Picture a founder we will call Dahlia. She has spent eight months and most of her savings perfecting a line of cold-brew coffee concentrate. The recipe is dialed in, the bottles look gorgeous, and one sleepless night she lands on a name she loves: AURORA. She buys the domain, prints a thousand labels, and tells her cousin to start the Instagram. What she has not done—what almost nobody does first—is ask whether she is allowed to use that name, whether anyone else already owns it, and how she might turn it into a legal asset she can defend, license, sell, or pass down.

That gap between I picked a name and I own a brand is the trademark process. It is a journey with a defined map: choose a strong mark, clear it, file the application, survive examination, get published, dodge or defeat any opposition, prove your use, register, and then maintain and police the registration for as long as you keep using the mark. Each stage has its own rules, its own deadlines, and its own ways to trip. This guide walks the whole arc end to end, in plain language, with worked examples and the governing law, so that a judge, a transactional lawyer, and Dahlia herself can all follow along.

A note on where this fits, because mclaw.io has a small library on this topic and you should land on the right shelf. If you want the gentle conceptual primer—what a trademark even is, the difference between ™ and ®, why a source identifier matters—start with our trademark basics article. If you want the nuts-and-bolts filing playbook (which forms, which fees, how to draft a goods-and-services description), see the trademark registration guide and our step-by-step how to file a trademark application with the USPTO. For the deeper doctrine—what qualifies as a mark and the legal standards it must meet—the four-part overview series (subject matter, substantive standards, obtaining protection and licensing, and infringement and related rights) is the treatise. This article is the wide-angle lifecycle: the story of a mark from a glimmer of an idea to a registration you renew for the third time twenty years later. Think of it as the trail map; the others are the trailhead sign, the detailed topo of the hardest mile, and the geology textbook.

First, the big picture: what is a trademark protecting?

A trademark is a source identifier. It is any word, name, symbol, design, sound, color, scent, or combination that tells a buyer where a product or service comes from. The legal magic is not in the word itself but in the link it forms in a consumer's mind between the mark and a single (even if anonymous) commercial source. That is why the law speaks not of "owning a word" but of owning rights in a mark for particular goods or services. DELTA is a faucet company, an airline, and a dental insurer, and all three coexist because they sell unrelated things to non-overlapping customers. (Class lines are part of how the system keeps them apart; see USPTO trademark classes.)

The federal statute governing all of this is the Lanham Act, codified at 15 U.S.C. §§ 1051–1141n. Section 45 (15 U.S.C. § 1127) supplies the core definitions, including "trademark," "service mark," and the all-important phrase "use in commerce." The U.S. Patent and Trademark Office (USPTO) administers the federal register; its examining procedures live in the Trademark Manual of Examining Procedure (TMEP), its adjudications happen before the Trademark Trial and Appeal Board (TTAB), and the day-to-day rules of practice are codified in Title 37 of the Code of Federal Regulations (37 C.F.R. parts 2, 7, and 11). Keep those reference points handy—the Lanham Act, the TMEP, the C.F.R., the USPTO, and the TTAB—because they recur at every stage below.

One thing the trademark process does not do is protect inventions or creative works. Patents protect inventions; copyrights protect original works of authorship. If Dahlia invented a novel cold-brew extraction machine she would seek a patent; if she wrote a jingle for the AURORA ad she would have a copyright. The brand name is the trademark. For the clean side-by-side, see copyright vs. trademark: what is the difference. And if you are weighing the calendar—register now or later—our piece on when to trademark your brand is the companion to read alongside this one.

Stage One: Choose a mark you can actually own (the distinctiveness spectrum)

Here is the counterintuitive truth that bites more startups than any other: the names that sound most appealing to a marketer are often the weakest in law, and the names that sound oddest at first are often the strongest. Trademark strength is measured along a continuum that lawyers call the distinctiveness spectrum, and the foundational case that organized it is Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir. 1976). Judge Henry Friendly sorted marks into a ladder of five categories, and where your mark sits on that ladder determines whether you can register it at all and how easily you can stop copycats. (We unpack this doctrine fully in trademark overview: substantive standards for protection; here we use it as the launch pad for the whole process.)

Generic terms sit at the bottom and are never protectable as marks for the thing they name. "Coffee" for coffee, "Bicycle" for bicycles, "Email" for email. You cannot fence off the common name of the product, because then competitors could not even describe what they sell. The Supreme Court refined the rule in USPTO v. Booking.com B.V., 591 U.S. 549 (2020), holding that adding a top-level domain to a generic term can—depending on consumer perception—yield a non-generic whole ("Booking.com" was registrable), but it left untouched the bedrock principle that a generic term standing alone cannot be a trademark. A related nightmare is genericide, where a once-valid mark becomes the common name through the public's own usage—aspirin, escalator, thermos, and cellophane all died this way. (Xerox and Google fight ferociously to avoid joining that graveyard, which is why their lawyers send polite letters reminding journalists to "search on Google," not "google it.")

