Picture this. You spent a long weekend at a whiteboard with two co-founders and a pot of cold coffee, and at 11 p.m. on a Sunday somebody finally said the name out loud and the whole room went quiet in that good way. You bought the domain before you went to bed. Three months later you have a logo, a launch tee, a thousand stickers, a deck full of that name in 48-point type, and a small but real pile of customers who know you by it.

Then an email arrives. It is polite, it is on law-firm letterhead, and it informs you that a company you have never heard of has been using a very similar name on related goods since before you were born, that they own a federal registration for it, and that they would like you to stop. Immediately. Including the domain. Including the stickers.

This is the moment a lot of founders learn what a trademark is. It is a brutal way to learn, because by then the cheap, easy fixes are gone and only the expensive, painful ones remain. The entire purpose of this article is to get you to think about trademarks before that email, when your options are still cheap and your leverage is still high. The single most important variable in trademark protection is not how clever your mark is or how good your lawyer is. It is timing. So let us talk, in plain English and with real specifics, about when to trademark your brand.

We will cover why earlier is almost always better, the priority rules that make timing matter so much, the intent-to-use application that lets you plant a flag before you have a single sale, the clearance search that should come before you fall in love with a name, the genuinely scary cost of being forced to rebrand, what to file first as you grow, how international timing works (including a six-month window that is easy to blow), and finally the honest question of whether to do this yourself or hire someone. If you want the broader procedural walkthrough of how a trademark actually gets registered, pair this piece with the trademark process and the trademark registration guide; if you want the gentlest possible primer on what a trademark even is, start with trademark basics. This article assumes you have a brand or are about to launch one, and it is focused on the when.

First, the one-sentence version of "what is a trademark"

A trademark is a word, phrase, logo, sound, color, or other symbol that identifies the source of goods or services and distinguishes them from everyone else's. That last part is the whole game. Trademark law does not exist to give you a monopoly on a cool name for its own sake. It exists to protect consumers from confusion about who is selling them what, and to protect the goodwill you build up so that competitors cannot free-ride on it. The legal hook in the United States is the Lanham Act, codified at 15 U.S.C. § 1051 and following, plus a thick body of common law that predates the statute. The Lanham Act's definitions section, 15 U.S.C. § 1127, is where most of the load-bearing terms live — "trademark," "use in commerce," and "abandonment" among them — and we will return to it more than once, because the statute's definitions quietly decide who wins.

Because the system is about preventing confusion, the core question in almost every trademark fight is the same: is there a likelihood of confusion between two marks? Courts weigh a list of factors — the similarity of the marks, the relatedness of the goods or services, the strength of the senior mark, evidence of actual confusion, the marketing channels, and so on. (The federal circuits each have their own named version of the test; the Second Circuit's Polaroid factors, from Polaroid Corp. v. Polarad Electronics Corp., 287 F.2d 492 (2d Cir. 1961), and the Ninth Circuit's Sleekcraft factors, from AMF Inc. v. Sleekcraft Boats, 599 F.2d 341 (9th Cir. 1979), are the famous ones; the Federal Circuit and the USPTO use the DuPont factors.) You do not need to memorize the factors today. You just need to internalize the consequence: if your name is confusingly similar to someone else's, and they got there first, you have a problem, and the size of that problem grows with every dollar you spend on the name. For a deeper tour of the confusion analysis, see navigating the maze of trademark confusion.

That word "first" is doing enormous work, and it is the reason timing matters so much. So let us pull on it.

Why "first" is everything: the priority rules

In the United States, trademark rights are fundamentally rooted in use in commerce, not in registration. This is unusual. Many countries are "first to file" — whoever races to the trademark office first wins, full stop. The United States is, at its core, "first to use." The business that first uses a mark in commerce on particular goods or services generally has priority over later users, even if the later user is bigger, richer, and louder.

But — and this is the twist that makes timing strategic rather than purely chronological — federal registration grafts a powerful set of advantages onto that use-based foundation, and one of those advantages is a form of nationwide priority that you cannot get from use alone.

Let us separate the two layers.

Common-law rights: real, but local and fragile

The moment you start using a mark in commerce, you acquire common-law trademark rights. These are real and enforceable. You can sue, you can send a cease-and-desist letter, and you can stop confusingly similar uses. But common-law rights have two big limitations. First, they are geographically limited to the area where you actually use the mark and the zone of your natural expansion — not the whole country. Second, you bear the burden of proving your first-use date and your geographic footprint, which can be a documentary slog of invoices, dated advertisements, and shipping records. We cover this terrain in depth in trademark rights under common law, and the leading doctrine — the Tea Rose–Rectanus rule, under which a good-faith remote junior user can carve out its own territory — comes from United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90 (1918), building on Hanover Star Milling Co. v. Metcalf, 240 U.S. 403 (1916). The modern application of that doctrine, and how knowledge of a senior user destroys the good faith you would need to claim a remote territory, is the subject of Stone Creek v. Omnia, 875 F.3d 426 (9th Cir. 2017).

