A friend once told me, with total confidence, that he had "copyrighted" the name of his coffee shop. He had done no such thing. He had registered an LLC, painted a sign, and assumed the law would do the rest. What he actually wanted was a trademark, what he actually had was a thin set of common-law rights, and the word he reached for—copyright—described a body of law that had almost nothing to do with his problem. He is not unusual. People say copyright when they mean trademark, patent when they mean copyright, and "I'll just keep it secret" when what they need is a filed application with a priority date. The vocabulary is slippery because the underlying ideas really are easy to confuse: they all involve intangible things, they all give somebody the right to say "no," and they all live under the same umbrella word.

But "intellectual property" is not one thing. It is four distinct legal regimes, each created by a different statute, each protecting a different kind of asset, each arising in a different way, each lasting for a different length of time, and—this is the part that matters most in a courtroom—each tested for infringement by a different question. Copyright protects original creative expression. Trademark protects the brand signals that tell a consumer where a product comes from. Patent protects inventions. Trade secret protects valuable information that gives you an edge precisely because nobody else knows it. Pick the wrong regime and you protect nothing. Pick none and you protect nothing. Pick the right combination and you turn an idea into a portfolio of durable, enforceable assets.

This guide is built to be the map. We will take the four regimes and lay them flat, side by side, along six axes: what each protects, where the right comes from, how long it lasts, what registration does, how infringement is proved, and what remedies follow. Then we will do something most overviews skip: explain why the four regimes stay in their lanes—the "channeling" doctrine that the Supreme Court has enforced in cases like Baker v. Selden, Sears and Compco, and Dastar—because understanding the fences between the regimes is what separates a working knowledge of IP from a glossary. We will finish by running a single invented product through all four protections at once, and by giving you a decision framework you can actually use. For the deeper dive on any one regime, this article cross-references companion pieces throughout, including a focused comparison of the two most-confused regimes and a regime-by-regime treatment of the legal protection of software.

Why the Confusion Is So Easy—and So Costly

The four regimes feel interchangeable because they share a family resemblance. Each is a government-backed monopoly of sorts; each lets the owner exclude someone else from doing something; each protects something you cannot touch. From a distance they blur. Up close, they could hardly be more different.

The confusion is costly in specific, recurring ways. A founder who thinks her logo is "copyrighted" may never register the trademark, then watch a competitor adopt a confusingly similar name with impunity in a region she never expanded into. An engineer who believes his algorithm is "patented" because he wrote it down has confused copyright in the code with a patent on the method—and may have blown his patent rights entirely by publicly disclosing the invention more than a year before filing. A company that treats its customer list as a trade secret while emailing it around unencrypted to anyone who asks has not taken the "reasonable measures" that trade-secret protection requires, and will discover the gap only when a departing salesperson walks out the door with it.

These are not edge cases. They are the ordinary failure modes of treating four regimes as one. So the single most valuable habit a creator or business owner can build is to ask, for every intangible asset, which regime governs this?—and to recognize that the honest answer is sometimes "more than one." Let us build the map that lets you answer it.

The Big Picture: Six Axes, Four Regimes

Before the detail, the orientation. The four regimes can be told apart on six dimensions, and if you internalize only the table in your head, you already understand most of what matters.

Subject matter. Copyright covers original works of authorship fixed in a tangible medium—books, songs, films, photographs, software, architecture. Trademark covers source identifiers—words, logos, slogans, and sometimes shapes and colors that tell consumers who made the goods. Patent covers inventions—new and useful machines, processes, compositions of matter, and articles of manufacture (utility patents), new ornamental designs (design patents), and new plant varieties (plant patents). Trade secret covers information of competitive value that is kept secret—formulas, processes, methods, customer data, pricing models.

Source of the right. Copyright arises automatically the instant an original work is fixed—no filing, no fee, no government action. Trademark rights arise from use in commerce, strengthened (but not created) by federal registration. Patent rights do not exist until the government grants them after examination—there is no such thing as a common-law patent. Trade-secret rights arise from secrecy maintained by reasonable measures, with no government involvement at all.

Term. Copyright and patent are fixed-term regimes; trademark and trade secret are potentially perpetual. A copyright generally lasts the life of the author plus 70 years. A utility patent lasts 20 years from filing. A trademark lasts as long as you keep using it and maintaining the registration—forever, in principle. A trade secret lasts as long as it stays secret—which can be more than a century or can end at noon tomorrow.