Descriptive marks sit just above generic and describe a feature, quality, function, or characteristic of the goods. "Creamy" for yogurt, "Speedy" for delivery, "Cold-Brew" for coffee concentrate. Descriptive marks are not registrable on the Principal Register out of the gate. The Lanham Act's Section 2(e)(1), 15 U.S.C. § 1052(e)(1), bars registration of marks that are "merely descriptive" of the goods. But there is an escape hatch: if a descriptive mark acquires secondary meaning—meaning consumers have come to associate it with one source through long use and marketing—it can be registered under Section 2(f), 15 U.S.C. § 1052(f). The classic illustration is the Supreme Court's discussion in Park 'N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189 (1985). Five years of substantially exclusive and continuous use creates a basis for the USPTO to accept acquired distinctiveness, but acceptance is discretionary, not automatic, and a highly descriptive term often demands far more proof. Descriptive marks are a long, expensive road; many applicants get parked on the Supplemental Register (a secondary roster for capable-but-not-yet-distinctive marks) while they build the evidence.

Suggestive marks are the sweet spot for many businesses. They suggest a quality of the goods but require a mental leap—a flash of imagination—to connect mark to product. COPPERTONE for suntan lotion, NETFLIX for streaming, JAGUAR for cars, GREYHOUND for buses. Suggestive marks are inherently distinctive: they are registrable immediately, no secondary meaning required, yet they still whisper something flattering about the product. The line between "merely descriptive" and "suggestive" is the single most litigated boundary in trademark prosecution, and reasonable examiners disagree about it constantly.

Arbitrary marks use real words that have nothing to do with the goods. APPLE for computers, CAMEL for cigarettes, AMAZON for an online store, SHELL for gasoline. The word is ordinary; its application is unexpected. These are strong and easy to register.

Fanciful marks are invented words coined specifically to function as a brand—KODAK, XEROX, EXXON, PEPSI, VERIZON. They are the strongest of all because they carry no baggage, mean nothing else, and are almost impossible for a competitor to claim they "needed" to use. The trade-off is marketing cost: you have to teach the public what a fanciful word means from scratch.

So back to Dahlia. AURORA for coffee concentrate is arbitrary—a real word (the dawn, the Northern Lights) with no connection to cold brew. That is a strong choice, legally. Had she gone with MORNING BREW she would be flirting with descriptiveness; COLD-PRESSED COFFEE CO. would be nearly generic and a registration headache. The lesson, which we drum into every founder, is to spend distinctiveness like the asset it is. A slightly less literal name today saves a fortune in enforcement and rebranding tomorrow.

Worked example (hypothetical). Two friends each open a bakery. Sam names his shop "Fresh Bread Bakery." Priya names hers "ZEPHYR Bakehouse." A year later a competitor opens "Fresh Loaf Bread" next door to Sam and "ZEPHYR Coffee & Bread" next door to Priya. Sam will struggle—"fresh bread" is descriptive bordering on generic, and he has almost nothing to enforce. Priya, holding an arbitrary mark, can move quickly. Same effort, wildly different leverage, decided entirely at the naming stage.

A note before we leave naming. Distinctiveness is the registrability question, but it is not the only Section 2 gate. The statute also bars (among others) marks that are primarily merely a surname, deceptive matter, and matter that falsely suggests a connection with persons or institutions. The Supreme Court has recently trimmed two of these bars on First Amendment grounds—striking the disparagement bar in Matal v. Tam, 582 U.S. 218 (2017), and the "immoral or scandalous" bar in Iancu v. Brunetti, 588 U.S. 388 (2019)—while upholding the "names" clause in Vidal v. Elster, 602 U.S. 286 (2024). For the full subject-matter map, see trademark overview: subject matter of trademark law.

Stage Two: Clearance—look before you leap

Choosing a strong mark is necessary but not sufficient. The next question is whether someone else already has rights that would block you. This is clearance, and skipping it is how brands end up sending out a thousand labels they later have to shred.

Clearance has two enemies to find. The first is a confusingly similar mark on the federal register or in a pending application—because under Section 2(d) of the Lanham Act, 15 U.S.C. § 1052(d), the USPTO will refuse a mark that so resembles a registered or prior-pending mark as to be likely to cause confusion. The second enemy is an unregistered, common-law user who got there first. U.S. trademark rights flow from use in commerce, not from registration; a business that has been selling under a name in a region can have superior rights even without a federal registration. We unpack that doctrine fully in trademark rights under common law, including the famous Tea Rose–Rectanus geographic-rights rule from United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90 (1918), and its modern application in Stone Creek v. Omnia, where knowledge of the senior user destroyed the junior user's good-faith defense.

A real clearance search therefore looks well beyond the USPTO's public search system. It covers state registrations, business-name and domain registries, app stores, social handles, and a broad internet sweep for unregistered users. A thorough professional search also evaluates related goods and phonetic and visual near-misses, not just identical hits—because confusion is about the marketplace, not about spelling. Our deep dive on how to conduct a comprehensive trademark clearance search is the companion resource here, and it is worth reading before you fall in love with a name.

There is a second, underappreciated reason to clear carefully: a documented clearance search and a clean attorney opinion can later shield you from a finding of willful infringement if a dispute arises. Good faith is a defense to the worst remedies. We cover that strategic upside in the shield of good faith. Clearance is not just defense against an examiner; it is insurance against an enhanced-damages verdict.

Worked example (hypothetical). Dahlia's search turns up "AURORA" registered for energy drinks and a small Oregon roaster using "AURORA COFFEE" on its café signage for the last six years without any federal registration. Both are red flags. The energy-drink registration could trigger a Section 2(d) refusal because beverages are related goods; the Oregon roaster has common-law priority in its region and might oppose or, worse, sue. Better to learn this before the labels print than after. Depending on the results, Dahlia's counsel might recommend narrowing her goods description, adjusting the mark (AURORA BREW CO.), negotiating a coexistence agreement, or pivoting to a different name entirely. That is exactly the kind of call clearance is meant to surface early and cheaply.

