Novelty is the easy hurdle. Non-obviousness is the one that rejects most patent applications, and it is where nearly every office action lands. An examiner is not asking whether anyone built your invention before — that question is usually settled quickly. The examiner is asking whether a reasonably skilled engineer in your field, holding all the published prior art in front of them, would have arrived at your combination anyway. If the answer is yes, the application fails even though nothing identical exists.
This article is general education from an engineering perspective. Projects House is an engineering firm, not a law firm, and nothing here is legal advice. Patentability and prosecution strategy are decisions to make with a registered patent attorney or patent agent.
Two completely different tests
US patent law applies two separate requirements, and conflating them is the most common source of inventor confusion:
- Novelty (anticipation). Does a single prior reference already disclose every element of your claim? This is a binary comparison against one document at a time. Adding one genuinely new element clears it.
- Non-obviousness. Would combining two or more references have been obvious to a person of ordinary skill in the art at the time of filing? This is a judgment call, made by an examiner who can freely combine documents from different sources.
That second test is why "nobody has ever done exactly this" is not an answer. An examiner can take a mechanism from one patent and a sensor from another, assert that combining them was routine engineering, and reject the claim. Understanding what a claim actually asserts helps enormously here — see patent claims explained.
Who is the "person of ordinary skill in the art"?
This hypothetical person is a legal construct, and defining them shapes the whole analysis. They are competent but not creative: someone with the ordinary training and experience of a working professional in the relevant field, aware of everything published in it, and capable of applying routine skill — but not of making inventive leaps.
The level of skill assumed matters. In a mature, highly technical field, the hypothetical person is presumed to know a great deal, and more combinations look obvious to them. In an emerging or interdisciplinary space, less is presumed, and more combinations survive. This is one reason inventions that bridge two unrelated disciplines often hold up better than refinements within one.
How an examiner constructs the rejection
An obviousness rejection almost always follows the same skeleton:
- Identify a primary reference that discloses most of the claim.
- Identify what is missing from that reference.
- Find a secondary reference disclosing the missing element, often in a different application entirely.
- Articulate a reason to combine — a stated motivation such as improving efficiency, reducing cost, or applying a known technique to a known problem.
- Assert a reasonable expectation of success in making the combination.
The pressure points are steps four and five. A rejection that simply asserts the combination "would have been obvious" without articulating a reason is vulnerable. So is one where combining the references would actually break the primary reference's stated purpose, or where the secondary reference teaches away from the combination. Responding to this is a structured exercise, described in how to respond to a patent office action.
What genuinely supports an inventive step
Some kinds of evidence carry real weight, both during prosecution and later if a patent is challenged:
- Unexpected results. The combination produces an outcome the prior art would not have predicted — a disproportionate performance gain, or a side effect that vanishes. This is the strongest argument available, and it depends on data.
- Solving a problem others tried and failed to solve. Documented failed attempts in the field are powerful.
- Teaching away. A prior reference that explicitly says your approach will not work, when in fact it does.
- Long-felt need. A known problem that persisted for years despite demand.
- Commercial success attributable to the claimed feature specifically, not to marketing.
- Skepticism from experts that the approach could work at all.
Notice how much of that is engineering evidence rather than legal argument. Comparative test data showing your configuration outperforms the obvious alternative is often the difference between an allowance and an abandonment. Generating that data is real work, and it is a good reason not to rush from concept to filing — particularly when refining something that already exists, as discussed in patenting an improvement to an existing product.
What does not help
- How long the development took, or how much it cost.
- That the combination was not obvious to you — the standard is the hypothetical skilled person, not the inventor.
- That the product sells well, without tying success to the claimed feature.
- Simply substituting one known equivalent for another, changing a material, scaling dimensions, or automating a manual process — these are routinely treated as within ordinary skill.
What this means for how the application is written
The obviousness analysis is largely decided by drafting choices made before filing:
- Describe the problem precisely, and why the obvious solutions fail. A specification that explains what goes wrong with the straightforward approach gives the attorney material to argue with later.
- Include quantitative results. Numbers in the specification support an unexpected-results argument. Numbers you did not include cannot be added afterward.
- Claim a range of scope. Broad independent claims plus progressively narrower dependent claims mean an obviousness rejection of the broadest claim does not sink the application.
- Search first. Knowing the closest references before drafting lets you write around them rather than discovering them in an office action. Approaches are covered in patent search keywords.
- Use the filing sequence deliberately. Filing early to secure a date, then developing supporting data before the full application, is a common strategy — see provisional patent applications.
You received an obviousness rejection — what now?
It is expected, not fatal. Most applications that eventually issue received at least one rejection along the way. The realistic paths are: argue the examiner's combination is improper, amend the claims to add a limitation the references do not disclose, submit evidence of unexpected results, or request an examiner interview to discuss what would be allowable. Which path fits depends on the references cited and how much room your specification left you. The broader sequence is laid out in how to patent an idea, and further background on the whole area is collected on our patents and intellectual property page.
Build the technical case behind the claim
The strongest response to an obviousness rejection is usually engineering evidence — a test that shows your configuration does something the obvious one does not. Send us your project details through the contact form and we will help design the comparison your attorney can use.