Registration. Patents must be granted to exist. Trademark registration is optional but powerful. Copyright registration is optional for existence but a practical prerequisite to suing and to the best remedies. Trade secrets are never registered—registration would destroy the very secrecy that defines them.

The infringement test. This axis is the one practitioners care about most, because it determines what a plaintiff must prove. Copyright infringement turns on copying of protected expression. Trademark infringement turns on likelihood of confusion. Patent infringement turns on whether the accused thing falls within the claims. Trade-secret liability turns on misappropriation—improper acquisition, disclosure, or use.

Remedies. All four offer injunctions and damages, but the menus differ in revealing ways—statutory damages and fee-shifting in copyright, profits and trebling in trademark and patent, and a unique blend of injunctions, damages, and even seizure in trade-secret cases.

Hold those six axes in mind. Now we take each regime in turn, then circle back to set them against one another and to explain the fences between them.

Copyright: Protecting Expression

Copyright is the regime that asks for the least and gives, at the threshold, the most. The Copyright Act protects "original works of authorship fixed in any tangible medium of expression," 17 U.S.C. § 102(a), and the protected categories are sweeping: literary works, musical works, dramatic works, pantomimes and choreography, pictorial and graphic and sculptural works, motion pictures, sound recordings, and architectural works. Software counts as a literary work. So does this sentence, the moment it was saved to disk.

Two thresholds define what copyright reaches. The first is originality, and the bar is famously low: the work must be independently created and possess "at least some minimal degree of creativity," as the Supreme Court put it in Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991). Feist is the case that holds the alphabetical white pages of a phone book uncopyrightable—facts are not original to anyone, and a garden-variety alphabetical arrangement of them adds no spark. The second threshold is fixation: the work must be set down in a medium stable enough to be perceived "for a period of more than transitory duration," 17 U.S.C. § 101. An improvised jazz solo, never recorded, is not fixed and not protected. Record it, and the recording is.

The most consequential limit on copyright is not a threshold at all—it is the idea-expression distinction, codified at 17 U.S.C. § 102(b): copyright protection extends to no "idea, procedure, process, system, method of operation, concept, principle, or discovery." Copyright protects the particular way an author expressed something, never the underlying idea, fact, system, or method. This is the doctrine that lets a thousand authors write boy-meets-girl novels, lets every cookbook list "flour, sugar, eggs," and lets anyone explain double-entry bookkeeping in their own words. We will return to § 102(b) when we discuss channeling, because it is the doctrinal hinge that keeps copyright from swallowing patent.

Copyright's signature feature is that it arises automatically upon fixation. No registration is required for the copyright to exist. But registration with the U.S. Copyright Office is where the practical leverage lives, and the gap between "exists" and "enforceable" trips up more creators than any other rule in the field. Registration is a prerequisite to filing an infringement suit for U.S. works, and the Supreme Court held in Fourth Estate Public Benefit Corp. v. Wall-Street.com, LLC, 586 U.S. 296 (2019), that registration must actually be made—the Office must act on the application—not merely applied for, before suit. Timely registration (before infringement, or within three months of publication) also unlocks statutory damages of up to $150,000 per work for willful infringement and the possibility of recovering attorney's fees, 17 U.S.C. §§ 412, 504(c), 505. Without timely registration, a plaintiff is limited to actual damages and the infringer's profits, which are often hard to prove and small. So copyright is free and automatic, and the registration is the inexpensive step that makes it worth having. The mechanics are covered in our companion guides to registering a copyright with the U.S. Copyright Office and to copyright registration generally.

Copyright confers a bundle of exclusive rights under 17 U.S.C. § 106: to reproduce the work, prepare derivative works, distribute copies, and publicly perform and display it. Those rights are bounded by limitations, the most important of which is fair use, 17 U.S.C. § 107, a four-factor inquiry weighing the purpose and character of the use, the nature of the work, the amount used, and—often decisively—the effect on the market for the original. The Supreme Court's most recent major word on fair use, Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith, 598 U.S. 508 (2023), narrowed the role of "transformative" use, holding that where a copied work is put to a commercial purpose substantially the same as the original's, adding new meaning or aesthetics does not by itself tip the first factor toward the copier. Fair use is a deep enough subject to merit its own treatment; what matters here is simply that copyright's exclusive rights are not absolute.