Stage Three: Pick your filing basis

Once the mark is chosen and cleared, you file an application with the USPTO—and the first substantive decision is your filing basis, the legal hook that says why you are entitled to register. The Lanham Act offers several, and the two that matter most for U.S. businesses are:

Section 1(a)—use in commerce. You are already using the mark in interstate (or U.S.–foreign) commerce on the goods or services you list. 15 U.S.C. § 1051(a). You must submit a specimen—real-world evidence of the mark as actually used, such as a product label, packaging, or a screenshot of a live e-commerce page where customers can buy. You also state the date of first use anywhere and first use in commerce.

A crucial and frequently misunderstood point: "use in commerce" means actual sale or transport of the goods (or rendering of the services) bearing the mark, not preparation, advertising of a coming service, or token gestures. The Federal Circuit drove this home in Couture v. Playdom, Inc., 778 F.3d 1379 (Fed. Cir. 2015), where it canceled a service-mark registration because the applicant had only posted a "website coming soon" page advertising services he had not yet rendered. Merely offering a service in the future is not using the mark; the service must have actually been provided to a customer. File a 1(a) application before you have genuine use, and you have built your registration on sand.

Section 1(b)—intent to use (ITU). You have a bona fide intention to use the mark but have not started yet. 15 U.S.C. § 1051(b). This is enormously useful for founders: it lets you reserve a place in line before launch. The catch is that a 1(b) application does not mature into a registration until you actually use the mark and prove it with a Statement of Use (more on that below). The "bona fide intention" must be genuine and ideally documented—business plans, prototypes, marketing drafts, manufacturing inquiries. A naked intent to warehouse names, unsupported by any objective evidence of real plans, can void the application; the TTAB regularly invalidates ITU filings where the applicant cannot produce a scrap of contemporaneous documentary support for its claimed intent.

Two other bases serve foreign applicants. Section 44 (15 U.S.C. § 1126) lets an applicant rely on a home-country application (44(d), claiming a priority date based on the foreign filing, available for six months under the Paris Convention) or registration (44(e)). Section 66(a) (15 U.S.C. § 1141f) is the Madrid Protocol route, an extension of an international registration to the United States, administered through the World Intellectual Property Organization. These bases generally do not require a pre-registration specimen, though use is still expected to maintain the registration later—and, as we will see, a Madrid-based U.S. registration is maintained under Section 71 rather than Section 8. For the cross-border timing playbook, see [registering a Madrid Protocol trademark] within our broader coverage and when to trademark your brand. For most domestic small businesses, though, the real choice is 1(a) versus 1(b), and it turns on one question: have you started selling yet?

Worked example (hypothetical). Dahlia hasn't sold a single bottle—she is still finalizing distribution. She should file under Section 1(b) to lock in priority now, then prove use later. Her cousin, who already sells handmade mugs branded "AURORA" on Etsy with live listings, would file under Section 1(a) and attach the listing as a specimen. Same family of products, two different bases, because one is selling and one is not yet.

Stage Four: The application itself (and how TEAS works)

We keep the mechanics brief here because the trademark registration guide and how to file a trademark application with the USPTO cover them in granular detail, and USPTO trademark classes covers classification in particular. But the lifecycle view needs the essentials.

Applications are filed electronically and only electronically. Since the USPTO's mandatory e-filing rule took effect on February 15, 2020, virtually all formal trademark correspondence—applications, office-action responses, and maintenance filings—must go through the Trademark Electronic Application System (TEAS). 37 C.F.R. § 2.23(a). Paper submissions are rejected and, eventually, destroyed, save for a short list of narrow exceptions (certain non-traditional-mark specimens, postal-service emergencies, and the like). You also need a verified USPTO.gov account, and the rules require the applicant's and (if any) attorney's actual email address and domicile address on the record. The core ingredients of the application are:

  • The applicant. The correct legal owner—the entity or person who controls the mark and the nature and quality of the goods. Naming the wrong owner (say, the founder personally instead of the LLC that actually sells) can be a fatal, sometimes un-fixable, defect that voids the application, because an applicant who did not own the mark at filing never had a valid application to begin with.
  • The mark and its format. A standard-character mark (the words, in any font—maximum flexibility) or a special-form/design mark (a specific stylization or logo, requiring a drawing). The drawing is the legally operative depiction of the mark, and it must match the specimen: the public sees the drawing when the mark publishes and on the eventual certificate. You can also seek non-traditional marks—sounds, colors, scents—with heightened proof.
  • The goods and services. A precise identification, classified under the international Nice system into one or more of 45 classes (34 for goods, 11 for services). Vague or overbroad descriptions invite refusals; the USPTO's Trademark ID Manual (the Acceptable Identification of Goods and Services Manual) is the safe harbor, and selecting pre-approved language from it both speeds examination and qualifies you for the cheaper TEAS Plus fee tier.
  • The filing basis, as discussed above, with a specimen if 1(a).
  • A verified statement (declaration) signed under penalty of perjury—carrying the statutory warning that willful false statements are punishable under 18 U.S.C. § 1001—plus the government filing fee, which is assessed per class. A multi-class application multiplies the fee by the number of classes.