Here is the practical upshot of common-law rights for the timing question: they exist automatically and for free the second you use the mark, which sounds great, but they are a weak shield in a national or online market. A coffee roaster using "Acme Roast" only in Tucson has rights in Tucson, not in Tulsa or Toronto. In an e-commerce world where everyone is arguably selling everywhere, "local" rights are cold comfort — and courts have struggled to map the Tea Rose–Rectanus concept of "territory" onto a website that any consumer in the country can reach.

Federal registration: nationwide priority, frozen at filing

A federal registration on the Principal Register, by contrast, does something common-law rights cannot. Under 15 U.S.C. § 1057(c), filing a federal application gives you nationwide constructive use as of your filing date. In plain language: the law treats you as if you had used the mark everywhere in the country on the day you filed, even places you have never sold a thing. Anyone who starts using a confusingly similar mark after your filing date is a junior user nationwide, and your registration is constructive notice to the entire country under § 1072. The constructive-use date is the quiet hero of trademark strategy — it is why a thoughtful founder can beat a better-funded competitor who simply moved slower.

That is the magic, and it is why the date you file is one of the most important dates in your company's history. Federal registration also gives you: the right to sue in federal court and to invoke federal remedies; a presumption that your mark is valid, that you own it, and that you have the exclusive right to use it (§ 1057(b)); the ability to record the mark with U.S. Customs and Border Protection to block infringing imports; the eventual possibility of "incontestability" after five years of continuous use (§ 1065), which dramatically narrows the defenses an infringer can raise; the potential to recover enhanced remedies; and the right to use the ® symbol, which deters copycats and supports stronger damages arguments. (Before registration, you use ™ for goods or ℠ for services — those symbols simply assert a claim of rights and require no government permission.)

Stack those two layers together and the strategic picture snaps into focus. Use gives you rights but only locally. A federal filing date freezes nationwide priority in place. Therefore, the earlier you file, the larger and more secure your priority becomes — and the smaller the window in which someone else can sneak in ahead of you. As one widely used practitioner reference bluntly puts it, even in the United States, where priority technically flows from use, an owner should file as soon as practical because the filing date sets a nationwide constructive-use date. Which brings us to the single most underused tool in the founder's kit.

The intent-to-use application: stake your claim before your first sale

Here is the question that trips up nearly every first-time founder: "I haven't launched yet — I can't have a trademark, right? Don't I have to be using it first?"

For common-law rights, yes — no use, no rights. But for a federal application, no. The USPTO lets you file on one of several "bases," and the two that matter most for U.S. businesses are:

  • Section 1(a), "use in commerce" (15 U.S.C. § 1051(a)): you are already using the mark on your goods or services in interstate commerce, and you can prove it with a specimen (a real-world example of the mark as used). "Use in commerce" is a defined term of art under § 1127 — it means bona fide use in the ordinary course of trade, not a single token sale rigged to reserve a name.
  • Section 1(b), "intent to use" (15 U.S.C. § 1051(b)): you are not using the mark yet, but you have a bona fide intention to use it in the near future.

The intent-to-use ("ITU") application is the founder's secret weapon, and it is criminally underused. It lets you file today, before launch, before your first sale, before you have so much as a product to ship — and capture a nationwide priority date as of that filing. If your application eventually matures into a registration, your rights relate back to that early filing date under the constructive-use rule of § 1057(c). In effect, you can reserve a name.

How the two tracks actually run

It helps to see the two filing bases as two slightly different obstacle courses that converge at the same finish line. A use-based application is examined, published for opposition, and — if unopposed — proceeds straight to registration. An intent-to-use application is examined and published the same way, but instead of registering on approval, it receives a Notice of Allowance: a conditional green light that says, in effect, "This mark is registrable; now go use it." You then have six months to file a Statement of Use (a sworn statement, with a specimen, that you are now using the mark in commerce), and you can buy more time with extension requests — up to five of them, in six-month increments, for a maximum of three years from the Notice of Allowance. Once you file an acceptable Statement of Use, the registration issues, and your priority dates back to your original ITU filing.

Run the clock on a concrete sequence. You file an ITU application in March. Acme Corp., oblivious, launches a confusingly similar brand in June, spends a fortune on it through the fall, and files its own application in October. Acme still loses — because your constructive-use priority date is March, before Acme started, and your eventual registration relates back to it. Acme's months of spending bought it a junior position and, very likely, a § 2(d) refusal of its own application. The whole expensive contest was decided in March, by a filing fee.

