Patent practitioners obsess over validity—will the claim survive examination, IPR, litigation? But validity is only half the enforcement equation. A perfectly valid claim is worthless if you can't detect when someone infringes it.
Detectability asks a simple question: Can you prove infringement?
The Detectability Problem
Consider two claims covering the same invention:
Claim A: "An apparatus comprising a sensor, a processor configured to analyze sensor data, and a display showing analysis results."
Claim B: "A method comprising receiving sensor data at a server, processing the sensor data using a machine learning model, and transmitting results to a client device."
Both might be valid. Both might cover the same competitive threat. But detecting infringement is radically different:
- Claim A: Buy competitor's product. Open it. See sensor, processor, display. Infringement proven.
- Claim B: The method runs on competitor's internal server. You can't see the machine learning model. You can't see the processing steps. You can only see what goes in and what comes out.
Factors Affecting Detectability
1. Claim Type
| Claim Type | Detectability | Why |
|---|---|---|
| Apparatus | High | Physical structure can be examined, reverse-engineered |
| System | Medium-High | Often includes observable components |
| Method | Low-Medium | Steps may occur internally, invisibly |
| Computer-Readable Medium | Low | Software analysis required; code often obfuscated |
2. Divided Infringement
Under Akamai Technologies v. Limelight Networks, direct infringement of a method claim requires a single entity to perform all steps. Claims requiring multiple actors create "divided infringement" problems:
⚠️ Multi-Actor Language
"A method wherein a user uploads content, a server processes the content, and an administrator approves the content" requires three actors. Unless one controls or directs the others, there may be no direct infringer.
Watch for language indicating multiple actors:
- "a user," "a server," "an administrator"
- "receiving from a first device," "transmitting to a second device"
- "a client," "a service provider," "an end user"
3. Internal vs. External Steps
Steps performed inside a black box are harder to detect than steps with observable inputs or outputs:
- External (detectable): Displaying a user interface, transmitting a particular data format, physical movement or actuation
- Internal (hard to detect): Algorithm execution, database queries, model inference, caching logic
4. Technical Specificity
Concrete structural limitations are easier to detect than abstract functional descriptions:
- High detectability: "a capacitive touchscreen," "a JSON-formatted message," "a lithium-ion battery"
- Low detectability: "means for receiving input," "processing circuitry," "a data structure"
Our Scoring Methodology
The detectability score starts at baseline and adjusts based on:
Positive Factors
- Apparatus claim type: +2.0
- High technical specificity (>4 technical terms): +1.5
- Low element count (≤4 elements): +1.0
Negative Factors
- Method claim type: -1.0
- CRM/Beauregard claim type: -1.5
- Divided infringement indicators (>2): -2.0
- High functional language density: -1.5
- High element count (>10): -1.0
Improving Detectability During Drafting
💡 Drafting Strategy
For every method claim, consider: "How will I prove a competitor does this?" If you can't answer, redraft or add an apparatus claim covering detectable structure.
Practical strategies:
- Parallel claim sets: For each method claim, draft a corresponding apparatus claim covering the same invention from a structural perspective
- Focus on outputs: Claim the observable result, not just the internal process
- Single-actor claims: Reframe multi-actor scenarios from one actor's perspective
- Product-by-process: When the process is undetectable, claim the resulting product structure
Why This Matters for Enforcement
In litigation, proving infringement requires evidence. For method claims, that often means:
- Expert reverse engineering of competitor's product
- Discovery of competitor's source code or internal documentation
- Inference from observable inputs and outputs
Each approach is expensive, uncertain, or both. Apparatus claims often need only: "Here's their product. Here's our claim. See the match."
A portfolio heavy on low-detectability claims may look strong on paper but prove unenforceable in practice. Our detectability score helps identify this risk before it becomes a problem.
Assess Your Claims
Find out which claims in your portfolio can actually be enforced.
Run Detectability Analysis →Strategic claim drafting requires balancing multiple factors. IP Services drafts claims designed for real-world enforcement, not just examination.