A heads-up on representation: applicants whose domicile is outside the United States are required to be represented by a U.S.-licensed attorney. 37 C.F.R. § 2.11. Domestic applicants may file pro se under 37 C.F.R. § 11.14(e), but the USPTO itself strongly encourages counsel, and the error rate among unrepresented applicants is high. Trademark prosecution looks deceptively simple and is full of quiet landmines.

Once filed, the application gets a serial number and a filing receipt, sits in a queue, and is eventually assigned to a USPTO examining attorney. Expect the overall path to registration to take roughly a year in a smooth case, and considerably longer if refusals, oppositions, or the USPTO's backlog intervene; recent backlogs have pushed first office actions out many months past filing. Throughout, you (or your attorney) must monitor status through the USPTO's Trademark Status and Document Retrieval (TSDR) system and keep your correspondence email current—missed mail is no excuse for a missed deadline.

Stage Five: Examination—when the USPTO pushes back

Examination is where the trademark process earns its reputation for difficulty. The examining attorney reviews the application for compliance with every applicable rule and conducts a search of the register for conflicting marks. If the attorney spots a problem, the USPTO issues an Office Action—a letter explaining the refusals and requirements. Here is a change that catches the unwary: since December 3, 2022, applicants generally have three months to respond to an office action (not the old six), with a single three-month extension available for a fee, after which the application goes abandoned. (The shorter clock came in under the Trademark Modernization Act of 2020, discussed below; Madrid Section 66(a) applications still run on a six-month response period.)

Office Actions come in two flavors. Some are minor and procedural—amend the goods description, clarify the entity type, disclaim an unregistrable component, or submit a substitute specimen. Others are substantive refusals that go to whether the mark can be registered at all. The two substantive refusals you must understand are likelihood of confusion under Section 2(d) and descriptiveness under Section 2(e).

The Section 2(d) refusal: likelihood of confusion

Under 15 U.S.C. § 1052(d), the USPTO will refuse registration of a mark that so resembles a mark already registered (or in a prior-filed pending application) that it is likely to cause confusion, mistake, or deception. This is the single most common substantive refusal. The examining attorney weighs the factors set out in In re E.I. DuPont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973)—the DuPont factors—which include the similarity of the marks in appearance, sound, connotation, and commercial impression; the relatedness of the goods or services; the channels of trade; the conditions of purchase (impulse buy versus careful, expensive decision); the strength of the cited mark; and the number and nature of similar marks on similar goods. In a TTAB or ex parte context the two heavyweight factors are usually the similarity of the marks and the relatedness of the goods, and the examiner must resolve doubt in favor of the prior registrant.

Crucially, the goods do not have to be identical, and the marks do not have to be identical. Confusion is the test, not copying. "AURORA" coffee versus "AURORA" energy drinks could draw a 2(d) refusal because beverages travel in overlapping channels and a consumer might assume a common source. (Litigation in court uses circuit-specific multi-factor tests—Polaroid in the Second Circuit, Sleekcraft in the Ninth, and so on—that parallel DuPont; for the courtroom version of this analysis see navigating the maze of trademark confusion and trademark overview: infringement and related rights.)

How do you overcome a 2(d) refusal? Options include arguing that the marks differ in meaning or impression (a different connotation can defeat similar spelling), that the goods are unrelated and move in different channels, that the cited mark is weak because the register is crowded with similar marks, that consumers are sophisticated and careful, or—often the most practical fix—amending your identification of goods to carve out the overlap. Sometimes the answer is a consent or coexistence agreement with the cited registrant, which the USPTO gives real (though not automatic) weight, especially when the agreement is detailed about how the parties will avoid confusion. And sometimes the right move is to argue the cited registration is vulnerable—dead wood that was never properly used—and pursue a separate cancellation or, under the new TMA tools described below, an expungement or reexamination proceeding.

The Section 2(e) refusal: descriptiveness (and its cousins)

Section 2(e), 15 U.S.C. § 1052(e), bars several categories of marks, but the workhorse is 2(e)(1): "merely descriptive." If your mark directly conveys a feature, quality, or characteristic of the goods—remember "Cold-Brew" for coffee—the examiner will refuse it on the Principal Register. The test asks whether the mark immediately tells a consumer something about the product without a mental leap; if it requires imagination, thought, or perception to reach a conclusion about the goods, it is suggestive and registrable, and if it conveys the information straightaway, it is descriptive and refused.

Section 2(e) has several siblings worth knowing because they trap unwary applicants:

  • 2(e)(2): primarily geographically descriptive—"NAPA" for wine from Napa, "BROOKLYN" for goods from Brooklyn.
  • 2(e)(3): primarily geographically deceptively misdescriptive—a geographic term that is false as to origin and material to the purchasing decision.
  • 2(e)(4): primarily merely a surname—which is exactly why trademarking your own name is harder than people expect; surnames need acquired distinctiveness, and names of living individuals require their written consent under Section 2(c).
  • 2(e)(5): functional matter—you cannot trademark a utilitarian product feature; that is the province of patents. This is the trade-dress functionality bar, and it is why design and configuration claims are so contested (see the intricate world of trade dress protection).

The responses to a 2(e)(1) descriptiveness refusal track the distinctiveness spectrum from Stage One. You can argue the mark is actually suggestive, not descriptive (the imagination test). You can claim acquired distinctiveness under Section 2(f) by showing the mark has come to identify your company as the source—often via the five-year presumption, advertising spend, sales volume, media coverage, look-for advertising, and survey evidence. Or you can accept registration on the Supplemental Register, which gives you the right to use the ® symbol, the ability to sue in federal court, a bar to confusingly similar later marks, and a place in line, while you build the secondary-meaning record needed to graduate to the Principal Register later.

