The claims are the numbered list at the end of a patent application, and they are the only part of the document that defines what competitors are forbidden to do. Everything else — the background, the description, the examples, the drawings — exists to support and interpret the claims. A product infringes a claim only if it contains every single element that claim recites, which produces the most important rule in the field: the more elements a claim lists, the narrower it is and the easier it is to design around.

Most inventors reading a draft application study the description and the figures carefully and skim the claims. That is exactly backwards. This article is educational and general — Projects House is an engineering firm, not a law firm, and claim drafting should be done by a registered patent attorney or agent.

What Claims Actually Do

The most useful analogy is a fence around a parcel of land. The description tells you what is built on the parcel and how it all works; only the fence establishes where your property ends and everyone else's begins. Each claim is a single sentence enumerating a set of technical features, and the boundary it draws is exactly as wide as that enumeration allows.

This has a counterintuitive consequence. Adding detail to a claim feels like strengthening it, but it does the opposite — every additional element is one more thing a competitor can omit in order to fall outside the fence. Understanding this also makes reading other people's patents far more productive, a skill covered in how to read a patent.

Independent vs Dependent Claims

  • An independent claim stands alone and defines the invention in the broadest terms the applicant dares to request. This is the claim that sets your real scope of protection.
  • A dependent claim refers back to an earlier claim and adds a feature — a specific material, a dimensional range, an extra component, a particular mode of operation. It is always narrower than the claim it references.
  • Method and apparatus claims are usually both filed, so that you cover both the thing and the process of making or using it. This matters when a competitor manufactures in one place and sells in another.

Dependent claims are not decoration. If the examiner rejects the independent claim, they function as a safety net: you can pull a feature up from a dependent claim and keep prosecution moving rather than starting over. That is precisely where drafting quality shows itself, as anyone who has worked through an office action response discovers.

Breadth Versus Likelihood of Allowance

All of the tension in claim drafting lives here. A very broad claim covers a lot of ground but collides easily with prior art, and an examiner will probably reject it. A very narrow claim gets allowed comfortably but protects one specific configuration, so any small change by a competitor puts them outside your fence.

The standard answer is a hierarchy: one independent claim as broad as the prior art landscape allows, with layers of dependent claims descending from abstract to concrete beneath it. That structure gives you an ambitious opening position and a series of prepared retreats.

Because the ceiling on breadth is set by what already exists, a serious search is a prerequisite for good drafting rather than a formality. See prior art search for how it is done and patent search cost for what it takes. If you are filing early to secure a date, understand how a provisional interacts with claim scope — provisional patent applications covers that.

Common Drafting Mistakes

  • Copying the product specification into the claim. Founders are proud of the design and load the independent claim with every dimension, fastener, and material. The result protects the one prototype that was built and nothing else.
  • Marketing language. Words like convenient, efficient, or smart draw no technical boundary and invite an indefiniteness rejection.
  • Claiming a feature the description never mentions. You cannot add subject matter later that the original filing did not support, so a thin specification permanently limits your ability to amend.
  • Ignoring design changes. If the product evolves materially after filing, the claims may no longer cover what you actually sell — a good reason to keep filing in step with engineering progress.
  • Confusing claim types. Utility claims and design protection cover completely different things; see utility patent vs design patent.

What This Means for You as the Engineer

Even when an attorney writes the document, the engineer and founder are the only people who know which variations of the solution are genuinely possible: which component could be substituted, which parameter range still works, which alternative mechanism reaches the same result. The earlier you hand that over, the broader and more durable the claims will be.

The key question in every internal review is simple: what could a competitor change to get the same benefit without falling inside our list? Work through that honestly and you will find the gaps while they are still fixable. If you are improving on something that already exists, patenting an improvement explains how narrow claims can still be commercially strong. And if you are on the receiving end, claims are also what an infringement analysis turns on — see patent infringement and what to do.

Claims Are a Business Decision

Once you understand what claims are, they stop looking like legal formality and start looking like the document that determines the value of the asset. Budget and timeline context is in how much a patent costs and how long a patent lasts, and the wider subject is covered across our patents and intellectual property hub.

If you want to work through which engineering variations your claims ought to cover before you file — the alternative mechanisms, the parameter ranges, the substitutable components — describe your invention through our contact form.