Startup and Founder Brand Toolkit: The First Two Years of Trademark Decisions

By Casey Scott McKay · 2024-07-29

This toolkit sequences every trademark decision a new company makes between the naming workshop and the Series A data room, month by month, and routes each one to the Marksy document that does the work. It explains why protectability and availability are separate questions answered by different tools, why a Section 1(b) intent-to-use filing is the single highest-return move most founders can make, and how the identification of goods a founder writes in month one fixes the company's enforcement radius for a decade. It carries one worked matter across twenty-four months — a two-founder medical-device and app company clearing, filing, prosecuting, and proving use on two marks in four classes — and prices the whole program in a budget table built from the fee schedule effective 18 January 2025. Separate sections cover examination and office-action triage, the six-month statement-of-use clock and the thirty-six-month outer limit, the ownership file that decides diligence, the regulated-industry detour, and the watching and enforcement work that starts the day a mark publishes. It closes with the five mistakes that cost the most to unwind — the founder's name on the application, the intent-to-use application assigned too early, clearance that stopped at the free database, the padded identification, and the logo the designer still owns — with a cost-to-avoid and cost-to-unwind figure for each. A branching reading path, a table of controlling authorities, and annotated routes to more than forty Marksy documents follow.

IP and Technology > Trademarks | Toolkit | Published 29 July 2024 - Updated 13 April 2026 | Casey Scott McKay - marksy.us

Summary. This toolkit sequences every trademark decision a new company makes between the naming workshop and the Series A data room, month by month, and routes each one to the Marksy document that does the work. It explains why protectability and availability are separate questions answered by different tools, why a Section 1(b) intent-to-use filing is the highest-return move most founders can make, and how the identification of goods written in month one fixes a company's enforcement radius for a decade. One worked matter runs across all twenty-four months, and the whole program is priced in a budget table built from the fee schedule effective 18 January 2025. Later sections cover examination and office-action triage, the six-month statement-of-use clock and the thirty-six-month outer limit, the ownership file that decides diligence, the regulated-industry detour, and the enforcement work that starts the day a mark publishes. It closes with the five mistakes that cost the most to unwind, each with a cost to avoid and a cost to fix.

Keywords: startup trademark strategy · founder brand · naming workshop · trademark clearance · intent-to-use application · constructive use priority · statement of use · notice of allowance · identification of goods and services · nice classification · trademark budget · chain of title · void ab initio application · section 1060 assignment restriction · series a ip diligence · contractor ip assignment · office action response · trademark watch service · rebranding cost · first two years

Start Here

A company commits most of what it will ever spend on trademarks in its first twenty-four months, and spends almost none of it then. The name is chosen in a conference room by people who have never read 15 U.S.C. § 1052(e)(1). The logo is bought from a designer on a one-page invoice. The application is filed by whoever has the credit card. Two years later a Series A lawyer asks four questions about that file, and the answers cost between nothing and four hundred thousand dollars depending on what happened in those first few months.

This toolkit is the curated route through everything Marksy publishes about that window. It is for founders and early general counsel building a brand file from zero, for outside counsel who just took on a seed-stage client and needs a defensible sequence, and for the operations person who inherited a folder named "Trademark" containing one screenshot.

It answers three questions:

What has to happen, and in what month? The order is not arbitrary. Doing clearance after the packaging is printed is not clearance; it is damage assessment.
What does the program actually cost? Government fees and professional fees, itemized, for a real two-mark, four-class program.
Which decisions are cheap now and ruinous later? Five of them, priced both ways.

If you read only one thing, read Trademark Clearance Searching: What a Knockout Search Can and Cannot Tell You. It is the single document in this collection with the highest ratio of pages read to dollars saved. It explains the structural asymmetry that founders get wrong — a free database screen produces reliable no answers and unreliable yes answers — and it walks the constructive-notice rule of 15 U.S.C. § 1072 that makes "we had never heard of them" legally irrelevant. Read it before the naming workshop, not after.

The matter carried through this toolkit. Sabler Health, Inc. is a two-founder Durham company building a post-surgical recovery system: a wearable sensor, a patient app, a clinician dashboard sold as software-as-a-service, and remote monitoring delivered by nurse practitioners. Two marks matter — the house mark SABLER and the product mark TRELLIS. Four classes are in play: 9 (downloadable app), 10 (the sensor), 42 (SaaS), 44 (telehealth services). The sensor needs FDA clearance on a schedule nobody controls. Everything that can go wrong in a first portfolio can go wrong here.

The First Twenty-Four Months, Mapped

Trademark law runs on two clocks that founders routinely confuse. The first is the rights clock, which is driven by use: American trademark rights arise from selling something under a name, and at common law they exist only where the mark is used and known. United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90 (1918). The second is the register clock, which is driven by filing dates and administrative deadlines and cares nothing about whether the business succeeds.

Federal registration is the bridge between them. Filing a Section 1(b) intent-to-use application gives a company something the common law never would: a nationwide priority date as of the filing day, contingent on the mark eventually registering. 15 U.S.C. § 1057(c). That single provision is why a company with no revenue and no customers should still file in month one. It buys a place in line ahead of every business that starts using the name tomorrow, anywhere in the country, and it does so for a base government fee of $350 per class.