Worked example (hypothetical). Suppose Dahlia had named her concentrate "SMOOTH POUR." An examiner might call that descriptive of a quality of the product. Her counsel could argue it is suggestive (it takes a small leap—"smooth" could describe taste, texture, or the pour itself), submit evidence that the phrase is used arbitrarily in the marketplace, or, if she has been selling for years, claim acquired distinctiveness under 2(f). With "AURORA," none of this arises—arbitrary marks sail past 2(e). Again, the naming choice pays dividends (or extracts costs) all the way down the line.

A special headache: the specimen refusal

Even when a mark is registrable in the abstract, an application can die over a bad specimen—the evidence of real-world use. Specimen problems are among the most common reasons applications stall, and the USPTO has only grown stricter, in part because it caught a wave of fraudulent, digitally fabricated specimens (often from overseas filing mills) and responded with Examination Guide 1-20 and ongoing scrutiny. Specimen refusals come in two species.

Technical problems mean the specimen does not actually show the mark in use in commerce: it is just a picture or digital mock-up of the mark, a photocopy of the application drawing, the mark in the specimen does not match the drawing, the mark is missing or illegible, or—a classic mismatch—an invoice or advertisement is submitted for goods (for goods you generally need the mark on the product, its packaging, labels, or a point-of-sale display, not an ad), or a label is submitted for a service (where advertising and materials showing the service being rendered are the right proof). 37 C.F.R. § 2.56; TMEP §§ 904, 1301.04.

Substantive problems mean the mark, though it appears, is not functioning as a trademark: it is used merely as a trade name, as ornamentation (a big chest logo on a T-shirt, with no source-identifying use), as background decoration, to name a character, as the title of a single creative work, or solely as a domain name. These cut deeper because they go to whether the public perceives the term as a brand at all.

The usual cure is to submit a substitute specimen that was in use as of the relevant date, and if that is impossible, to argue against the refusal or amend the drawing. A word of caution lawyers internalize and founders often don't: the office-action response is a public record, so any concession you make about your mark can be quoted back at you in a later opposition, cancellation, or infringement fight. Choose your words with that future adversary in mind.

Final refusals and appeals

If your response does not overcome every issue, the examiner can issue a final Office Action. At that point your choices narrow: file a request for reconsideration with new arguments or evidence, or appeal to the Trademark Trial and Appeal Board. The TTAB is an administrative tribunal inside the USPTO, and an ex parte appeal is a paper proceeding (with optional oral argument) in which the Board reviews whether the refusal was correct. A further appeal runs either to the U.S. Court of Appeals for the Federal Circuit or, by civil action, to a federal district court under 15 U.S.C. § 1071. Most applications never reach this stage; many refusals are won at the response stage with good lawyering and a willingness to amend.

Stage Six: Publication and opposition (the TTAB enters)

Clear examination—either with no refusals or by overcoming all of them—and the examining attorney approves the mark for publication in the Trademark Official Gazette, the USPTO's weekly bulletin of marks headed toward registration. Publication is the trademark world's version of reading the banns: it gives the public formal notice and a window to object.

For 30 days after publication, any party who believes it would be damaged by registration may file a Notice of Opposition (or request an extension of time to oppose, up to 90 additional days for cause, and longer with the applicant's consent). 15 U.S.C. § 1063. An opposition is a contested inter partes proceeding before the TTAB, and it functions much like a streamlined federal lawsuit. The opposer must have a real interest and a reasonable basis (statutory standing under Lexmark Int'l, Inc. v. Static Control Components, Inc., 572 U.S. 118 (2014), applied to Board proceedings) and a valid ground—most commonly that the applicant's mark creates a likelihood of confusion with the opposer's earlier mark (Section 2(d) again, this time raised by a private party), or that the mark is generic, merely descriptive without secondary meaning, functional, abandoned, or that the applicant lacked a bona fide intent to use it.

Opposition proceedings have pleadings, a discovery phase, trial by submitted testimony and evidence, briefing, and a decision—but no live courtroom trial; the "trial" is conducted on paper and deposition transcripts under the Board's rules and the Trademark Trial and Appeal Board Manual of Procedure (TBMP). Discovery in particular is its own discipline, with interrogatories, document requests, and admissions governed by Board-specific rules and a mandatory discovery conference; we cover that machinery in detail in discovery practice in TTAB trademark proceedings. A related vehicle, cancellation, lets a party attack a mark after it has registered on similar grounds; opposition and cancellation are the TTAB's twin inter partes proceedings, both surveyed in trademark overview: infringement and related rights.

Two quieter tools deserve mention here. A Letter of Protest, formalized by the Trademark Modernization Act, lets a third party tip off the examining attorney about a problem with a pending application (for example, evidence of genericness or a conflicting prior mark) before publication—cheaper and faster than an opposition, though the protester gets no party rights. And after registration, the same Act created two new ex parte ways to clear deadwood, discussed in Stage Eight.

If no one opposes—the overwhelmingly common outcome for ordinary marks—the application proceeds to the next step automatically. If an opposition is filed and the applicant prevails (or settles, often via a coexistence agreement), the application likewise moves forward. If the opposer wins, registration is refused.