A few guardrails on the ITU, because the doctrine has teeth:

"Bona fide intention" is a real requirement, not a magic word. The intention must be genuine and demonstrable by objective evidence — a business plan, product development, branding work, supplier conversations, regulatory steps, that kind of thing. The Trademark Trial and Appeal Board has invalidated applications where the applicant had no documentary evidence of any real intent to use the mark, treating the filing as a naked attempt to warehouse a name. The leading case is M.Z. Berger & Co. v. Swatch AG, 787 F.3d 1368 (Fed. Cir. 2015), which confirmed that the lack of a bona fide intent to use is a valid statutory ground to oppose or cancel an application and that the question turns on objective circumstances, not the applicant's say-so. Translation: do not file ITU applications for fifty names you might use someday. File for the one (or few) you actually plan to launch, and keep a paper trail showing you mean it.

You cannot just sit on an ITU forever. Three years from the Notice of Allowance is the hard ceiling. If you never actually use the mark, the application goes abandoned and your priority evaporates. The relation-back benefit only pays off if you eventually cross the finish line with real use.

You still have to clear the same hurdles. An ITU application is examined for likelihood of confusion, descriptiveness, and everything else, exactly like a use-based one. The basis affects when you prove use, not whether your mark is registrable. An ITU filing buys you a priority date; it does not buy you immunity from a senior user who got there first.

The takeaway for the timing question is enormous: you do not have to choose between "launch now and risk it" and "wait until I'm sure." You can file an intent-to-use application the moment you have settled on a name and have genuine plans to use it. That is, frequently, the single most valuable trademark step a pre-launch company can take, and it costs a small fraction of what a rebrand costs. We will get to those numbers shortly.

But before you file anything: clear the name

Filing is step two. Step one is making sure the name is actually available — that you are not about to pour money into a name that belongs, legally, to someone else. This is the clearance search, and skipping it is the most common and most expensive mistake founders make.

The logic is simple. Trademark priority generally runs to the first user. If a senior user already owns rights in a confusingly similar mark for related goods, then no amount of filing, marketing, or wishing changes the fact that you are the junior user. You could file an ITU application, sail through to a Notice of Allowance, and still get blocked — because the senior user can oppose your application at the TTAB, petition to cancel after registration, sue you for infringement, or simply send the letter that started this article. Your filing date helps you against people who came after you; it does nothing against people who came before.

So before you commit, you search. A good clearance search is layered:

  • Knockout search. A quick pass through the USPTO's public database (the modern Trademark Search system that replaced TESS) plus basic web and state searches to catch the obvious, fatal conflicts. This is the "is anyone clearly already doing this?" filter. It is cheap and you can do a basic version yourself.
  • Comprehensive (full) search. A professional search that sweeps federal registrations and applications, state registrations, common-law uses (business names, domains, product names that were never registered but still create rights), and close variants — phonetic equivalents, foreign-language translations under the doctrine of foreign equivalents, alternate spellings, and design elements for logos. Confusion is not about identical spelling; "Kwik" and "Quick" can collide, and so can "Cyclone" and the Spanish "Ciclón."
  • Legal opinion. An attorney reviews the search results and tells you whether the name is clear, risky, or dead, and why. This is where the layered facts turn into an actual risk judgment — distinguishing the conflicts that are fatal from the ones that are survivable or addressable through a coexistence strategy.

We have a dedicated, detailed walkthrough in how to conduct a comprehensive trademark clearance search. For now, internalize the sequencing: search, then commit, then file. Founders love to do it in the wrong order — fall in love with a name, build a brand around it, then discover the conflict. The cost of reversing course at that stage is exactly what we will quantify in a moment. There is also a defensive bonus to a documented clearance, beyond avoiding the conflict in the first place: a good-faith reliance on a competent search and an attorney opinion can help rebut a later claim that any infringement was willful, which is the kind of finding that triggers enhanced damages and fee awards. We unpack that protective effect in the shield of good faith.

One more clearance point that matters for timing: clearance is not only a one-time event at launch. It is wise to re-clear when you add a major new product line, expand into a new industry (and therefore new trademark classes), or go international, because confusion is assessed within the field of the goods and services. A name that is clear for software might be occupied for apparel.

Choose a name that is actually capable of being protected

Timing intersects with a second decision that founders often make on vibes alone: the choice of mark itself. The strength of a mark — how much protection it gets and how easily it registers — depends on where it falls on what lawyers call the distinctiveness spectrum, articulated in the canonical case Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir. 1976). From weakest to strongest:

  1. Generic terms ("Email" for email; "Bicycle" for bicycles) are never protectable as trademarks. They are the name of the thing itself, and no one gets to own the dictionary. A mark can even slide into genericness over time — "aspirin," "escalator," and "thermos" all started as brands and died as words, the cautionary tale every brand owner is taught.
  2. Descriptive terms ("Creamy" for yogurt; "Cold and Creamy" for ice cream) merely describe a feature, quality, or characteristic. They are not protectable unless and until they acquire secondary meaning — proof that consumers have come to associate the term with a single source. Acquiring secondary meaning typically takes years of use and substantial marketing, which is exactly the kind of delay a startup cannot afford. Descriptive marks initially land on the Supplemental Register rather than the Principal Register and miss out on the best benefits, including the § 1057(b) presumptions and the path to incontestability.
  3. Suggestive marks ("Netflix," "Coppertone") hint at a quality but require a mental leap. They are inherently distinctive and protectable from day one — the sweet spot of "evocative but ownable."
  4. Arbitrary marks (real words used in an unrelated context — "Apple" for computers, "Amazon" for retail) are strong and inherently distinctive.
  5. Fanciful marks (invented words — "Kodak," "Exxon," "Xerox") are the strongest of all, because they mean nothing except your brand.