What the filing does not buy is a registration. An intent-to-use application matures only when the applicant proves actual use in commerce, on a schedule that begins at the notice of allowance and ends thirty-six months later with no extension, no excuse, and no tolling. 15 U.S.C. § 1051(d). Between those two events sits everything else: examination, publication, opposition, allowance, specimens.

Four decisions inside the first two years do most of the damage or most of the good.

The name itself. Where a mark sits on the distinctiveness spectrum of Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir. 1976), determines whether it registers on the first try, how much of the market it can police, and whether enforcement is a letter or a lawsuit. A descriptive name is not merely a weak asset; it is a recurring bill. It draws a Section 2(e)(1) refusal that costs money to answer, it cannot be registered on the Principal Register without proof of acquired distinctiveness, and it leaves competitors free to describe their own products in nearly identical words.

Who owns it. The applicant must be the party that owns the mark — meaning the party that controls the nature and quality of the goods. 15 U.S.C. § 1051(a); 37 C.F.R. § 2.71(d). Naming the wrong party is not a typographical problem. An application filed by an entity that does not own the mark is void ab initio, and shared officers and a shared address do not create the related-company relationship of 15 U.S.C. § 1055 that might otherwise save it. Great Seats, Ltd. v. Great Seats, Inc., 84 U.S.P.Q.2d 1235 (T.T.A.B. 2007).

What the application covers. The identification of goods and services is the operative legal text of the whole file. It defines what the examiner compares against prior marks, what a court will treat as the scope of the registration, and what the owner must eventually prove use of. Write it too narrowly and the enforcement radius is a pinhole. Write it too broadly and the registration becomes a target for a post-registration audit, an expungement petition, or an abandonment defense.

Whether the company can prove use when it has to. Use in commerce means the mark is actually on goods sold or transported, or services actually rendered to the public — not advertised, not planned. Couture v. Playdom, Inc., 778 F.3d 1379 (Fed. Cir. 2015); Aycock Eng'g, Inc. v. Airflite, Inc., 560 F.3d 1350 (Fed. Cir. 2009). A specimen has to show the mark functioning as a source identifier at the point of sale, which is why webpage screenshots without ordering information keep failing. In re Siny Corp., 920 F.3d 1331 (Fed. Cir. 2019).

Here is the shape of the two years.

| Month | Event | The decision it forces | | --- | --- | --- | | −2 to 0 | Naming workshop | Distinctiveness tier; three to five candidates, not one | | 0 | Knockout screening | Kill weak candidates before anyone falls in love | | 0–1 | Full search on survivors; use investigations | Go, go-with-conditions, or pick again | | 1 | Written clearance opinion; file § 1(b) application | Owner, classes, identification, basis | | 1–2 | Domains, handles, entity name, DMCA agent | Secure the surfaces a registration does not reach | | 3–9 | Examination; office action, if any | Three-month response clock; argue or amend | | 8–14 | Publication; 30-day opposition window | Watch; negotiate consent or coexistence if opposed | | 10–15 | Notice of allowance issues | The six-month clock starts; docket all six deadlines | | 12–24 | Launch; capture specimens; file SOU or extensions | Divide, delete, or extend, class by class | | 12–24 | Watch service; first enforcement letters | Log, escalate, or ignore — but decide deliberately | | 18–24 | Series A diligence | Produce a chain of title that does not need explaining |

Two structural points about that table. First, the register clock is far slower than a founder's intuition — filing to registration on a use-based application typically runs a year or more, and an intent-to-use application that waits on a product launch can run three. Second, almost every irreversible decision happens in months 0 through 2, when the company has the least money and the least patience for legal process. That is the design problem this toolkit is built around. The staged, deadline-driven view of prosecution is in What Happens After You File: The Examination Timeline, a short orientation worth handing to a non-lawyer founder before the first call; the fuller institutional version, covering all five filing bases and what each forecloses, is the Trademark Application and Prosecution Toolkit.

Month −2 to 0 — The Naming Workshop

The naming workshop is a legal event that companies stage as a marketing event. Bring counsel into it, because the cheapest possible intervention is the sentence "that one is descriptive, keep going."

Start with Choosing a Strong Trademark: The Distinctiveness Spectrum, the house primer on the five categories and the self-test that matters most: would a competitor need this word to describe its own product? It is two pages, it is written for founders rather than lawyers, and it is the right thing to circulate to the whole naming team the week before the workshop. Reach for it at the moment someone proposes a name that "tells customers exactly what we do."

When the team has fallen in love with a descriptive name anyway — and they will — the honest answer is in From Descriptive to Distinctive: How a Weak Mark Acquires Secondary Meaning. It explains the Inwood primary-significance test, who carries the burden, and why the evidentiary bar rises with every degree of descriptiveness. Use it to price the alternative: a descriptive name is a name you must buy twice, once in marketing spend and again in Section 2(f) evidence.