Worked example (hypothetical). Back to that AURORA energy-drink company. If it didn't catch Dahlia during examination, it gets a second bite when her mark publishes: it can file a Notice of Opposition within 30 days, arguing likelihood of confusion. Now Dahlia faces a choice—litigate before the TTAB, negotiate a coexistence agreement that limits each party to its lane (she stays in cold-brew concentrate sold in grocery refrigerator cases; they stay in energy drinks sold in convenience stores), or, if the conflict is real, rethink her brand. This is precisely why early clearance is worth its weight: the cost of discovering the energy-drink company at the opposition stage is vastly higher than discovering it during a pre-filing search.

Stage Seven: From allowance to registration

What happens after publication depends on your filing basis—and this is where the two roads (1(a) use and 1(b) intent-to-use) finally diverge in a visible way.

If you filed under Section 1(a) (or 44 or 66(a))—that is, your use is already established or you are on a foreign basis—and no opposition derails you, the USPTO simply registers the mark and issues a certificate of registration. You are done with the application phase. The mark is now on the Principal Register, you may use the ® symbol, and the maintenance clock (Stage Eight) starts ticking.

If you filed under Section 1(b) (intent-to-use), there is one more hurdle, because you reserved the mark before using it and now must prove you actually put it into commerce. After publication with no opposition, the USPTO issues a Notice of Allowance (NOA). Read that term carefully: a Notice of Allowance is not a registration. It is an official statement that your mark has cleared opposition and is allowed—conditioned on your proving use.

From the date of the NOA, you have six months to either:

  1. File a Statement of Use (SOU)—a verified statement that you are now using the mark in commerce on the listed goods, accompanied by a specimen and a fee; or
  2. File a Request for an Extension of Time to file the SOU (a six-month extension), with the ability to file additional extensions up to a statutory maximum of 36 months from the NOA. 15 U.S.C. § 1051(d).

(If you actually began using the mark between filing your 1(b) application and the NOA, you can instead file an Amendment to Allege Use earlier in the process, but only before the mark is approved for publication—after that, the SOU is the vehicle.) Miss both the SOU and an extension and the application goes abandoned; reviving it requires a petition within a tight window. When you do file the SOU, the examining attorney reviews it—and yes, the USPTO can issue an office action even at this late stage if the specimen is deficient or the use does not match the application. Clear that review and a registration issues, typically a couple of months later.

The intent-to-use track is a gift to founders—it lets you claim constructive use priority as of your filing date even though you actually used the mark months later (15 U.S.C. § 1057(c))—but it demands diligence. Those six-month deadlines arrive faster than anyone expects, and there is no sympathy for a missed Statement of Use.

Worked example (hypothetical). Dahlia filed 1(b) in January, sailed through examination by August, published in September, drew no opposition, and received her Notice of Allowance in November. She launches AURORA cold-brew in January of the following year, files her Statement of Use in February with a photo of a real retail shelf showing the bottled product for sale, the examiner accepts it, and her registration certificate arrives in April. Her priority date, however, reaches all the way back to the prior January's filing—a fourteen-month head start over anyone who started using a similar mark in the interim. That backdated constructive-use priority is the entire point of the ITU system.

Stage Eight: Maintaining the registration (it is not "set it and forget it")

A registration is a living thing that must be fed on a schedule, and more good marks are lost to missed maintenance deadlines than to any courtroom defeat. Federal registration is not a one-time purchase; it is a subscription you renew by proving continued use. With limited exceptions, every maintenance filing goes through TEAS. 37 C.F.R. § 2.23(a). The key post-registration filings are:

Section 8 Declaration of Continued Use. Between the fifth and sixth year after registration, the registrant must file a declaration—with a current specimen for each class and a fee—attesting that the mark is still in use in commerce on the registered goods, or that any nonuse is excusable. 15 U.S.C. § 1058; 37 C.F.R. § 2.160. Miss it and the registration is automatically cancelled—there is a six-month surcharged grace period, but after that your only remedy is to start the entire application over from scratch (your underlying common-law rights survive, but the federal registration and its presumptions are gone). (If your registration is a Madrid Protocol extension under Section 66(a), the equivalent continued-use filing is a Section 71 declaration, 15 U.S.C. § 1141k.)

Section 15 Declaration of Incontestability (optional but valuable). After five years of continuous use, the registrant may file a Section 15 declaration to render the registration incontestable. 15 U.S.C. § 1065. Incontestability is a powerful upgrade: it forecloses many challenges to the mark's validity, and the Supreme Court confirmed in Park 'N Fly that an incontestable registration cannot be cancelled merely on the ground that the mark is descriptive. Park 'N Fly, Inc. v. Dollar Park & Fly, Inc., 469 U.S. 189 (1985). It is not a force field—incontestable marks can still be attacked for genericness, abandonment, fraud, or functionality, among other listed grounds—but it removes the descriptiveness and ownership arguments from an infringer's arsenal. Practitioners usually file the Sections 8 and 15 declarations together as a combined filing at the five-to-six-year mark.

Section 9 Renewal (combined with Section 8). Between the ninth and tenth year, and every ten years thereafter, the registrant must file a combined Section 8 declaration of continued use and a Section 9 renewal application. 15 U.S.C. § 1059; 37 C.F.R. § 2.182. So the rhythm is: a check-in at year 5–6, then renewals at year 9–10, 19–20, 29–30, and so on, forever, as long as you keep using the mark.