Here is why this belongs in a timing article: the cheapest, fastest, safest path to a registrable mark is to pick a suggestive, arbitrary, or fanciful name from the start. Descriptive names feel intuitive and "explain themselves," which is why founders gravitate to them — and they are a trap. They are hard to register, hard to enforce, and force you to spend years building secondary meaning you might never get. The earlier you understand the spectrum, the earlier you can choose a name that the law will actually protect, which saves you from discovering at the worst possible moment that your beloved descriptive name was never ownable in the first place. (One special case worth flagging: marks that are "primarily merely a surname" face a similar acquired-distinctiveness hurdle under § 2(e)(4) — relevant if you want to brand under your own last name; see trademarking your own name.)

So the front-of-funnel checklist looks like this: pick a distinctive name, then clear it, then file (ITU if you are pre-launch, use-based if you are already selling). Do those three things in that order and you have done 90% of what good trademark timing requires.

The cost of waiting: a tour of the things that go wrong

Let us get concrete about what "waiting" actually costs, because abstractions ("you should protect your brand early") never move anyone. Here are the failure modes, with the price tags attached.

Failure mode 1: someone files ahead of you and freezes you out

You launch "Acme Roast" coffee in Tucson and grow happily for two years on common-law rights you never bothered to register. Meanwhile, a company in Denver picks the same name, files a federal ITU application a year after your launch, and registers. Now things get ugly. The Denver company has a federal registration and nationwide constructive priority as of its filing date. You have common-law rights, but only in Tucson and your natural zone of expansion. You can probably keep operating in Tucson (your prior use protects you there, and you may be able to seek a concurrent-use registration limited to your territory), but your dreams of selling nationwide, getting into Whole Foods, or expanding to Denver are dead, because the federal registrant blocks you everywhere else. You cannot get your own unrestricted federal registration for the name. Your brand is permanently boxed in — by a company that started after you, simply because they filed and you did not. A few hundred dollars and an afternoon of filing, two years earlier, would have flipped that entire outcome.

Failure mode 2: a senior user surfaces and you have to rebrand

This is the email from the opening. You build a brand, and a genuine senior user — someone with rights predating yours — appears and demands you stop. Now you are the junior user, and the law is not on your side. Your realistic options are: negotiate a coexistence agreement (possible if the goods or geography are different enough), pay them for a license or assignment (expensive, and they have no obligation to sell), or rebrand. If you have to rebrand, here is the partial bill:

  • New name development, logo, and brand identity.
  • New website, domain acquisition, and SEO — and you forfeit years of accumulated search ranking and backlinks under the old name.
  • New packaging, labels, signage, uniforms, business cards, and any physical inventory bearing the old mark (often written off entirely).
  • New marketing collateral, ad creative, and the cost of re-running awareness campaigns to teach the market your new name.
  • Legal fees for the dispute itself, plus possible damages, a settlement payment, or — if the senior user proves willfulness — disgorgement of your profits.
  • The intangible but very real loss of brand equity and customer recognition you spent years building.

For a small business, a forced rebrand routinely runs from the low five figures into six figures; for a funded startup mid-flight, it can be catastrophic, both financially and in terms of momentum and morale. And the cruelest part is the timing asymmetry: the longer you waited before discovering the problem, the more you have invested in the doomed name, so the more the rebrand costs. Waiting does not just risk the problem — it inflates the size of the problem every single day.

Failure mode 3: you lose your foreign rights to a squatter

In much of the world, it is first-to-file, not first-to-use. If you wait to think about international protection, a trademark squatter can register your brand in a key market — China is the textbook example — and then either block you from entering that market or demand a ransom to sell you back your own name. We will get to the timing fix (the six-month Paris priority window) below, but file this under "things waiting can cost you." It is among the most expensive mistakes in the whole field precisely because the cure — buying back, litigating bad faith, or abandoning a market — is so much pricier than the few hundred dollars a defensive foreign filing would have cost.