The whole axis, from fanciful marks at one end to genericide at the other, is assembled in the Distinctiveness and Genericness Toolkit. Founders should read its first third before naming; brand teams should read its last third before writing style guidelines, because Genericide: How Escalator, Aspirin, and Thermos Lost Their Trademarks describes the failure mode of a name that works too well.

There is a second screen most workshops skip. Section 2 of the Lanham Act bars more than confusion and description: surnames, primarily geographically descriptive terms, deceptive matter, and names of living individuals without consent. The Section 2 Bars: Surnames, Geography, Deception, and the First Amendment walks the whole statute and — this is the part that matters at the workshop — separates the refusals that acquired distinctiveness can cure from the ones that are permanent. Read it whenever a founder wants to use their own last name, a city name, or the name of a person who is still alive.

Trap. Sabler is a surname. It is also the co-founder's surname, which triggers nothing under 2(c) — that clause protects other living individuals — but it does invite a 2(e)(4) primarily-merely-a-surname refusal, curable under 2(f) but not for free. Sabler Health knew this going in and budgeted for it. Knowing about a probable refusal in month zero converts a surprise into a line item.

Month 0 to 1 — Clearance, and the Difference Between Screening and Clearing

Protectability and availability are different questions. The workshop answers the first. Clearance answers the second, and it is where the money is.

Begin with the Trademark Clearance and Brand Selection Toolkit, which arranges the whole sequence — distinctiveness, knockout, full search, opinion, filing — in the order the work actually happens. It is the right first stop for anyone who has never run the process end to end and needs to see the shape before the detail.

Then the three documents that do the work.

Trademark Clearance Searching: What a Knockout Search Can and Cannot Tell You is the conceptual piece: what a free federal-database screen structurally cannot see (common-law users, state registers, business names, domains, unpublished foreign priority filings), why the DuPont factors rather than the search box define the target, and what a bad search buys the other side in a willfulness fight after Romag Fasteners, Inc. v. Fossil, Inc., 590 U.S. 212 (2020). Read it first, and read it whole.

Running a Full Trademark Clearance Search: A Practitioner's Guide to Screening, Analysis, and the Written Opinion is the execution manual: twelve stages from the intake call to the signed opinion, with a five-axis hit-triage rubric, model opinion language, a seventeen-day annotated timeline with real dollar figures, and a rush protocol for the client whose product is already at the printer. This is the document counsel works from; a founder should read the budget stage and the rush protocol and skip the rest.

Trademark Clearance Search Checklist: From Knockout to Written Opinion is the same protocol as tickable actions across twelve phases, including the variant set, the common-law sweep, the use investigations, and the international screen. Print it when you are actually running a clearance and want a record that each step happened.

The step founders cut, and should not, is the use investigation — paying an investigator to determine whether a threatening reference is actually in use, on what goods, in what channels, since when. It takes three to five business days and it does not compress. It is also the step that most often changes the answer.

Practice tip. If the client accepts risk on a screening-only search, put the acceptance in writing and say what was not done. A clean screenshot filed in a folder called "Legal" is worse than no file at all, because it looks like diligence that was never performed.

Sabler ran knockouts on four candidates on 6 and 7 January, killed two, ordered full searches on the survivors, and got the opinion on 24 January. Total: about $9,400 including vendor spend, against a rebrand exposure the founders priced at $310,000 once the sensor housing tooling was cut.

Month 1 — Filing: Owner, Basis, Classes, and the Sentence That Fixes Your Reach

Four decisions, in this order.

Owner. The applicant is the entity that controls quality. For Sabler that is Sabler Health, Inc. — not Dr. Amara Sabler personally, and not the Delaware holding company that exists only on a cap table. Get this wrong and the application is void, not amendable. The failure mode, its cure, and its non-cure are laid out in Trademarks in the Deal: Chain of Title, Security Interests, and the Anti-Assignment-in-Gross Rule, which treats the founder-owned registration as the single most common defect in venture-backed companies. Read its chain-of-title section before the first filing, not at the term sheet.

Basis. Almost every startup files under Section 1(b). Intent-to-Use Applications: Claiming a Trademark Before You Sell a Thing explains what the verified statement of bona fide intent actually commits you to — an objective, documentary standard after M.Z. Berger & Co. v. Swatch AG, 787 F.3d 1368 (Fed. Cir. 2015), not a question of sincerity — and why the absence of any contemporaneous documents is itself enough to prove the intent was not bona fide. It also sets out the constructive-use priority of § 1057(c), the thirty-six-month cap, and the assignment restriction of § 1060(a)(1). This is the most important article in the toolkit for anyone about to file, and the section on documenting intent should be read as a records-retention instruction: keep the product roadmap, the vendor quotes, the class map, and the launch plan, dated.

Classes and text. The Nice Classification System: Why Your Identification of Goods Decides Your Trademark's Reach is the article to read the night before you draft. It makes the point founders miss: the class number is a filing cabinet, but the sentence inside the class is the deed. It maps the enforcement radius that follows from the words you choose, explains why "computer software" standing alone is always refused, and prices the January 2025 fee restructuring that replaced TEAS Plus and TEAS Standard with a base fee plus behavior-based surcharges.