The duration of copyright is long. For a work by an individual author, protection lasts for the life of the author plus 70 years, 17 U.S.C. § 302(a). For works made for hire and for anonymous or pseudonymous works, the term is 95 years from publication or 120 years from creation, whichever ends first, § 302(c). When the term expires, the work enters the public domain, free for anyone to use—which is precisely the fate that set up the Dastar case discussed below.

One modern access point deserves a mention: under the CASE Act, the Copyright Office launched the Copyright Claims Board (CCB) in June 2022, a voluntary three-member tribunal that hears small copyright claims (damages capped around $30,000) as a cheaper alternative to federal court. A respondent can opt out and force the dispute into court, but for individual creators and small businesses chasing modest infringements, the CCB is a forum that simply did not exist before.

Trademark: Protecting Source

If copyright protects what you made, trademark protects who you are—in the marketplace, at least. A trademark is a word, name, symbol, design, or combination that identifies and distinguishes the source of one party's goods from another's; a service mark does the same for services. The federal statute is the Lanham Act, 15 U.S.C. §§ 1051 et seq., and the animating idea behind the whole regime is source identification. A trademark is a promise of consistency: the swoosh tells you who made the shoe, the mark on the soda can tells you what is inside before you taste it. Because the mark's job is to prevent consumer confusion about source, the test for infringement is likelihood of confusion—whether the defendant's use is likely to confuse ordinary consumers about who stands behind the goods.

Not every word can be a mark, and the law sorts marks along a spectrum of distinctiveness that determines how much protection (if any) they get. Arbitrary or fanciful marks—APPLE for computers, KODAK for film—are inherently distinctive and strongly protected. Suggestive marks—NETFLIX, COPPERTONE—hint at the product and are also inherently distinctive. Descriptive marks—"Cold and Creamy" for ice cream—are protectable only if they acquire secondary meaning, i.e., consumers come to associate them with a single source. And generic terms—"bicycle" for bicycles—can never be marks, because allowing one seller to own the common name of the thing would hand them a monopoly over the product category itself. This spectrum traces to Judge Friendly's classic taxonomy in Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir. 1976), and it is foundational enough that we treat it at length in our trademark basics guide.

Trademark rights arise from use in commerce, not from registration, and this is the single most counterintuitive fact about the regime for newcomers. A business that uses a mark to identify its goods acquires common-law rights in the geographic area of that use, with nothing filed. But federal registration on the Principal Register transforms those rights: it provides nationwide constructive notice, a legal presumption of ownership and of the exclusive right to use the mark on the listed goods, the ability to invoke federal jurisdiction, and—after five years of continuous use—the prospect of incontestability, which forecloses many challenges to the mark's validity. We map the full hierarchy in our piece on common-law rights, the Supplemental Register, and the Principal Register, and we catalog the registration advantages in benefits of federal trademark registration. The short version: use creates rights, and registration supercharges them.

Two persistent confusions are worth naming. First, a business name is not automatically a trademark. Forming "XYZ, Inc." with a state corporation commission, or filing a "doing business as" name, lets you operate under that name but confers no trademark rights; a business name becomes a mark only when used to identify the source of goods or services, not merely to name the entity (see TMEP § 1202.01). Second, trademark protects a name only in context. The same word can be owned by different companies in unrelated fields—DELTA for faucets and DELTA for airlines coexist precisely because no consumer thinks the airline made the faucet. And a mark can be lost to genericide if the public starts using it as the name of the product rather than a brand: ASPIRIN, ESCALATOR, and CELLOPHANE were all once trademarks.

Trademarks can last forever, so long as the owner keeps using the mark and files the periodic maintenance documents (the Section 8 declaration of continued use and the Section 9 renewal) on schedule. A separate but related doctrine—functionality—keeps trademark from protecting useful product features, and it is one of the channeling fences we discuss below. Finally, the Trademark Modernization Act of 2020 added ex parte expungement and reexamination proceedings to clear unused "deadwood" registrations, and—crucially for enforcement—restored a rebuttable presumption of irreparable harm for trademark owners seeking injunctions, 15 U.S.C. § 1116(a), resolving a circuit split that had festered after the Supreme Court's patent decision in eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006).

Patent: Protecting Inventions

Patent is the regime that demands the most and, in exchange, gives the strongest right of all: the right to exclude even an independent inventor. The bargain at the heart of the patent system is disclosure for exclusivity. Society grants the inventor a time-limited right to exclude others from making, using, selling, offering to sell, or importing the invention, 35 U.S.C. § 271(a); in return, the inventor must teach the public how to make and use the invention, 35 U.S.C. § 112, so that the knowledge enters the public domain when the patent expires. The Patent Act, Title 35 of the U.S. Code, governs.