Failure mode 4: you weaken your own position by sleeping on it

Trademark rights can be lost through neglect. Under 15 U.S.C. § 1127, a mark is deemed abandoned if its use is discontinued with intent not to resume (and three consecutive years of nonuse is prima facie evidence of that intent), or if the owner's own conduct causes the mark to lose its significance as a source identifier — through uncontrolled "naked" licensing, assignment in gross, or simply letting the term go generic. Separately, if you fail to police your mark and let others use confusingly similar marks unchecked, your mark weakens, and equitable defenses such as laches (unreasonable delay in enforcing your rights, to a defendant's prejudice) can bar or limit your claims against an infringer who relied on your silence. We treat those doctrines in understanding equitable defenses. The theme is consistent: trademark law rewards diligence and punishes delay, at the front end and the back end alike.

Put the four failure modes together and the lesson is stark. The downside of filing too early is small — a modest fee, and the discipline of clearing your name first. The downside of filing too late ranges from "boxed into one city forever" to "six-figure emergency rebrand" to "locked out of China." This is one of the most lopsided risk-reward calculations in all of business law. That asymmetry is the entire argument for acting early.

So when, exactly? A practical timeline for founders

"Earlier is better" is true but not actionable on its own. Here is a concrete sequence keyed to the stages of a typical company.

Stage 0 — Naming (before you commit to anything)

This is the cheapest and highest-leverage moment. Do a knockout search on every finalist name before you fall in love. Steer toward suggestive, arbitrary, or fanciful marks and away from descriptive and generic ones. Check that the matching domain and key social handles are available, but never treat domain availability as a proxy for trademark availability — they are completely different systems, and a free domain tells you nothing about whether you can own the name. (The converse trap is just as common: a registered trademark does not entitle you to a matching domain that someone else lawfully holds.) If a name survives the knockout search, you have a candidate worth a real clearance.

Stage 1 — Name chosen, pre-launch (you have a winner but no sales yet)

Run a comprehensive clearance search and get a legal opinion. If the name clears, file a federal intent-to-use application for your core mark in your core class. This freezes your nationwide priority date before you have spent a dime on marketing. This is, for most startups, the optimal moment to file — early enough to lock in priority, deliberate enough that you have cleared the name and know it is worth protecting. If you are also raising money, note that investors increasingly diligence IP; a cleared name and a pending application are exactly the kind of tidy, low-risk story that makes diligence go smoothly, and an unresolved naming conflict is exactly the kind of thing that spooks a lead investor at the worst possible moment. See navigating the capital-raising maze for how IP fits into fundraising.

Stage 2 — Launch (you are now using the mark in commerce)

Begin using ™ (or ℠ for services) immediately — no registration required. If you filed ITU at Stage 1, you will convert it to a registration by filing your Statement of Use once you are genuinely selling. If you skipped Stage 1 and are launching now, file a use-based (Section 1(a)) application as soon as you have a valid specimen. Either way, the goal is to convert real-world use into a federal registration as fast as the process allows. The procedural mechanics live in how to file a trademark application with the USPTO, and the document-by-document requirements in the trademark registration guide.

Stage 3 — Growth (new products, new markets, new classes)

As you add product lines, you add trademark classes — registrations are tied to specific goods and services under the international Nice classification, and your apparel registration does not automatically cover your new software product. Re-clear and file for important new marks (product names, taglines, logos) as they emerge. Start policing: monitor for infringers and confusingly similar new applications, and act on them, because the right to complain erodes if you ignore problems. A useful discipline here is to define a "circle of protection" around your marks — clear, written criteria for which third-party uses fall inside the circle and warrant a challenge and which fall outside it and can be left alone. Consistent criteria keep you from over-enforcing (and looking like a trademark bully) or under-enforcing (and weakening your mark), and they keep you from arguing two marks are confusingly similar in one matter only to have that argument thrown back at you in the next. A brand-protection program at this stage is described in brand protection online.

Stage 4 — Maturity (keep it alive)

A federal registration is not "set it and forget it." You must file a Declaration of Use under Section 8 between the fifth and sixth years (and a Section 15 declaration of incontestability is worth filing at the same time), and renewals under Section 9 every ten years. Periodic IP audits — confirming each registration still matches what you actually sell, in the classes that matter, in the countries where you operate — catch problems while they are cheap. Miss a maintenance deadline and your registration is canceled, and you may have to start over, having lost the priority date you worked so hard to lock in. The full lifecycle, including maintenance, is mapped in the trademark process.

If you remember only one thing from this section, make it this: the optimal moment to file for most founders is Stage 1 — after you have cleared the name and before you have spent money marketing it — using an intent-to-use application. Everything else is execution.

"What should I file first?" Sequencing on a budget

Budgets are real, and most startups cannot federally register every word, logo, tagline, and product name on day one. So prioritize. The rough order of importance:

  1. Your primary brand name (the word mark). Word marks are the most flexible and valuable form of registration because they protect the name itself regardless of font, color, or styling. If you can only afford one filing, make it the standard-character word mark for your house brand in your core class.
  2. Your logo, especially if it is distinctive and you expect to invest in it heavily. A logo (a "design mark" or "composite mark") is protected separately from the word mark. Many companies file the word mark first and add the logo as the brand matures or once the logo design is stable — because if you redesign the logo, the old registration's value drops and you may need to refile.
  3. Flagship product names and major taglines, in order of how central they are to the business and how much you are spending to promote them.
  4. Secondary classes. If your single product genuinely spans multiple classes (say, you sell both physical goods and a software service under the same name), cover the classes that matter most first. For a primer on how the class system works, see USPTO trademark classes.

