Drafting an SNQ Request That Survives: Lessons From 4,500 Reexams

95% of utility-patent ex parte reexamination requests clear the SNQ gate (see the SNQ standard explained). The 5% that do not share a small set of weaknesses. A request that survives the determination, and the harder downstream question of being credited as the load-bearing argument in the order itself, is drafted with both outcomes in mind from the start. This is what differentiates the strong requests from the routine ones.

The 5% trap

Across the 4,500+ closed utility-patent reexams, SNQ denials concentrate on two grounds. Cumulativeness denials cite that the references were considered during prosecution, or are materially the same as references that were. Technical insufficiency denials cite that the references do not actually disclose what the request claims.

Both can be defended against in the request. Both require specific drafting choices that many practitioners skip in favor of generic invalidity arguments.

Map every reference against the prosecution record

The first move is mechanical and rarely done well. For every reference cited in the request, identify whether it appears anywhere in the prosecution record: original examination, any prior reexams, parent applications, divisionals, IDS submissions. If a reference is in the record, the request must explicitly distinguish the SNQ being raised from the question previously considered.

The 2002 amendment to 35 U.S.C. 303(a) clarified that prior consideration of a reference does not preclude a new question. But the question must be new, not the reference. If the request fails to articulate what the new question is, the examiner has an easy path to a cumulativeness denial.

Frame the question, not just the reference

A strong SNQ request leads with the question, not the prior art. The framing should be something like: "The substantial new question raised by Reference X is whether the specific feature disclosed in section 4.2 of X, in combination with the prior art recognition that distributed processing is non-novel, would have rendered claim 1 obvious to a person of ordinary skill at the priority date."

That framing is harder to deny than "Reference X discloses elements A, B, and C." A reference-led framing reads as a 102/103 rejection. A question-led framing reads as an SNQ argument.

Element-by-element teaching analysis

The technical insufficiency denial ground is where many requests get sloppy. The request must walk through what each cited reference actually discloses, element by element, against the challenged claim. Vague characterizations ("Reference X generally discloses adaptive control systems") invite denials in a way that specific quoted disclosures do not.

The examiner is looking at the request alongside the patent. Any gap between what the request claims the reference says and what the reference actually says will be visible.

Use the examiner's known patterns

The reexamination examiner pool is small, roughly 110 active examiners. Their historical patterns are visible across closed cases: rejection-basis preferences, denial rate, and the depth of analysis they typically apply at the SNQ stage. A request drafted with the assigned examiner's patterns in mind has a structurally higher probability of producing an order in the form the requester needs.

Two illustrative patterns:

• An examiner whose rejection mix runs 91% 103 across 50+ closed cases is statistically unlikely to issue an SNQ order on a pure 102 ground.
• An examiner whose denial rate runs three times the average requires a request with substantially more developed technical argument to clear the gate.

The data exists. Most requesters do not use it.

What strong requests share

The pattern across requests that clear SNQ and produce orders on the requester's preferred grounds:

• A 1-2 page summary of the substantial new question, distinct from the prior art elements.
• A reference-by-reference statement of whether and how each cited piece appears in the prosecution record.
• Element-mapped disclosures with verbatim quotes, not paraphrase.
• An awareness of the assigned examiner's historical denial pattern and rejection-basis preferences.

A request that hits all four lands inside the 95% grant rate and improves its odds at every downstream stage.