Drafting an Identification of Goods and Services: A Practitioner's Guide to the ID Manual, Scope, and Amendment is where you go to actually write it. Thirteen stages, model identification text for downloadable software, SaaS, and retail services, the restriction language that wins DuPont factors without hollowing out the rights, and the one-way ratchet of 37 C.F.R. § 2.71 — you can narrow an identification, never broaden it. Use it during drafting and again the first time an examiner asks for definiteness.

Goods and Services Identification Checklist: Classes, Scope, and Specimen Fit is the pre-flight version, and its Phase 6 is the one that saves startups money: name the specimen for every class before you file, or drop the class. A class you cannot produce a specimen for is a class you will pay extension fees on for three years and then delete.

Pre-flight. Run the Pre-Filing Trademark Application Checklist last. It is deliberately short — legal name of the owner, dates, evidence, fees, proofread, docket — and it is the item-by-item confirmation that nothing obvious is missing before you hit submit.

Sabler filed on 3 February in four classes, using ID Manual entries in Classes 9, 42, and 44 and paying the $200 custom-wording surcharge in Class 10 because the Manual had nothing that described the sensor honestly. Government fees: $1,600.

Month 3 to 12 — Examination, Refusals, and the Three-Month Clock

Roughly half of applications draw at least one office action. Treat it as the examining attorney's opening argument, not a rejection.

The deadline changed and founders still get it wrong: most applications now get three months to respond, extendable once by three months for a fee, with Madrid Section 66(a) applications keeping the old six-month window. The 3-Month Office Action Deadline: What It Means for Applicants states the rule, the clock's start date (the issue date, not the day you read it), and the abandonment consequence. Docket it the day it arrives.

For triage — which refusal is this, and what actually works against it — the Office Action Response Toolkit: Refusals, Deadlines, and the Arguments That Work maps the whole terrain and routes you to the right response document. Then:

Responding to a §2(d) Likelihood-of-Confusion Refusal — the workhorse. Read it when the examiner cites a prior registration; its value is in showing which DuPont factors an examiner will actually weigh on an ex parte record.
How to Overcome a Descriptiveness §2(e)(1) Refusal — the second most common refusal, and the fork where you decide between arguing suggestiveness, amending to the Supplemental Register, and claiming 2(f).
Overcoming a Section 2 Refusal: A Practitioner's Guide to Surname, Geographic, and Deceptiveness Arguments — the guide Sabler needed for its surname refusal, with the Benthin factors and the evidence that moves them.
Section 2 Refusal Response Checklist: Diagnosis, Evidence, and Filing — run this before drafting so you diagnose the subsection correctly; the arguments that beat 2(e)(4) do nothing against 2(e)(3).
Office Action Response Checklist and the Response to Office Action — Template — the mechanical layer, for making sure the response is complete and properly signed.

If the refusal is a 2(f) problem, Claiming Acquired Distinctiveness at the USPTO: A Practitioner's Guide to Secondary Meaning Evidence sets out what an examiner will accept, and the Secondary Meaning Evidence Checklist tells you what to start collecting now — sales figures by year, advertising spend, unsolicited media, declarations — because a two-year-old company almost never has five years of substantially exclusive use to point to.

Publication opens a thirty-day opposition window under 15 U.S.C. § 1063(a), extendable. That is the moment a competitor's watch service sees you. It is also the moment your own watch should already be running, which is the subject of Trademark Watch Services: What to Monitor — a short guide worth reading in month two so the service is live before publication rather than after.

Month 10 to 24 — Allowance, Specimens, and the Six-Month Clock

This is where startups lose applications they already paid to prosecute.

The notice of allowance starts a six-month clock. The first extension is available as of right; the second is as of right; further extensions require good cause in six-month increments up to a hard ceiling of thirty-six months from allowance. 15 U.S.C. § 1051(d)(2); 37 C.F.R. § 2.88. There is no thirty-seventh month. A petition to revive under 37 C.F.R. § 2.66 can rescue a missed interim deadline; nothing rescues the outer limit.

From Notice of Allowance to Registration: A Practitioner's Guide to Statements of Use and Extension Requests is the operating manual for this phase: thirteen stages, a decision tree at the fork (file, extend, delete, or divide), model verification and good-cause language, the ten refusals a statement of use actually draws, a substitute-specimen response, and a complete fee table with three cost scenarios. Reach for it the day the notice of allowance issues, not the month before the deadline.

Statement of Use Filing Checklist: Specimens, Dates, and the Six-Month Clock is the eleven-phase working version — deadline ladder, ownership audit, class-by-class use verification, specimen capture, fixing first-use dates to documents rather than memory, and rescue procedures. Use it as the file's cover sheet for the whole allowance period.

The Request for Extension of Time to File a Statement of Use — Template is the fill-in-the-blanks version for the routine filings in between. It is a template, not a strategy: the good-cause showing under TMEP § 1108.02(c) has to describe real ongoing efforts, and filing identical boilerplate five times running builds exactly the documentary record an opposer's document request is aimed at.