There are three kinds of patents. Utility patents protect the way an invention works—machines, processes, compositions of matter, articles of manufacture—and last 20 years from the filing date of the application, 35 U.S.C. § 154(a)(2). Design patents protect the ornamental appearance of an article and last 15 years from issuance for applications filed on or after May 13, 2015, 35 U.S.C. § 173. Plant patents protect new asexually reproduced plant varieties and run 20 years from filing. The line between design patents and trade dress is subtle enough that we devote a separate analysis to design patents versus trade dress.

To be patentable, an invention must clear several statutory bars. It must be patent-eligible subject matter under 35 U.S.C. § 101—and here the law excludes abstract ideas, laws of nature, and natural phenomena, the boundary the Supreme Court drew in Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014), and Mayo Collaborative Services v. Prometheus Laboratories, 566 U.S. 66 (2012). It must be novel under § 102 (not already known or disclosed), non-obvious under § 103 (not an obvious variation on the prior art to a person of ordinary skill), and useful. Since the America Invents Act of 2011, the United States operates on a first-inventor-to-file system, meaning priority generally goes to whoever files first—so delay can be fatal, and a public disclosure can start a one-year clock running against the inventor's own rights. The contours of patentability and the prosecution process are the subject of our patent basics and general information concerning patents guide and our plain-English patent guide.

The defining feature of patent infringement is that it is claim-based and strict. A patent ends with numbered claims—single-sentence definitions of the metes and bounds of the invention—and infringement turns entirely on whether an accused product or process falls within those claims, either literally or under the doctrine of equivalents. Intent is irrelevant to liability: an infringer who never heard of the patent and invented the same thing independently still infringes. This is the feature that most sharply distinguishes patent from trade secret, and it is the reason the patent-versus-secret choice (below) is so consequential. Because there is no common-law patent, a patent right does not exist until the USPTO grants it—a point that surprises inventors who assume that being "first to think of it" counts for something. Under first-to-file, it usually does not.

Trade Secret: Protecting Secrecy

The fourth regime is the one most often forgotten in the classic copyright-trademark-patent trio, and for many companies it is the most valuable of all. A trade secret is information that derives independent economic value from not being generally known and that the owner takes reasonable measures to keep secret. That two-part definition—value from secrecy, plus reasonable efforts to maintain it—appears in both the Uniform Trade Secrets Act (adopted in some form by nearly every state) and the federal Defend Trade Secrets Act of 2016 (DTSA), 18 U.S.C. §§ 1836 et seq., which created a federal civil cause of action for misappropriation, see § 1839(3). Classic examples run from the storied Coca-Cola formula to manufacturing processes, source code, customer lists, pricing models, and proprietary know-how.

Trade secrets are protected differently from everything else in this guide. There is no registration—indeed, registration would defeat the purpose, since it requires disclosure—and so there is no government office and no application. Protection arises and persists for as long as the information stays secret, which means a trade secret can last forever or evaporate the instant the secret becomes public. The legal protection runs against misappropriation: acquiring the secret through improper means (theft, bribery, breach of a confidentiality duty) or disclosing or using it in breach of such a duty, § 1839(5)–(6). The DTSA also offers an extraordinary remedy unavailable anywhere else in IP—an ex parte civil seizure order to recover misappropriated secrets in extraordinary circumstances, § 1836(b)(2)—and requires employers to give employees notice of the DTSA's whistleblower immunity in confidentiality agreements, § 1833(b).

The catch—the structural vulnerability that defines the regime—is that trade-secret law protects only against improper acquisition. A competitor who independently develops the same formula, or who reverse-engineers a product lawfully bought on the open market, infringes nothing. Both are entirely permissible. This is the mirror image of patent's strict liability, and it sets up the central strategic fork in all of IP, which we take up next. To do trade-secret protection right, an owner must actually earn the protection through reasonable measures—NDAs, access controls, "confidential" markings, exit interviews, network segmentation—because sloppy secrecy forfeits the right. Our companion guides on the protection of trade secrets and on building a trade secret protection program from scratch walk through the program a serious owner builds.

The Four Regimes, Side by Side

Having met each regime, set them against one another along the six axes. The contrasts are where the practical wisdom lives.

