what triggers termination? What happens to the
   business after termination?
</instructions>

<output_format>
Use numbered sections matching the six areas above. Within each section,
use the four sub-points (a, b, c, d) labeled clearly.
At the end, provide an Executive Summary of the top five issues requiring
legal review, in priority order, with a one-sentence explanation of why
each matters.
Maintain professional consulting language throughout.
</output_format>

The Analysis Output

Claude's response ran approximately 3,500 words across the six sections. Selected findings:

Area 1: Customer Relationship Protection (HIGH priority)

What the agreement says: The vendor has an unrestricted right to "develop relationships with potential customers in the territory" as long as they notify the client within 10 business days of initiating contact with any company the client has identified as an active prospect.

Why it matters: The 10-day notification window, not a prohibition, means the vendor can approach the client's customers and prospects without restriction. "Active prospect" is undefined — a vendor who waits for the prospect list could argue that any company not on the most recent prospect list is a legitimate target.

Negotiating posture: The notification-not-prohibition structure is aggressive for a partnership of this type. More partner-friendly agreements include a 12-24 month exclusivity on named accounts and a defined process for prospect list updates.

Priority: HIGH. This provision directly contradicts the client's stated primary concern about customer relationship protection.

Area 2: Intellectual Property (HIGH priority)

What the agreement says: Section 8.3(c) grants the vendor "a perpetual, irrevocable, worldwide license to use, adapt, and develop derivative works based on Jointly Developed Materials." Section 8.1 defines Jointly Developed Materials broadly to include "any methodology, process, or approach developed or refined in connection with this Partnership."

Why it matters: The client's proprietary service delivery methodology is likely to be "refined in connection with this Partnership" — they will use it to service joint clients. Under this language, the vendor may acquire perpetual irrevocable rights to the client's core intellectual property simply by virtue of the partnership occurring.

Negotiating posture: This provision is extremely aggressive and would be flagged in any competent legal review. Typical language limits vendor licenses to materials created specifically for the vendor relationship, with explicit carve-outs for each party's pre-existing IP and independently developed improvements.

Priority: HIGH. The IP exposure here could be existential for the client's business model.

Area 4: Exclusivity and Competitive Restrictions

What the agreement says: The client agrees not to partner with any "Competing Products" in the territory during the partnership term and for 18 months after termination. Competing Products are defined as "any software or service that competes with any product in Vendor's current or future product portfolio."

Why it matters: "Future product portfolio" means the competitive restriction expands as the vendor adds products. If the vendor adds a category the client currently serves independently, the client's independent activity in that category arguably becomes a breach. The 18-month post-termination period compounds this: if the partnership ends acrimoniously, the client's business is significantly constrained for a year and a half.

Negotiating posture: The "future portfolio" language is non-standard. A fixed definition of the competitive field at contract signing is more common. Post-termination restrictions of 6-12 months on current products only is more typical.

Priority: HIGH.

Phase 4: The Executive Summary and Client Brief

After reviewing Claude's analysis, Elena spent 45 minutes: - Verifying three specific citations against the actual document text (all accurate) - Adding two items from her own first-read notes that Claude had assessed as LOW priority but that Elena assessed as MEDIUM given context she had not fully shared - Writing an introductory paragraph explaining the review process and the pre-legal nature of the analysis - Reformatting from Claude's output into her firm's standard Word template

The result: a 12-page analysis brief ready for outside counsel, structured so the attorney could go directly to the five HIGH priority items (covering approximately 25 pages of the agreement) rather than reviewing all 82 pages from scratch.

The attorney reviewed the brief and the agreement and confirmed: all five HIGH priority items were genuinely significant. She identified two additional items (both at a level below what Elena had flagged) and estimated that the pre-analysis saved approximately three hours of her review time — roughly $1,350 at her hourly rate.

What the Workflow Looked Like in Time

Elena's previous two-day estimate for this type of work (without AI assistance) was based on the reading and note-taking time, not counting the drafting of the brief. With Claude, the active work was concentrated into approximately three hours with a faster, more structured output.

What Claude Got Right and What Required Elena's Judgment

Claude's Accurate Findings

All five HIGH priority items Claude identified were confirmed by outside counsel as genuinely significant. The IP provision (Section 8.3(c)) was particularly well-analyzed — Claude correctly identified both the broad definition of "Jointly Developed Materials" and the connection to the client's core methodology, a non-obvious inference that required reading multiple sections together.

The document completeness check (missing Product Schedule and appendix) caught a real gap that would have compromised the analysis if overlooked.

What Required Elena's Professional Judgment

Strategic prioritization: Claude assessed the competitive restriction provision's "future portfolio" language as HIGH priority. Elena agreed with the flag but added context in her brief: given the vendor's stated product roadmap (which she had from separate conversations), the immediate practical risk was lower than the technical legal risk. This nuance required knowledge Elena had outside the document.

Negotiating leverage assessment: Claude's assessments of what was "typically negotiable" were based on general commercial contract patterns. Elena adjusted several of these based on her knowledge of this specific vendor's standard practices and the client's specific leverage position.

Tone calibration: Claude's analysis was thorough but somewhat more alarming in tone than Elena wanted for the client brief. A client who reads "this provision could be existential for your business model" before talking to a lawyer may make poor decisions. Elena restated several findings in language that conveyed urgency without producing panic.

The two additional items: Elena added two items from her first-read notes that Claude had assessed at LOW priority. Both were genuinely low risk, but one connected to a specific regulatory concern in the client's industry that Elena knew about from prior work in that sector — context that was not in the document and that Claude could not have known.

Transferable Principles

Confirm document completeness before analysis. The two-minute confirmation check caught a missing exhibit. That check is always worth doing.

Front-load context about what matters. Elena's background section specified the client's two primary concerns (customer relationship protection and IP protection) and their negotiating leverage. This calibrated Claude's priority assessments toward the client's actual interests rather than generic contract risk.

Use a systematic framework, not a free-form request. The six-area structure produced a more complete analysis than "identify the issues in this contract" would have. Structured requests produce structured outputs; vague requests produce vague outputs.

Verify specific citations against the source. Elena checked three citations. All were accurate — but the habit is what makes the trust appropriate. Verifying a few spots in a long analysis is quick and provides confidence for using the rest.

Your contextual knowledge is irreplaceable. Claude cannot know the vendor's product roadmap, the client's industry-specific regulatory environment, or Elena's judgment about what language a particular type of client can handle emotionally. These contributions came from Elena, not from the analysis tool.

The value of AI in this workflow is in structure and coverage, not in replacing judgment. Claude produced a comprehensive, well-structured first analysis in 40 minutes of model time (plus 25 minutes of Elena's review). Elena spent 45 minutes adding judgment, context, and polish. The total was better than either could produce alone.