Two moves save more startups than any other. The first is the amendment to allege use under 15 U.S.C. § 1051(c) and 37 C.F.R. § 2.76, filed if use begins before the mark is approved for publication — it skips the notice of allowance and the extension fees entirely. Miss that window and you hit the blackout period, three to five months in which neither filing is available. The second is the request to divide under 37 C.F.R. § 2.87: split the classes that are in use from the ones that are not, register the ready ones now, and keep the rest on the original clock.

Sabler used both. TRELLIS in Classes 9 and 42 launched in month fourteen; those classes were divided out and registered. Class 10 sat waiting on FDA clearance and Class 44 waited on the first nurse-practitioner visit, both carried on extensions. When specimens for the app came back refused because the screenshot showed no ordering information — the Siny problem exactly — the substitute-specimen playbook in Specimen Refusals: Why the USPTO Rejected Your Proof of Use resolved it in one filing.

The Ownership File Nobody Opens Until Diligence

Three documents decide whether Series A diligence is a formality or a fire drill, and none of them is a trademark filing.

Who Owns the Work: Employees, Contractors, Joint Authors, and Work Made for Hire is the one to read first, and the fact that stops founders cold is this: logos are not on the statutory list of commissioned works eligible for work-made-for-hire treatment under 17 U.S.C. § 101. Paying a designer does not make you the author; only a signed assignment under § 204(a) transfers the copyright. Read it the week you engage any outside designer, developer, or agency.

Copyright Ownership and Chain-of-Title Checklist: Assignments, Work-for-Hire, and Termination Windows converts that into a document-collection exercise. Run it in month two, when the designer still answers email, rather than in month twenty when they have moved to Lisbon.

Trade Secrets and the DTSA: Protecting What You Cannot Register covers the asset class with no registry, where protection exists only so long as the owner behaves as though secrecy matters. For a startup, the practical output is small and cheap: NDAs before disclosure, an onboarding assignment for every employee, and an exit checklist. It is the right read the first time you hire.

The trademark-side ownership document is the anti-assignment-in-gross rule. A trademark cannot be transferred apart from the goodwill it symbolizes, 15 U.S.C. § 1060(a)(1), and an intent-to-use application cannot be assigned at all before an allegation of use except to a successor to the applicant's ongoing business. Founders routinely trip this during a Delaware reincorporation or an F-reorg. See again Trademarks in the Deal, and use the Trademark Assignment Agreement — Template with the Trademark Assignment Recordal Checklist when a transfer is genuinely required — recordation matters because § 1060(a)(4) protects a later bona fide purchaser against an unrecorded prior assignment.

Where the company's rights are unregistered — a second product name in use but never filed — the territorial question governs. Where Your Trademark Rights End: Tea Rose-Rectanus, Dawn Donut, and the Geography of Common-Law Priority explains why those rights extend only as far as market penetration and reputation, and the Common-Law Priority Evidence Checklist tells you what to preserve to prove them. Both are worth reading before a diligence request asks you to describe unregistered marks.

If Your Goods Are Regulated

Some products cannot be registered at all, and no amount of prosecution skill changes that. The USPTO reads "use in commerce" to mean lawful use in commerce, and an examiner who establishes a per se violation of a federal statute refuses the application regardless of state licensure.

The Lawful Use Requirement: Why the USPTO Refuses Cannabis, Kratom, and Vape Marks traces the doctrine, the per se standard under TMEP § 907 and 37 C.F.R. § 2.69, the 0.3% hemp line the 2018 Farm Bill drew, and the delta-8 question the Ninth Circuit answered in AK Futures LLC v. Boyd Street Distro, LLC, 35 F.4th 682 (9th Cir. 2022). Read it before quoting a client in this space, because the answer to "can we register this" is often no.

Registering a Cannabis-Adjacent Trademark: A Practitioner's Guide to Hemp, CBD, State Registrations, and Ancillary Goods is the fifteen-stage execution manual for building a portfolio anyway — sorting the business into regulatory buckets, allocating goods among federal filings, state registrations, and common-law-only protection, and drafting identifications that foreclose the violation on their face. Reach for it once the client has decided to proceed.

Regulated-Industry Trademark Filing Checklist: Cannabis, Alcohol, Firearms, and Supplements generalizes the same eleven-phase workflow across TTB, ATF, FDA, and the drug schedules. Use it at intake for any client whose product needs a federal license, a label approval, or a marking variance — the regulatory clock and the trademark clock have to be docketed together.

Sabler's sensor is a Class II device. It is lawful, so § 907 is not in play, but the timing problem is identical: the thirty-six-month cap runs whether or not the FDA does.

Month 12 to 24 — Watching, Enforcing, and the Surfaces a Registration Does Not Reach

Registration is a right to exclude, not a system that excludes. Somebody has to look.

A watch service catches confusingly similar applications during the opposition window, which is the cheapest place to stop a conflict. Trademark Watch Services: What to Monitor covers scope; the Brand Enforcement Toolkit: Watching, Warning, and Escalating covers what to do with the hits, and its graduated ladder — log, correspond, complain, oppose, sue — is the right frame for a company that cannot afford to treat every hit as an emergency.

When escalation is warranted, Sending an Effective Cease-and-Desist Letter and the Trademark Cease-and-Desist Letter — Template are the starting point, and the strength of the letter depends entirely on what your registration covers, which is the identification decision from month one arriving with a bill. The substantive standard behind any threat is in Trademark Infringement: Proving Likelihood of Confusion.