On subject matter, the regimes are nearly non-overlapping by design: expression (copyright), source identifiers (trademark), inventions (patent), and secret information (trade secret). The overlaps that do exist—software, which can be copyrighted as expression, patented as a method, and held as trade secret all at once—are the interesting cases, and we work one through below.

On source of the right, the spread runs from nothing required to everything required. Copyright needs only fixation. Trade secret needs only secrecy plus reasonable efforts. Trademark needs use (registration optional). Patent needs a granted application and nothing less.

On term, the regimes split cleanly into fixed (copyright, patent) and potentially perpetual (trademark, trade secret). This is not an accident of drafting; it reflects purpose. Patents and copyrights are time-limited because the constitutional grant authorizes Congress to secure rights only "for limited Times" to "promote the Progress of Science and useful Arts," U.S. Const. art. I, § 8, cl. 8—the public is meant to inherit the invention and the expression eventually. Trademarks and trade secrets are not "limited Times" creatures at all; the trademark protects an ongoing consumer-information function, and the trade secret protects only so long as the secret survives.

On registration, the lesson is procedural but high-stakes: patents must be granted; copyright and trademark registration are optional for existence but decisive for enforcement and remedies; trade secrets are never registered. Confusing the optional with the mandatory is how rights get lost.

On the infringement test, the four questions are genuinely different, and a litigator who imports one regime's test into another's case will lose. Did the defendant copy protected expression? (copyright). Is confusion likely? (trademark). Does the accused thing fall within the claims? (patent). Was the secret misappropriated? (trade secret). Note especially that copyright and patent both reach copying-like conduct but ask opposite questions about independent creation: independent creation is a complete defense to copyright infringement (you cannot infringe what you never saw) but no defense at all to patent infringement (the claims exclude everyone, original or not). Trade secret splits the difference—independent development is fine, but theft is not.

On remedies, every regime offers injunctions and monetary relief, but the levers differ. Copyright offers statutory damages and fee-shifting for timely-registered works, 17 U.S.C. §§ 504(c), 505. Trademark offers the defendant's profits, actual damages, and—in exceptional cases—trebled damages and fees, 15 U.S.C. §§ 1117(a)–(b), plus statutory damages for counterfeiting. Patent offers damages "in no event less than a reasonable royalty," with trebling for willfulness and fees in exceptional cases, 35 U.S.C. §§ 284–285. Trade secret offers actual loss, unjust enrichment or a reasonable royalty, exemplary damages up to twice the award for willful and malicious misappropriation, and the DTSA's unique seizure remedy, 18 U.S.C. § 1836(b)(3).

The Fences Between the Regimes: Channeling

Here is the part that overviews usually skip, and it is the part that makes the four-regime system actually cohere. Because the regimes overlap at the edges—and because the fixed regimes (patent, copyright) eventually surrender their subject matter to the public, while the perpetual regimes (trademark, trade secret) do not—there is a constant temptation to use a perpetual regime to do an end-run around an expired or unavailable fixed one. If you could trademark the functional shape of a product, you would get a forever-monopoly on an invention that patent law says must enter the public domain after 20 years. If you could copyright a system or method, you would get a 70-years-after-death monopoly on something patent law would protect for only 20 years and only after rigorous examination. Courts will not allow this. The doctrine that keeps each regime in its lane is called channeling, and three landmark cases define it.

Baker v. Selden, 101 U.S. 99 (1879), is the fountainhead of the idea-expression distinction and the patent-copyright fence. Selden held a copyright in a book describing a novel bookkeeping system, complete with example ledger forms. Baker used the same system with slightly different forms. The Supreme Court held that the copyright in the book did not give Selden a monopoly on the system it described. The explanation of an art may be copyrighted; the art itself can be protected, if at all, only by patent. To let copyright reach the useful system would be "a surprise and a fraud upon the public," because the public, seeing only a copyrighted book, would assume the underlying method was free to use. Baker v. Selden is the reason § 102(b) exists, and it is the reason your recipe's narrative can be copyrighted while the recipe's method cannot. It channels functional, useful subject matter out of copyright and toward patent.

