The inventor describes what they built. The attorney asks a few good questions, gets a couple of alternatives on the record, and drafts the application. It goes to the patent office. A year or two later, a patent issues describing exactly what the company built.[1]
That is not a failure. It is also not finished. It is only the first pass, and most companies stop right there.
The Myth
The myth is that a thorough disclosure meeting produces a strong patent.
A complete description of what you built is the raw material for a patent. It is the starting point for further analysis, not the finished analysis.
Why the Belief Holds
A good attorney does more than transcribe. They ask what else the invention could look like, get a use case or two beyond the original one, and make sure the specification covers more than that first example. That is useful work, and it is where most disclosure meetings stop.
The specification looks complete because it is thorough. Thoroughness is not the same test as whether it captures what matters commercially.
A thorough description of what you built and a patent scoped to what matters commercially are two different documents, produced by two different lines of inquiry. The first line of inquiry starts with the invention and asks how else it could be built. The second line of inquiry starts with the customer. It asks what they are willing to pay, then works backward to figure out how much of the invention is the reason, and how much is just how this team happened to build it this time.
Almost nobody runs the second line of inquiry, because nobody in the room is positioned to run it. The inventor can tell you how the thing works. They cannot tell you, with any objectivity, which parts are the reason a customer buys it and which parts are incidental to their own environment, legacy stack, and vendor relationships. The attorney has no visibility into the sales calls, the competitive losses, or the CRM notes on why a deal was won or lost.[2]
The CEOs who catch this ask one question before the disclosure meeting is called finished: what is the sales team’s read on why customers buy? Does the patent we’re about to get protect that, or just how we happened to build it this time? That question belongs to whoever owns the product and the competitive picture, not the inventor describing their own work, and not an attorney seeing the technology for the first time.
The Second Pass
The second pass starts with a market question, not a technical one: what does the customer get from this? The customer might be the buyer, the user, or the person who approves the purchase. The question is not what the invention does mechanically. It is what the market’s behavior reveals about what they are willing to pay.
That behavior often contradicts what the inventor assumes. A customer paying the highest price in the category is not automatically buying the best feature set. Sometimes the price is the feature: the number itself signals quality, exclusivity, or risk reduction that a cheaper competitor cannot credibly claim, regardless of what the underlying components do. An invention built to justify a premium price protects something real, even when no single technical element in it is unique.
“Customer” also needs a harder look than the word implies. The person using a product and the person approving its purchase are often different people. Sometimes that split is organizational: for anything with real cost attached, the purchase runs through a buying committee, a procurement function, or a manager who never touches the product. Sometimes it is a gift: someone chooses and pays, and someone else opens the box and uses it, with no say in the decision. A feature that delights the end user carries no patent value if the decision that matters gets made by someone who never sees it. The patent has to protect whatever moves the buyer, not just whatever the user notices.
Once you have an answer to that, you go looking for every other way to deliver the same thing, starting with what your competitors do.
Take a product that uses one sensor to detect a condition, where a competitor’s product uses a different sensor for the same condition. Both deliver the same result to the customer. Before anyone writes a patent application naming a specific sensor, ask the business question first: are we and this competitor delivering the same value, built two different ways? If so, is there an advantage to building it the harder way?[3]
Sometimes the answer is no advantage at all. Then the sensor should never have anchored the patent; the protection belongs on whatever the two approaches share. Sometimes the sensor is cheaper, faster, or more capable in a way the customer can see or feel: fewer false positives, a lower price, a smaller footprint, a faster response. If there is a customer-visible advantage, document it. Write the specific capability the sensor enables into the patent, not the sensor itself.
This is also where a patent examiner’s favorite rejection appears. Swapping one known sensor for another is exactly the substitution an examiner calls obvious under 35 U.S.C. § 103,[4] absent a specific reason it is not. “We used a different sensor” is not a non-obviousness argument. “Every competitor’s approach requires calibration drift correction that adds cost and latency the customer can measure, and ours does not, because of how this sensor behaves” is one. It works only if true. It works only if someone did the work to find out before the patent was written.
The same question applies past a single component. A product can deliver a result through a local device, a mobile app, and a cloud service. The customer perceives one experience. The work behind it is split across three pieces of hardware owned by three parties.
Write the patent the way the product was built, and it maps every step to whichever device performs it in your current architecture. The result accurately describes your implementation. It may not match how a future competitor builds the same experience. Worse, it can split the invention across actors so no one of them infringes alone.
Write the patent around the customer-facing result instead. Name the mechanism that produces the advantage. Do not over-specify which device performs which step. That protection comes closer to a real constraint on anyone trying to deliver the same value.
What Excellent Looks Like
The process starts with the same market question everywhere it applies: what is the customer buying, and what would it take for someone else to sell them the same thing?
The disclosure holds many implementation choices: this sensor, this architecture, this data flow. Run each one through the same filter. Is this choice the reason the customer buys, or just how we happened to build it? If a competitor delivered the same customer-visible result through a different technical path, would the customer notice or care? If no, drop the detail from the patent. If yes, that detail is the invention. It deserves the sharpest, most specific patent language the specification can support. Gather the evidence for why it was not an obvious substitution before the application is submitted, not after an examiner raises the rejection.
This has to happen before a competitor test can mean anything, because the test itself is a design constraint on enforcement, not an afterthought. A patent’s value as a licensing asset, collateral for a loan, an entry ticket into a standards pool, or consideration in an acquisition all depend on one fact. Someone else, delivering the same value to the same customer, cannot avoid the patent without sacrificing something real.
A competitor can simply swap in a different sensor, device split, or off-the-shelf component. If that swap delivers the same customer experience at no real cost, the competitor never takes a license. They just make the swap. The patent described one company’s parts list, a constraint on nothing.
That is the purpose of naming specific competitors and describing how each would build the same result. It is not an academic exercise. It is the test of whether the patent survives contact with the one thing that determines whether it is worth anything: does a business, facing the choice between paying and switching, have an easy way to switch?
The Takeaway
The disclosure meeting that only captures what you built is the first pass, not the finished analysis. The second pass asks what the market is willing to pay. It stretches the invention across every other way that value could be delivered: a different sensor, a different split of work across devices, a different vendor’s component. It keeps only the patent language a competitor cannot walk around without sacrificing something the customer would notice.
That second pass is a business decision, not a legal one. It has to happen before the attorney starts drafting, not after an examiner or a competitor’s counsel points out that the patent only describes one company’s product. If your last few patents went straight from disclosure meeting to drafting with nobody asking what the market is willing to pay, that’s worth a short conversation before the next one does the same thing.
Sources
1 The Invention Disclosure Meeting Is Where Patent Value Is Decided argues the single most important event in a patent’s life is not filing, prosecution, or litigation — it is the disclosure meeting, and covers the two-meeting structure (a CIPO builds the business case before the attorney is ever briefed) for fixing it. ↩
2 Your Attorney Drafted Claims on Your Product — Not Your Competitor’s gives a four-question self-audit for testing whether a claim describes your own implementation or a real constraint on a named competitor, and covers why a published application teaches a competitor’s attorney exactly how to clear it. ↩
3 Stop Patenting Your Invention. Start Patenting Your Competitor’s Product. walks the competitor-comparison exercise in full: name real competitors, describe how each would implement the same feature, strip out what is specific to your own build, and keep only the common thread as the claim. ↩
4 35 U.S.C. § 103 (a claim is unpatentable as obvious if the differences between it and the prior art would have been obvious to a person having ordinary skill in the art). ↩