A note on word marks versus logos that founders consistently get wrong: registering a stylized logo does not fully protect the underlying words, and registering the plain word mark does not protect a specific logo design. They are different rights. For a name you intend to keep, the standard-character word mark is usually the priority, because it travels with the name through every redesign.

Also resist the urge to file in classes you are not actually using or genuinely planning to use. U.S. trademark law requires bona fide use or intent in every class you claim. Padding your application with aspirational classes can render the registration vulnerable to cancellation for fraud or lack of bona fide intent — a real risk highlighted by the M.Z. Berger line of cases mentioned earlier, and one the USPTO has been actively policing as part of its effort to clean up an over-cluttered register. File for what you are doing and credibly about to do, not for every category you can imagine.

International timing: the six-month window you cannot afford to miss

If you have any ambition beyond your home country — selling abroad, manufacturing overseas, or simply protecting against squatters in markets where copycats thrive — international timing deserves its own slot in your plan, because there is a deadline most founders have never heard of.

Paris Convention priority: a six-month head start, if you use it

The Paris Convention for the Protection of Industrial Property gives you a powerful timing tool: a right of priority. Under Article 4 of the Convention, once you file a trademark application in one member country (say, the United States), you have six months to file corresponding applications in other member countries for the same mark and the same goods and services, and claim the priority date of your original filing. (Six months is the trademark window; for patents, the Paris priority period is twelve months — do not mix them up.)

In plain terms: file in the U.S. on March 1, and if you file in the EU, the UK, Japan, or any other Paris member by September 1, those foreign applications are treated as though you filed them back on March 1. That relation-back can be decisive in a first-to-file country, where a squatter who filed in, say, May would otherwise beat you. The catch is the deadline. Miss the six months and the priority claim is gone — you can still file abroad, but you lose the head start, and in fast-moving markets that head start is sometimes the whole ballgame.

So if international protection is even plausibly in your future, your U.S. filing date does double duty: it locks in domestic priority and it starts the six-month Paris clock for the rest of the world. That is yet another reason not to dawdle on the initial U.S. filing.

The Madrid Protocol: one application, many countries

For actually filing abroad, the Madrid Protocol (administered by WIPO) lets you file a single "international application," based on your home-country application or registration, and designate any number of member countries — currently well over a hundred. It is administratively efficient and cost-effective compared to hiring local counsel to file separately in each country, and you can add countries later by "subsequent designation." Two timing wrinkles to know:

  • A Madrid international registration depends on your "basic" home application for the first five years. If your U.S. application is refused or canceled in that window, the international registration can fall with it (the dreaded "central attack," though the Protocol allows you to "transform" the lost designations into national applications). This argues for making sure your home application is solid before leaning hard on it for Madrid.
  • Madrid and the Paris priority claim work together: you can file your Madrid international application within the six-month Paris window and claim priority back to your original U.S. filing date across all designated countries at once.

The practical sequencing for an internationally minded startup, then: file in the U.S. first (ITU or use-based), and if global expansion is realistic, file your Madrid application (or direct national applications) within six months to capture Paris priority everywhere at once.

When Madrid is not enough: China and bad-faith squatting

If a specific market is mission-critical, consider filing there directly and early rather than relying solely on Madrid — and China is the recurring example, for two reasons. First, China is rigorously first-to-file, so the party that files first generally wins regardless of who used the mark earlier or elsewhere. Second, China has long had an active cottage industry of squatters who monitor foreign brands and rush to register them before the foreign owner arrives, then either block market entry or sell the registration back at a markup. China's trademark regime has been amended to push back on bad-faith filings — the law now bars applications made without an intent to use and gives examiners and the courts more room to reject or cancel bad-faith registrations — but enforcement is uneven and the cleanest defense remains the oldest one: file first, before you announce the product, before your manufacturer's employees see the name, and before a watcher spots your brand. For a market like China, the entire contest is often decided by who reaches the registry first, which means the right time to file is "before anyone in-country has a reason to."

DIY versus hiring an attorney: an honest accounting

Founders are scrappy, and the USPTO's system is designed to be usable by non-lawyers, so the temptation to self-file is strong. Here is a candid view of where DIY works and where it bites.

Where DIY is reasonable. Running an initial knockout search yourself is fine and even advisable — it is free and it kills obviously doomed names early. Educating yourself (you are doing it now) is invaluable. And for a clearly distinctive, fanciful name in a single class with no apparent conflicts, a careful founder can file a straightforward application; domestic applicants are permitted to file pro se under 37 C.F.R. § 11.14(e).