Sears, Roebuck & Co. v. Stiffel Co., 376 U.S. 225 (1964), and its same-day companion Compco Corp. v. Day-Brite Lighting, Inc., 376 U.S. 234 (1964), draw the patent-versus-state-law fence. Stiffel held design and mechanical patents on a "pole lamp"; Sears sold a near-identical copy; the patents turned out to be invalid. Stiffel won under state unfair-competition law in the lower courts—but the Supreme Court reversed, holding that once an article is unprotected by a valid patent (or copyright), federal patent policy affirmatively permits copying, and a state may not use unfair-competition law to forbid what federal law leaves free. Compco applied the same logic to a lighting-fixture design. The Sears/Compco principle is that the public's right to copy unpatented, uncopyrighted articles is a federal entitlement; the perpetual machinery of state law cannot be used to recreate a patent that was never granted or has expired. (Later cases like Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141 (1989), confirmed and refined this preemption principle.)

Dastar Corp. v. Twentieth Century Fox Film Corp., 539 U.S. 23 (2003), polices the copyright-trademark fence and is the most practically important channeling case for modern businesses. Fox owned the copyright in a World War II television series and let it lapse into the public domain. Dastar took the public-domain footage, lightly repackaged it, and sold it as its own product. Fox—unable to sue for copyright infringement, because the copyright was gone—tried instead to use the Lanham Act, claiming Dastar's failure to credit Fox was a "false designation of origin." The Supreme Court rejected the maneuver. "Origin" in the Lanham Act, the Court held, means the origin of the physical goods, not the origin of the ideas or expression embodied in them. Reading trademark law to require attribution of authorship would create "a species of mutant copyright law" and would do exactly what Sears/Compco forbids—give a perpetual trademark right over expression that copyright had deliberately released to the public. Dastar is the reason you cannot use trademark law to manufacture a never-ending credit obligation once a copyright expires, and it is essential reading whenever a plaintiff reaches for the Lanham Act to remedy what is really a copyright or patent grievance.

There is a fourth, statutory fence worth knowing: the functionality doctrine in trademark and trade-dress law. A product feature that is functional—that affects the cost or quality of the article, or is essential to its use—cannot be protected as trade dress, no matter how much it identifies a source, because functional features are the province of patent. In TrafFix Devices, Inc. v. Marketing Displays, Inc., 532 U.S. 23 (2001), the Supreme Court held that an expired utility patent covering a product feature is "strong evidence" that the feature is functional and thus unprotectable as trade dress—again channeling functional matter away from the perpetual regime and toward the fixed one. We unpack this line in our analysis of design patent functionality.

The throughline of all four fences is the same: the perpetual regimes may not be used to recapture what the fixed regimes have released, or to protect what the fixed regimes alone are meant to cover. Once you see channeling, the four-regime system stops looking like four arbitrary buckets and starts looking like a deliberately engineered allocation of subject matter. That is the difference between memorizing the regimes and understanding them.

One Product, Four Protections: A Worked Example

Abstractions land best on a concrete object, so let us build one. (The following product is hypothetical, invented to illustrate how the regimes layer; it describes no real company.) Imagine a startup launches a new smart fitness watch. Walk the watch through the four regimes and watch the portfolio assemble itself.

The technology inside is a novel sensor array and a new method for measuring heart-rate variability. That is an invention, and the right tool is a utility patent. The startup files promptly—first-inventor-to-file means delay risks losing priority to a competitor and risks a self-inflicted disclosure bar. If granted, the utility patent excludes everyone, including a rival who independently develops the same sensor method, for 20 years from filing. After that, the method enters the public domain and competitors may practice it freely—exactly the bargain patent law strikes. This is why our freedom-to-operate analysis and patent-prosecution guides matter before launch, not after.

The distinctive look of the watch—the sculpted case, the unusual crown, the band's ornamental contour—is not how it works but how it looks, so it is the province of a design patent, protecting the ornamental design for 15 years from issuance. Care is required here: any feature that is functional rather than ornamental falls outside design-patent and trade-dress protection under TrafFix, and may need to ride on the utility patent instead.

The software and the user interface—the firmware, the app, the carefully drawn graphics of the watch face—are original works of authorship, protected by copyright automatically on fixation and enforceably upon registration. Copyright will protect the expression of the code and the screens, but not the functional ideas they implement; for that functional layer, the startup leans on patent and trade secret. (The interplay among copyright, patent, and trade secret in software is intricate enough that we devote a whole article to the legal protection of software.) The copyright will outlive everyone in the room—life-plus-70, or 95/120 years for the work-made-for-hire code.