Digital surfaces need their own sweep. Domains, marketplaces, platforms, handles, and paid search are mapped in the Online Brand Protection Toolkit; Cybersquatting and the ACPA explains the 15 U.S.C. § 1125(d) claim, and Filing a UDRP Complaint to Recover a Domain is the faster, cheaper route in most startup-scale disputes.

If the company's product is a website or an app, its pre-ship legal stack is a separate discipline with its own six-dollar filing that founders keep skipping. The Legal Layers of a Website: IP, Contracts, Privacy, and the DMCA Before You Ship maps all five layers, and the Website and App Launch Legal Checklist turns them into roughly 110 actions including the DMCA designated-agent registration under 17 U.S.C. § 512(c)(2).

Two failure modes deserve a read before the first partnership deal. Naked Licensing: How Sloppy Quality Control Kills a Trademark explains why an informal "sure, use our logo" can forfeit the mark entirely; Use It or Lose It: Trademark Abandonment, Non-Use, and the Three-Year Presumption explains what happens to the product line you quietly shelved. Both are short, and both describe injuries that startups inflict on themselves.

The Budget

Government fees are those effective 18 January 2025 under 37 C.F.R. § 2.6; confirm current amounts before every filing. Professional-fee ranges assume competent trademark counsel at U.S. small-firm rates, not a filing mill and not an AmLaw 50 partner.

| Month | Item | Government fee | Professional fee | | --- | --- | --- | --- | | −2 | Naming workshop attendance; protectability screen, 4 candidates | — | $600 – $1,200 | | 0 | Knockout screening, 4 candidates | — | $1,200 – $2,500 | | 0–1 | Full U.S. searches, 2 survivors (vendor $700–$1,750 each) | — | $2,000 – $4,500 | | 1 | Use investigations, 2–3 references (vendor $300 each) | — | $700 – $1,400 | | 1 | Written clearance opinion | — | $2,500 – $4,500 | | 1 | § 1(b) application, 4 classes, base fee | $1,400 | $1,200 – $2,400 | | 1 | Custom identification surcharge, 1 class | $200 | — | | 1–2 | Domains, handles, entity/DBA filings, DMCA agent ($6) | ~$400 | $0 – $600 | | 3–9 | Office action response, 1 substantive refusal | $0 | $900 – $3,000 | | 3–9 | Extension of time to respond, if needed | $125 | $150 | | 8–14 | Watch service, first year | — | $400 – $1,200 | | 10–24 | Extension requests, 2 classes × 2 periods | $500 | $600 – $1,200 | | 12 | Request to divide | $100 | $300 – $600 | | 14 | Statement of use, Classes 9 and 42 | $300 | $700 – $1,400 | | 15 | Substitute specimen after refusal | $0 | $400 – $900 | | 20–24 | Statement of use, Classes 10 and 44 | $300 | $700 – $1,400 | | 18–24 | Series A IP diligence response and cleanup | — | $1,500 – $5,000 | | Two-year total, four classes, two marks | | ≈ $3,325 | ≈ $13,850 – $31,750 |

Three observations. Clearance is roughly a third of the professional spend and prevents the only six-figure outcome on the list. Government fees are noise — the $200 custom-identification surcharge that founders resist is the cheapest accuracy anyone will ever sell them. And the single largest controllable variable is office actions: a well-drafted identification and a properly cleared mark is the difference between one refusal and three.

A leaner program exists. One mark, two classes, ID Manual wording throughout, a knockout plus a single full search, no litigation, no divide: roughly $1,600 in government fees and $7,000 to $11,000 in professional fees over two years. Portfolio-level budgeting, docketing, and reporting, once there is more than one mark to manage, is the subject of Building and Managing a Trademark Portfolio and the Trademark Portfolio Management Toolkit.

The Five Mistakes That Cost the Most to Unwind

| # | Mistake | Cost to avoid | Cost to unwind | | --- | --- | --- | --- | | 1 | Application filed in the founder's personal name | $0 — ask one question at intake | Refile, lose the priority date; $1,500–$6,000 plus the gap | | 2 | § 1(b) application assigned in a reincorporation before any allegation of use | $0 — sequence the paper | Application and any resulting registration void; refile at zero | | 3 | Clearance stopped at the free federal database | $4,000–$9,000 | Rebrand at scale: $150,000–$400,000+, plus exposure | | 4 | Identification padded with goods the company will never sell | $0 — cut at drafting | Audit, partial cancellation, abandonment defense, fraud allegation | | 5 | Logo never assigned in writing by the designer | $0–$500 | Buy it back at leverage, or redraw; diligence escrow |

1. The founder's name on the application. The applicant must own the mark, and ownership means quality control. Great Seats holds that an application filed by the wrong existing entity is void ab initio and that shared officers and premises do not create related-company status under § 1055. Assignment cannot cure a void application; only refiling can, and refiling means a new priority date.