Where DIY tends to go wrong. The expensive mistakes cluster in a few predictable places:

  • Identifying the goods and services. The description of goods/services and the choice of classes is more art than it looks. Too narrow and your protection has holes; too broad and you invite refusals or later cancellation. Picking the wrong class, or copying a competitor's overbroad identification, is a classic self-filer error.
  • Specimens. The USPTO is strict and increasingly skeptical about specimens — the real-world proof of use. Mock-ups, digitally altered images, and "merely ornamental" uses (a slogan splashed across the front of a T-shirt, for instance) get refused, and a bad specimen can sink an otherwise fine application.
  • Reading the clearance results. Anyone can pull up a list of similar marks. Knowing which ones are actually fatal, which are survivable, and which call for a coexistence strategy is a judgment call that comes from experience with the likelihood-of-confusion factors.
  • Office Actions. If the examining attorney issues a refusal — most commonly a § 2(d) likelihood-of-confusion refusal or a § 2(e) descriptiveness refusal — responding effectively requires legal argument, sometimes with evidence and case citations. This is where self-filers most often stall out and lose their filing fee.

There is also a structural reason to involve counsel that has nothing to do with skill: the USPTO now requires foreign-domiciled applicants to be represented by a U.S.-licensed attorney. And as a practical matter, an experienced trademark attorney's clearance opinion is not just a filing aid — it is part of your defense, for the willfulness reasons discussed above and detailed in the shield of good faith.

The honest bottom line: for a simple, clean, distinctive single-class mark, a diligent founder can self-file and often does fine. The moment there is any complexity — a borderline-descriptive name, possible conflicts in the search, multiple classes, international ambitions, or an Office Action — the cost of competent help is trivial next to the cost of getting it wrong. And remember the asymmetry: a few hundred to a couple thousand dollars in legal fees up front is rounding error compared to a six-figure forced rebrand later.

A worked example, start to finish

Let us run our hypothetical founders, the team behind a pre-launch productivity app, through the whole sequence so the timeline feels real. (This example is illustrative and hypothetical; names and dates are invented.) Call the company Lumen Labs and the product Tideline — an invented, suggestive name that hints at organizing the ebb and flow of your tasks without describing the software, which puts it comfortably in protectable territory.

Stage 0. At the whiteboard, before anyone buys a domain, a co-founder runs a knockout search of "Tideline" in the USPTO database and a few web searches. There is a clothing brand using "Tideline," but apparel and productivity software are unrelated, so confusion is unlikely — promising, but not conclusive. The name survives the knockout and becomes the leading candidate.

Stage 1. With a winner in hand and no sales yet, Lumen Labs commissions a comprehensive clearance search. It surfaces the apparel mark (different class, low risk), a defunct application that went abandoned (no longer a barrier), and no live conflicts in software. Counsel green-lights the name and documents the opinion. Lumen Labs files a federal intent-to-use application for the standard-character word mark "TIDELINE" in the relevant software class. The filing date — say, April 12, 2026 — is now their nationwide constructive-use priority date. They also note that the EU is a likely future market and calendar a six-month Paris deadline of October 12, 2026.

Stage 2. Tideline launches in August. The team starts using ™ on the app and website on day one. Because real customers are now using the product, Lumen Labs files its Statement of Use with a clean specimen — a screenshot of the live app showing the mark as it appears to users — and the registration issues, relating back to the April priority date. They also file a Madrid application in early October, within the Paris window, designating the EU and the UK, capturing the April priority date abroad.

Stage 3. A year later, Lumen Labs adds a hardware accessory and a paid analytics service. That is new classes. They re-clear and file additional applications. They also set up a trademark watch; when a copycat files for "Tydeline" for a competing app, Lumen Labs is notified and opposes it at the TTAB, winning easily because it holds a registration with an April 2026 priority date that predates the copycat by more than a year.

Stage 4. Five years in, Lumen Labs files its Section 8 declaration of continued use and a Section 15 declaration, making the registration incontestable and dramatically narrowing the defenses any future infringer can raise. The brand they named at a whiteboard is now a defensible, appreciating asset.

Notice what made the whole thing work: not brilliance, not luck, but order and timing. Clear, then commit, then file early on an ITU basis, then convert, then watch the Paris clock, then maintain. Every painful failure mode from earlier in this article — the freeze-out, the forced rebrand, the foreign squatter — was avoided not by winning a fight but by never having to have one.