The brand—the watch's name, the company logo, the tagline on the box—is a set of source identifiers, protected by trademark. Rights begin with use in commerce and are supercharged by federal registration on the Principal Register. If the company keeps selling watches under the name and files its maintenance documents, the trademark can last as long as the company does—forever, unlike the patents and even the copyright.

The confidential edge—the proprietary manufacturing process that lets the startup build the sensor at half a rival's cost, the tuning parameters of the heart-rate algorithm that were not disclosed in the patent, the customer and usage data—are trade secrets, protected for exactly as long as the company keeps them secret through NDAs, access controls, and disciplined information hygiene. Note the deliberate division of labor: the startup patented the sensor method (disclosure for a 20-year exclusive) but kept the tuning parameters as a trade secret (secrecy for a potentially perpetual but fragile edge). That is not redundancy; it is strategy.

Step back and look at what one product has become: a layered portfolio in which each asset is matched to the regime that fits it, each arose in a different way, and each expires—or does not—on a different schedule. The utility patents lapse first, opening the core technology to competitors. The design patent follows. The copyrights endure for generations. The trademark can run as long as the brand sells watches. The trade secrets last precisely as long as the secrecy holds. A founder who understood "only copyright" or "only patent" would have left most of this product naked. Mapping each asset to the right regime—and layering them deliberately—is what comprehensive IP protection actually looks like.

The Patent-Versus-Trade-Secret Fork

Of all the choices in this guide, one recurs in nearly every product decision and deserves its own discussion: when you have a patentable innovation that you could also keep secret, which do you choose? The two regimes are near-perfect opposites, and they are mutually exclusive at the core, because a patent requires public disclosure and a trade secret requires secrecy. You generally cannot do both with the same information.

Patent gives you a strong but time-limited right. You disclose the invention to the world, you get 20 years of exclusivity, and that exclusivity reaches everyone—including an independent inventor and a reverse-engineer. The price is disclosure (your competitors get a roadmap), cost and delay (prosecution takes years and real money), and the hard ceiling of the term (after 20 years, it is everyone's).

Trade secret gives you a potentially perpetual but fragile right. You disclose nothing, you incur no filing cost, and the protection can last forever—but it protects you against only improper acquisition, never against independent development or lawful reverse engineering. The Coca-Cola formula has outlived a century of patents that would long since have expired; but if a chemist reverse-engineers it tomorrow, the law offers no recourse.

The decision turns on a few practical questions. Can the innovation actually be kept secret? A manufacturing process behind a factory wall can be; a feature visible in the shipped product (and thus reverse-engineerable) usually cannot, which pushes it toward patent. How long will it stay valuable? If the technology will be obsolete in five years, a 20-year patent's term advantage is illusory and secrecy may suffice. How easily could a competitor independently reach the same result? If the answer is "easily and soon," trade secret's fatal vulnerability is exposed and patent's exclusion of independent inventors becomes worth its price. And can you bear the disclosure? Some companies will not file because the published application teaches rivals too much. There is no universally right answer—only a fit between the innovation and the regime, which is exactly the judgment good IP counsel is paid to make.

What Each Regime Does Not Protect

Knowing the gaps is as valuable as knowing the coverage, because the gaps are where confident people lose rights.

Copyright does not protect ideas, facts, systems, methods, names, titles, or short phrases. It protects the expression, never the idea (§ 102(b), Baker v. Selden). A plot premise, a recipe's ingredient list, a business method, a song title, a brand name—none is copyrightable, though the creative expression around them often is. The instinct that copyright covers a name or slogan is simply wrong; that is trademark's domain.

Trademark does not protect a name in the abstract, and never protects a generic term. A mark protects a name only as a source identifier, only in connection with particular goods or services, and only against confusingly similar uses. The same word can belong to different owners in unrelated fields, and the common name of a product can never be owned—the genericide that claimed ASPIRIN and ESCALATOR is the permanent risk of a too-successful brand. Trademark also does not protect functional product features; TrafFix channels those to patent.

Patent does not protect abstract ideas, laws of nature, natural phenomena, or an invention you merely conceived but never properly and timely filed. A formula, a fundamental principle, or a substance as it exists in nature is ineligible under § 101 and Alice/Mayo; only a specific, novel, non-obvious, useful application qualifies. And under first-inventor-to-file, the inventor who sat on the idea loses to the one who filed.

Trade secret does not protect against independent discovery or reverse engineering. It guards only against misappropriation. A competitor who independently develops your formula, or who lawfully reverse-engineers your shipped product, infringes nothing. This is the central vulnerability that distinguishes trade secret from patent, and misunderstanding it is how companies "protect" something that was never actually protected.

