ethos
← Back to Projects
📜

Terms

Contract review that catches risks

Thesis

Reading contracts is tedious and risks get missed. Document comparison can surface unusual terms against your templates.

The Problem

Nobody actually reads contracts: You skim, you sign, you hope for the best. That one weird clause about IP assignment? Missed it. The unlimited revisions language? Didn't notice. The payment terms that net 90 instead of 30? Surprise. Lawyers are expensive. Reading carefully takes forever. So you don't, and sometimes you pay for it.

Implementation Approaches

Approach 1

Template Comparison (Recommended)

Recommended

Compare incoming contracts to your known-good templates

⏱️1-2 weeks
📊Medium complexity

Implementation

  • Upload your standard contract templates
  • Upload incoming contract for review
  • AI highlights: what's different, what's missing, what's risky
  • Risk scoring: green/yellow/red by clause
  • Suggested negotiation points

Pros

  • +Clear value: compare to what you know is safe
  • +Builds library of 'your terms' over time
  • +Fast, review in minutes not hours
  • +Can flag specific red flags (unlimited revisions, IP traps)

Cons

  • Need template library to compare against
  • Can't catch risks not in your templates
  • Legal nuance might be missed
Approach 2

Risk Pattern Detection

AI trained on common contract risks across industries

⏱️2 weeks
📊High complexity

Implementation

  • Pre-trained on common contract pitfalls
  • Scans for: payment terms, IP clauses, liability caps, termination
  • Industry-specific risk patterns
  • No template needed, works standalone

Pros

  • +Works without your templates
  • +Broader risk coverage
  • +Can learn from community patterns
  • +Higher value, more comprehensive

Cons

  • More complex to build and train
  • Generic patterns might miss your specific concerns
  • Higher bar for accuracy
Approach 3

Clause Library

Searchable library of good vs bad clause examples

⏱️1 week
📊Low complexity

Implementation

  • Curated examples of risky vs safe clauses
  • Search by clause type (IP, payment, termination)
  • Side-by-side: what you should push for vs accept
  • Community contributed over time

Pros

  • +Simpler to build, curated content
  • +Educational, helps users learn
  • +Can be free/freemium entry point
  • +Builds audience for full product

Cons

  • Passive, user has to search
  • Doesn't analyze their specific contract
  • Less immediate value

Validation Plan

Hypothesis to Test

Small agencies will pay $29/mo to review contracts 10x faster and catch risks

Validation Phases

1

Manual Comparison

1 week
  • Get 5 real contracts from agency contacts
  • Get their standard templates
  • Manually do comparison using Claude
  • Show output: 'Would this have helped?'
Identify at least 2 risks per contract that user missed
2

Template MVP

2 weeks
  • Build upload + comparison flow
  • Support 3 contract types (MSA, SOW, NDA)
  • Test with 5 agencies on real contracts
  • Measure: time saved, risks caught
Review time under 10 minutes, risk detection accuracy 80%+
3

Paid Pilot

2 weeks
  • Add risk pattern detection
  • Launch to waitlist with $29/mo pricing
  • Track: contracts reviewed, conversion rate
  • Gather feedback on what risks matter most
10+ paying users, weekly active usage

Kill Criteria

Stop and move on if any of these become true:

  • Risk detection has too many false positives/negatives
  • Users don't have standard templates to compare against
  • Legal concerns about AI contract advice
  • Price sensitivity below $15/mo