Key takeaways

  • Timing is the most important variable. The downside of filing too early is tiny; the downside of filing too late ranges from being geographically boxed in to a six-figure emergency rebrand. That asymmetry is the whole argument for acting early.
  • U.S. rights are use-based, but federal registration adds nationwide priority frozen at your filing date through the constructive-use rule of 15 U.S.C. § 1057(c). The earlier you file, the bigger and safer your priority.
  • The intent-to-use application (§ 1(b)) lets you stake a nationwide claim before your first sale — provided you have a genuine, documentable intent to use the mark (M.Z. Berger). It is the founder's most underused tool.
  • Clear before you commit. Trademark priority runs to the first user, so a clearance search and legal opinion before you build a brand is the cheapest insurance you will ever buy — and it doubles as a willfulness defense if a dispute ever comes.
  • Pick a protectable name. Suggestive, arbitrary, and fanciful marks are strong and registrable from day one; descriptive and generic names are traps (Abercrombie).
  • Sequence your filings. Standard-character word mark for your house brand first, then logo, then flagship products and key classes. File for what you use and genuinely intend to use — no aspirational padding.
  • Mind the international clock. A U.S. filing starts a six-month Paris Convention priority window for foreign filings; the Madrid Protocol lets you reach many countries with one application; in first-to-file markets like China, file early and directly, before you announce.
  • Maintain and police what you register. Section 8 between years five and six, Section 9 renewals every ten years, active monitoring, and a consistent enforcement "circle" — or you lose it to abandonment, genericide, or laches.
  • Get help when it gets complicated. DIY is fine for clean, distinctive, single-class marks; the moment there is a conflict, a descriptiveness issue, multiple classes, an Office Action, or an international angle, competent counsel pays for itself many times over.

Frequently asked questions

Do I have to be selling something before I can trademark my brand? No. You can file a federal intent-to-use application (Section 1(b)) before you make a single sale, as long as you have a bona fide intention to use the mark in the near future. You will eventually need to prove actual use (via a Statement of Use) to complete the registration, but the filing — and the nationwide constructive-use date that comes with it — can happen pre-launch.

Is my business name registration or my LLC filing the same as a trademark? No, and this trips up a lot of founders. Registering an LLC or a "doing business as" name with your state, or buying a domain, does not give you trademark rights. Those are separate systems. You can hold a state business registration and still infringe someone else's trademark — or have your own name taken by a federal registrant. Trademark rights come from use in commerce and are strengthened by federal registration at the USPTO.

How long does federal trademark registration take, and what does it cost? Expect roughly a year or more from filing to registration in a smooth case, and longer if there is an Office Action, an opposition, or a USPTO backlog. Government filing fees are charged per class and are modest (in the low hundreds of dollars per class), but the USPTO periodically restructures its fee schedule, so check current rates before you file. Attorney fees vary; a clean single-class application with a clearance opinion is typically a few hundred to a couple thousand dollars. All of it is trivial next to the cost of a forced rebrand.

What is the difference between ™ and ®? You may use ™ (for goods) or ℠ (for services) at any time to assert a claim of trademark rights — no registration or permission required. The ® symbol may be used only after your mark is federally registered. Using ® before registration is improper and can cause problems, so stick with ™/℠ until your registration issues.

Can a competitor really take my brand just by filing first? In the U.S., not exactly — our system is fundamentally first-to-use, so genuine prior use gives you defensible rights. But a competitor who files a federal application before you can secure nationwide priority that boxes your unregistered, locally used brand into the region where you can prove prior use, and can block you from registering or expanding nationally. Filing early — ideally an ITU application before launch — is how you avoid handing them that opening. In many foreign countries, which are first-to-file, the answer is closer to "yes," which is why international timing matters.

I already have a brand I've been using for years without registering. Is it too late? Not at all. You can still run a clearance search and file for federal registration, and your years of prior use count toward priority and may support common-law rights in your area. The sooner the better, though — every day you wait is another day a competitor could file ahead of you for the parts of the country where you have not yet established use. Do the clearance and file.

Should I trademark my name, my logo, or both? If budget forces a choice, start with the standard-character word mark for your primary brand name, because it protects the name itself across every font and styling and survives logo redesigns. Add the logo as a separate filing when the design is stable and you are investing in it. The two are different rights, and registering one does not automatically protect the other.

When do I need to think about international trademarks? The moment international expansion is realistically on your roadmap — or the moment you start manufacturing in or selling to a first-to-file country prone to squatting (China being the classic example). Your U.S. filing date starts a six-month Paris Convention priority window; file your foreign applications (often via the Madrid Protocol) within that window to relate them back to your U.S. date. Miss the six months and you lose the head start, though you can still file later.

Can I lose a trademark I already registered? Yes. Under 15 U.S.C. § 1127, a mark can be deemed abandoned through nonuse with intent not to resume (three years of nonuse is prima facie evidence), or through the owner's own conduct — uncontrolled licensing, assignment in gross, or letting the mark become the generic name for the product. You can also let your registration lapse by missing the Section 8 or Section 9 maintenance deadlines, and you can compromise your ability to enforce by sleeping on infringement (laches). Registration is the start of stewardship, not the end of it.

Related articles

This article provides general information and is not legal advice. Trademark rights, timing, and procedures depend on your specific facts and on law that changes over time; for advice about your brand, consult qualified trademark counsel.