A Decision Framework: Choosing and Combining

For the owner with an asset in hand, the practical task is a short sequence of questions. First, characterize the asset. A brand or source signal → trademark. A functional invention → patent (or trade secret). A creative work → copyright. Valuable confidential information → trade secret. Many assets fit more than one box—software is the paradigm case—and the boxes can be combined.

Second, run the patent-versus-secret fork for any patentable innovation: disclosure and a fixed monopoly, or secrecy and a perpetual but fragile edge, judged against secrecy feasibility, useful life, and reverse-engineerability.

Third, do what each regime requires, on time. File patent applications promptly—first to file wins, and your own public disclosure starts a clock. Register copyrights early to preserve the right to sue and to claim statutory damages and fees. Pursue trademark registration to upgrade common-law rights into nationwide ones, and clear your mark first with a proper trademark clearance search. Implement real secrecy measures—NDAs, access controls, exit procedures—to keep trade secrets alive.

Fourth, maintain and police. Pay trademark and patent maintenance fees, keep using the mark, renew on schedule, and watch the market for infringers, because a right you never enforce is a right that erodes. And fifth, think internationally and early, because IP rights are territorial: a U.S. patent, trademark, or trade-secret protection generally reaches only within the United States. The Madrid Protocol streamlines multinational trademark filing, the Patent Cooperation Treaty preserves international patent options while deferring national-stage costs, the Berne Convention extends copyright protection across member countries without formalities, and the TRIPS Agreement sets a baseline of trade-secret protection among WTO members. For patents and trademarks especially, deadlines abroad are unforgiving—waiting can forfeit foreign rights for good.

The unifying lesson is that IP protection rewards the deliberate. Copyright is the lone regime that arises on its own, and even it needs registration to enforce. Everything else demands affirmative, timely steps. The companies that protect their intangible assets well are the ones that identify each asset early and reach for the right tool before a dispute, not after.

Frequently Asked Questions

Can I copyright my company name or logo? Not usefully. A name or short slogan is generally too short to be copyrightable, and even a stylized logo's source-identifying function is a trademark matter, not a copyright one. A logo with enough original graphic creativity may carry a thin copyright in its artwork, but the protection that actually stops competitors from confusing consumers is trademark. The two regimes are confused so often that we wrote a dedicated comparison.

Do I need to register a copyright for it to exist? No. Copyright exists automatically on fixation. But you generally cannot sue for infringement of a U.S. work until the registration is made (Fourth Estate, 2019), and you cannot recover statutory damages or attorney's fees unless you registered before the infringement or within three months of publication (17 U.S.C. § 412). Registration is cheap insurance; skip it and you keep the right but lose most of its teeth.

If I keep my invention secret, can I still patent it later? Up to a point, and the risk is real. A trade secret kept too long can become un-patentable if it counts as prior art against your own later application, and a public use or sale can trigger the on-sale and public-use bars. More practically, the two regimes are mutually exclusive for the same information: patent requires you to disclose what trade secret requires you to hide. Decide deliberately; do not drift.

Can the same product be protected by all four regimes at once? Yes—commonly. As the smart-watch example shows, the invention (patent), the look (design patent), the software and UI (copyright), the brand (trademark), and the confidential process (trade secret) each map to a different regime, and a sophisticated owner layers them. The regimes are not mutually exclusive across different assets; they are only mutually exclusive where channeling forbids the overlap (e.g., functional features cannot be both patented and held as trade dress, and a system cannot be copyrighted to avoid patenting it).

Why can't I just trademark a useful feature of my product so it lasts forever? Because the functionality doctrine and the channeling principle forbid it. TrafFix (2001) holds that functional features—and especially those once covered by a utility patent—cannot be protected as trade dress, because that would let a perpetual regime recapture what patent law releases to the public after the patent expires. Functional features belong to patent, full stop.

What is the single most common mistake? Treating the four regimes as one and assuming protection is automatic. It is automatic for exactly one regime (copyright, on fixation), and even there enforcement requires registration. For everything else—patents, trademark registrations, and the secrecy program behind trade secrets—rights come only from deliberate, timely action.

Related Articles

This article is provided for general informational purposes and does not constitute legal advice. Intellectual property law is detailed, fact-specific, and full of exceptions to the rules summarized here. Consult qualified IP counsel about protecting any particular asset.