That "forever" is the trademark's superpower and its catch. Unlike a patent (20 years) or a copyright (life plus 70 years), a trademark can last indefinitely—some American brands are well over a century old—but only if you keep using it and keep filing. Stop using the mark with intent not to resume, and it can be deemed abandoned; the Lanham Act presumes abandonment after three consecutive years of nonuse. 15 U.S.C. § 1127.

The new deadwood tools. The Trademark Modernization Act of 2020 (effective December 18, 2021) added two ex parte procedures that any party can use against someone else's registration that is not genuinely in use—a response to a flood of registrations (many of foreign origin) covering goods never actually sold here. Expungement under 15 U.S.C. § 1066a lets a petitioner ask the USPTO to cancel a registration on the ground that the mark was never used in commerce for some or all of the goods; it is available between the third and tenth year after registration. Reexamination under 15 U.S.C. § 1066b targets use-based registrations where the mark was not in use on the relevant date (the filing or amendment-to-allege-use date), and must be requested within the first five years. Both are cheaper and faster than a full cancellation before the TTAB, and they are now a standard arrow in the quiver when a cited registration is blocking your application and looks like a paper tiger.

The lesson for owners: calendar the deadlines, keep dated specimens of real-world use, and treat each maintenance window as seriously as the original application.

Stage Nine: Enforcement and policing (the USPTO won't do it for you)

Here is a sentence that surprises nearly every new registrant: the USPTO does not police your mark for you. It examines and registers; you enforce. While the office tries to keep confusingly similar marks off the register, it will not chase infringers, monitor the marketplace, or sue on your behalf. Trademark rights are a "use them or arguably weaken them" proposition—a brand owner who tolerates widespread infringement can find its mark diluted, its zone of protection shrunk, and (in extreme cases) its mark drifting toward genericide.

Effective policing is a program, not an afterthought. It typically includes watch services that monitor new USPTO filings and marketplace uses, periodic searches of domains and app stores, the new TMA Letter of Protest to head off conflicting applications before they register, and a graduated enforcement ladder. The first rung of that ladder is usually a cease-and-desist letter, which puts an infringer on notice and demands they stop. Drafting one well is an art—too soft and it is ignored, too aggressive and you risk a declaratory-judgment suit or a "trademark bully" reputation. We cover both sides in drafting a trademark cease-and-desist letter and responding to a trademark cease-and-desist letter.

When a letter does not resolve the matter, escalation paths include a TTAB opposition or cancellation (to keep a conflicting mark off the register or strike it down), a federal infringement suit under Section 32 (15 U.S.C. § 1114) for registered marks or Section 43(a) (15 U.S.C. § 1125(a)) for unregistered marks and trade dress, UDRP or ACPA domain disputes for cybersquatting, and recordation with U.S. Customs and Border Protection to intercept counterfeit imports at the border. The remedies in court can include injunctions, the infringer's profits, the plaintiff's damages (sometimes trebled), and, in exceptional cases, attorney's fees—and, thanks to the Trademark Modernization Act, a trademark plaintiff who proves infringement now enjoys a rebuttable presumption of irreparable harm when seeking an injunction (15 U.S.C. § 1116(a)), restoring a presumption that many courts had abandoned after eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006). For the broader online-protection strategy that ties policing together, see brand protection online, and for the full enforcement doctrine see trademark overview: infringement and related rights.

The throughline is that registration is the beginning of brand protection, not the end. A certificate in a drawer protects nothing; a registration backed by consistent use, vigilant monitoring, and a willingness to enforce is what actually defends a brand. And a defended brand is an asset you can license and assign—just remember that trademark licenses require genuine quality control and assignments must carry the associated goodwill, topics we treat in trademark overview: obtaining protection and licensing.

Putting the whole arc together

Step back and the trademark process is a coherent story with a logic to it. You choose a mark strong enough to own (the Abercrombie spectrum). You clear it so you are not building on someone else's land. You file electronically through TEAS under the right basis—1(a) if you are selling, 1(b) if you are reserving. You survive an examiner who may push back under Section 2(d) for confusion, Section 2(e) for descriptiveness, or over a defective specimen. You publish and weather any opposition before the TTAB. If you filed ITU, you prove your use with a Statement of Use after the Notice of Allowance. You register. And then you maintain (Sections 8, 15, and 9) and police, potentially for the rest of the company's life.

Each stage has a deadline that, if missed, sends you back to start—and several of those deadlines tightened recently, with the response window shrinking to three months and new ex parte challenges making thinly used registrations easier to dislodge. Each stage rewards work done at the previous stage—the founder who chose an arbitrary mark and cleared it thoroughly glides through examination and opposition; the founder who picked a descriptive name with no clearance gets a 2(e) refusal, a 2(d) refusal, a specimen refusal, and an opposition, often all at once. The trademark process is forgiving of patience and brutal toward shortcuts.