2. The intent-to-use application assigned too early. Section 1060(a)(1) forbids assignment of a § 1(b) application before an amendment to allege use or a statement of use, except to a successor to the applicant's ongoing and existing business. Clorox Co. v. Chemical Bank, 40 U.S.P.Q.2d 1098 (T.T.A.B. 1996), voided both the application and the registration that issued from it. A Delaware flip, an F-reorg, or a lender's collateral grant can all trip it. Check the box before signing anything.

3. Clearance that stopped at the free database. Constructive notice under 15 U.S.C. § 1072 means a prior federal registration binds you whether or not you knew about it, and Romag removed willfulness as an absolute precondition to a profits award under § 1117(a). The screen produces reliable no answers. It does not produce a yes.

4. The padded identification. Founders list everything the company might someday sell. Every unused item is a lack-of-intent target — Kelly Services, Inc. v. Creative Harbor, LLC, 846 F.3d 857 (6th Cir. 2017), strikes the affected goods rather than the whole application, which is mercy, not safety — and an audit or expungement target after registration. Fraud requires an intent to deceive under In re Bose Corp., 580 F.3d 1240 (Fed. Cir. 2009), a high bar, but the allegation alone is expensive.

5. The logo the designer still owns. A commissioned logo is not a work made for hire, so absent a signed § 204(a) assignment the company holds at most an implied nonexclusive license. That is enough to keep using the logo and not enough to survive a representation that the company owns its IP.

Runner-up, and it is close. The thirty-six-month cap on an intent-to-use application. There is no relief from it. Divide early, register what you can, and refile the rest rather than watching a filing date die.

A Suggested Reading Path

The linear path — a founder starting from zero.

Choosing a Strong Trademark — thirty minutes, before the naming workshop.
Trademark Clearance Searching: What a Knockout Search Can and Cannot Tell You — the whole thing.
Intent-to-Use Applications — why to file before you sell anything.
The Nice Classification System — before drafting the identification.
Pre-Filing Trademark Application Checklist — the morning you file.
What Happens After You File: The Examination Timeline — so nobody panics in month seven.
Statement of Use Filing Checklist — the day the notice of allowance issues.
Docketing Deadlines: Never Miss a Renewal — the habit that makes the rest survive.

The branching path — counsel with a specific situation.

The client already launched under an uncleared name. Running a Full Trademark Clearance Search (rush protocol) → Where Your Trademark Rights End → Establishing and Proving Common-Law Trademark Rights → Common-Law Priority Evidence Checklist.
The mark drew a refusal. Office Action Response Toolkit → the matching response guide → Office Action Response Checklist.
The name is descriptive and the founders will not move. From Descriptive to Distinctive → Claiming Acquired Distinctiveness at the USPTO → Secondary Meaning Evidence Checklist.
The product is regulated. The Lawful Use Requirement → Regulated-Industry Trademark Filing Checklist → Registering a Cannabis-Adjacent Trademark.
A term sheet just arrived. Trademarks in the Deal → Trademark Due Diligence Checklist → Copyright Ownership and Chain-of-Title Checklist.
The product design itself is the brand. Trade Dress and the Functionality Doctrine → Protecting Trade Dress.
Expansion abroad is on the roadmap. The Madrid Protocol: How International Registration Works → Filing an International Trademark via the Madrid Protocol → Madrid Protocol Application Checklist. File within six months of the U.S. application to claim Paris Convention priority.