Key takeaways

  • Strength is chosen at the naming stage. The distinctiveness spectrum from Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir. 1976)—generic, descriptive, suggestive, arbitrary, fanciful—dictates everything downstream. Reach for suggestive, arbitrary, or fanciful.
  • Clearance is cheap insurance. Search the federal register, state filings, common-law users, and the open market before you commit, and keep the record—it can also defeat a later willfulness finding.
  • Your filing basis is a strategic choice. Section 1(a) for marks already in genuine use (preparation is not enough—Couture v. Playdom, 778 F.3d 1379 (Fed. Cir. 2015)); Section 1(b) intent-to-use to reserve constructive-use priority before launch (then prove use via a Statement of Use).
  • File electronically and watch the new clock. TEAS e-filing is mandatory (37 C.F.R. § 2.23(a)), and since December 2022 most office-action responses are due in three months.
  • Expect pushback. Section 2(d) likelihood-of-confusion, Section 2(e) descriptiveness, and specimen refusals are routine and often beatable—with amendments, arguments, evidence of acquired distinctiveness under Section 2(f), or the Supplemental Register.
  • Publication opens a 30-day opposition window before the TTAB; oppositions function like streamlined lawsuits.
  • Registration is not the finish line. File your Section 8 declaration at years 5–6, consider Section 15 incontestability, renew under Sections 8 and 9 at years 9–10 and every decade after, and never stop using the mark. The TMA's expungement and reexamination tools can clear deadwood blocking you.
  • You police your own mark. The USPTO won't do it for you; build a monitoring-and-enforcement program—the TMA's restored irreparable-harm presumption now helps when you go to court.

Frequently asked questions

How long does the whole trademark process take? For a clean use-based application with no refusals or oppositions, expect roughly a year from filing to registration, though USPTO backlogs can push that longer. An intent-to-use application takes longer because the mark cannot register until you prove use after the Notice of Allowance, and you have up to 36 months of extensions to get there. Refusals and oppositions add time on top.

What is the difference between ™ and ®? You may use the ™ symbol (or ℠ for services) on any mark you claim rights in, registered or not—it simply asserts a claim. The ® symbol may be used only after the mark is federally registered with the USPTO (on either the Principal or Supplemental Register). Using ® before registration is improper and can be held against you. For the basics of mark symbols and types, see trademark basics and our trademark FAQs.

Do I have to register to have any rights? No. In the United States, trademark rights arise from use in commerce, so an unregistered user can have enforceable common-law rights—usually limited to the geographic area of use. Federal registration adds powerful benefits: nationwide constructive priority, a legal presumption of validity and ownership, the ability to use ®, access to federal-court remedies, the foundation for incontestability, and the right to record with Customs. See trademark rights under common law for what use-based rights do and do not get you.

What is a Section 2(d) refusal, in plain English? It is the USPTO telling you that your mark is too close to a mark already on the register (or in an earlier pending application) for related goods or services, so consumers might be confused about who made the product. The examiner weighs the DuPont factors—mark similarity, relatedness of goods, trade channels, and more. You can often overcome it by distinguishing the marks or goods, showing the cited mark is weak, narrowing your goods, or obtaining a consent agreement.

My mark was refused as "merely descriptive." Is it dead? Not necessarily. You can argue the mark is suggestive rather than descriptive (does it require a mental leap?), or you can show acquired distinctiveness under Section 2(f)—often using the presumption that arises after five years of substantially exclusive use, plus advertising, sales, and recognition evidence. Alternatively, you can register on the Supplemental Register now and aim for the Principal Register once secondary meaning matures.

Why did my application get a "specimen refusal," and how do I fix it? Because the evidence you submitted didn't show the mark being used as a trademark in real commerce. Maybe it was a digital mock-up rather than an actual label, the mark didn't match your drawing, you submitted an advertisement for goods (the wrong type), or the mark looked like mere decoration or a trade name. The usual cure is to submit a substitute specimen that was in use as of the relevant date; failing that, argue against the refusal or amend the drawing. Whatever you say is public record, so phrase it carefully.

What is a Notice of Allowance, and why isn't it a registration? A Notice of Allowance is issued only in intent-to-use (Section 1(b)) cases. It means your mark cleared examination and opposition and is "allowed"—but because you reserved the mark before using it, you must still prove actual use by filing a Statement of Use (or extension requests) within six-month windows, up to 36 months total. Only after the USPTO accepts your proof of use does the registration issue.

What happens at the TTAB? The Trademark Trial and Appeal Board is the USPTO's in-house tribunal. It hears ex parte appeals (you challenging an examiner's refusal) and inter partes proceedings—oppositions (objecting to a mark before it registers) and cancellations (attacking a registered mark). These run like streamlined lawsuits with pleadings, discovery, and briefing, but the trial is on paper rather than in a live courtroom. See discovery practice in TTAB proceedings.

How do I keep my registration alive? File a Section 8 declaration of continued use between years 5 and 6 (commonly combined with a Section 15 incontestability declaration), then a combined Section 8/Section 9 renewal between years 9 and 10 and every ten years after that—each time with a current specimen showing real use. (Madrid-based registrations use a Section 71 declaration instead of Section 8.) Keep using the mark; three years of nonuse can create a presumption of abandonment.

Can I do this myself, or do I need a lawyer? Domestic applicants are permitted to file without counsel (37 C.F.R. § 11.14(e)), but the process is full of quiet traps—wrong owner, overbroad goods, the descriptiveness line, specimen rules, and unforgiving deadlines—and the unrepresented error rate is high. Foreign-domiciled applicants are required to use a U.S.-licensed attorney (37 C.F.R. § 2.11). Most businesses come out ahead with experienced counsel, especially when a refusal or opposition appears.

Related articles

This article is general information, not legal advice. Trademark law is fact-specific, deadlines are unforgiving, and outcomes vary by mark, market, and jurisdiction. Consult a qualified trademark attorney about your particular situation before acting.