Primary Authorities

| Authority | Holding or rule | Why it matters in year one or two | | --- | --- | --- | | 15 U.S.C. § 1051(a), (b) | Registration on actual use or on verified bona fide intent to use | The two doors; almost every startup uses the second | | 15 U.S.C. § 1051(c), (d) | Amendment to allege use; statement of use within six months of allowance, extendable to 36 months total | The clock that converts an application into a registration or kills it | | 15 U.S.C. § 1057(c) | Filing is constructive use nationwide as of the filing date, contingent on registration | The reason to file before launch | | 15 U.S.C. § 1052(d), (e), (f) | Confusion bar; descriptiveness, surname, and geographic bars; acquired distinctiveness | Where applications die and how some are saved | | 15 U.S.C. § 1055, § 1127 | Use by a controlled related company inures to the owner | The doctrine that sometimes rescues a founder-named filing | | 15 U.S.C. § 1060(a)(1), (a)(4) | No assignment of a § 1(b) application before an allegation of use; bona fide purchaser protection on recordation | The most destructive trap in startup prosecution | | 15 U.S.C. § 1072 | Registration is constructive notice of the registrant's claim | Why ignorance of a prior registration is no defense | | 15 U.S.C. § 1063(a), § 1058, § 1065 | 30-day opposition window; § 8 declaration in years 5–6; § 15 incontestability | The three post-filing dates to docket at once | | 15 U.S.C. § 1117(a), § 1125(a), (d) | Remedies; unregistered-mark protection; cybersquatting | What enforcement is worth and what it reaches | | 17 U.S.C. §§ 101, 201(b), 204(a) | Work made for hire; initial ownership; signed writing required for transfer | Why the designer, not the company, may own the logo | | 37 C.F.R. §§ 2.6, 2.66, 2.69, 2.71, 2.76, 2.87, 2.88 | Fees; petition to revive; lawful-use inquiry; amendments; AAU; divide; SOU | The rule numbers this whole workflow runs on | | TMEP §§ 907, 1108.02(c), 1402 | Lawful use; good cause for extensions; identification practice | The examiner's own instructions | | Abercrombie & Fitch Co. v. Hunting World, Inc., 537 F.2d 4 (2d Cir. 1976) | The five-category distinctiveness spectrum | The frame for every naming decision | | In re E. I. du Pont de Nemours & Co., 476 F.2d 1357 (C.C.P.A. 1973) | Thirteen likelihood-of-confusion factors | The analytic target of clearance and of every 2(d) refusal | | M.Z. Berger & Co. v. Swatch AG, 787 F.3d 1368 (Fed. Cir. 2015) | Bona fide intent is a statutory requirement tested objectively on the totality | Keep contemporaneous documents or the ITU is vulnerable | | Great Seats, Ltd. v. Great Seats, Inc., 84 U.S.P.Q.2d 1235 (T.T.A.B. 2007) | Application by the wrong existing entity is void ab initio | The founder-owned filing, and why paper cannot cure it | | Clorox Co. v. Chemical Bank, 40 U.S.P.Q.2d 1098 (T.T.A.B. 1996) | Assignment of an ITU before use voided the application and the registration | Sequence every corporate restructuring around it | | Couture v. Playdom, Inc., 778 F.3d 1379 (Fed. Cir. 2015) | Advertising a service without rendering it is not use in commerce | The bar the statement of use must clear | | In re Siny Corp., 920 F.3d 1331 (Fed. Cir. 2019) | A webpage specimen must show point-of-sale ordering information | The most common specimen refusal for software and DTC brands | | Kelly Servs., Inc. v. Creative Harbor, LLC, 846 F.3d 857 (6th Cir. 2017) | Lack of intent as to some goods strikes those goods, not the application | Padding is a live risk, not a free option | | Community for Creative Non-Violence v. Reid, 490 U.S. 730 (1989) | Common-law agency test decides employee status for work made for hire | Why contractor deliverables need signed assignments | | United Drug Co. v. Theodore Rectanus Co., 248 U.S. 90 (1918) | Common-law rights are territorial and remote good-faith users are protected | What unregistered use actually gets you | | Romag Fasteners, Inc. v. Fossil, Inc., 590 U.S. 212, 218 (2020) | Willfulness is not an absolute precondition to a profits award under § 1117(a) | Raises the price of adopting without clearing | | B&B Hardware, Inc. v. Hargis Indus., Inc., 575 U.S. 138 (2015) | TTAB likelihood-of-confusion decisions can have issue-preclusive effect | An opposition is not a low-stakes skirmish |

Forms and Templates

Request for Extension of Time to File a Statement of Use — Template — the routine filing between allowance and launch. Pair it with the good-cause language in the statement-of-use guide; the template gives you the shell, not the showing.
Response to Office Action — Template — structure and required statements for a complete response. Use it with the diagnosis you get from the response checklist, not instead of one.
Trademark Assignment Agreement — Template — for transfers that are genuinely necessary, with goodwill language that satisfies § 1060. Check the ITU restriction before executing anything.
Trademark Assignment Recordal Checklist — the recordation step founders skip; § 1060(a)(4) rewards it.
Trademark Portfolio Inventory — Template — the spreadsheet a Series A data room asks for. Start it at the first filing and it takes ten minutes to produce in month twenty.
Trademark Coexistence Agreement — Template — when clearance turns up a conflict that can be negotiated rather than avoided. Read it before you assume a conflict is fatal.
Trademark Cease-and-Desist Letter — Template — the first enforcement step, and the one that most often ends the matter.
Trademark License Agreement — Template — for the first partnership or co-branding deal, with the quality-control provisions that prevent a naked-licensing forfeiture.

Related Toolkits and Checklists

The Trademark Clearance and Brand Selection Toolkit is the deep version of this toolkit's months −2 through 1, and the right handoff when clearance turns into a real project. The Trademark Application and Prosecution Toolkit covers months 1 through 24 at institutional depth, including the four filing bases a startup rarely uses until it goes international.

Once there is more than one mark, the Trademark Portfolio Management Toolkit takes over budgeting, docketing, and board-level reporting, and the Trademark Maintenance and Survival Toolkit picks up at the § 8 filing in years five and six. The Annual Trademark Portfolio Review Checklist is the once-a-year hour that keeps both honest.

For the deal at the end of the two years, the IP Due Diligence Toolkit for Mergers, Financings, and Asset Sales shows what a buyer's or investor's counsel will actually look for, and Trademark Due Diligence in Mergers and Acquisitions: An IP Buyer's Guide shows how they will look for it. Reading the other side's manual is the fastest way to build a file that passes.

Founders shipping software should keep the Website and App Launch IP Toolkit beside this one; the two overlap only at the brand layer, and the rest of that stack — assent screens, the DMCA agent, the software bill of materials — has its own deadlines. And when the whole arc from naming to enforcement is the question, The Brand Owner's Master Toolkit is the parent document this one sits inside.