What this checklist is for. Clearance is the investigative work of confirming a proposed name is available and protectable before a business builds anything on it. There is no legal duty to search — and that is exactly the trap, because nothing stops you from walking eyes-closed into rights that already exist. This checklist operationalizes the five-movement framework set out in our full article, How to conduct a comprehensive trademark clearance search: knockout, full search, analysis, documentation, decision. Work it top to bottom for each candidate. It is educational, not legal advice.

Phase 1 — Scope and set-up

  • Generate a short list of five or six candidate marks, not one. Founders arrive in love with a single name; the knockout will kill some, so you want survivors.
  • For each candidate, define what you will actually sell — the goods and services, plus their natural extensions — because relatedness, not the name alone, drives conflict.
  • Identify the realistic geographic and channel footprint (national e-commerce vs. one storefront), since common-law rights are territorial.
  • Decide the risk tolerance up front: a venture-backed national launch and a neighborhood café can look at the identical conflict and rationally reach opposite conclusions.
  • Confirm whether the name raises a registrability flag (surname, geographic term, descriptive/laudatory wording, short letter string) — these need tailored searching and may be hard to protect even if available.

Why this matters. Clearance answers two different questions: availability (does this conflict with prior rights?) and protectability (will the law and the USPTO protect it?). A name can be gloriously available and still worthless because it is generic or merely descriptive. Scope both questions before you search. For choosing a mark that is worth clearing in the first place, see Selecting a protectable trademark checklist.

Phase 2 — The knockout search (quick triage)

  • Open the USPTO Trademark Search tool (the cloud-based system that replaced the retired TESS). Any guide still telling you to "search TESS" is pointing at a system that no longer exists.
  • Search the exact mark for identical or closely related goods/services.
  • Search phonetic equivalents (CALYX → KALYX, CALIX) — consumers hear marks as often as they read them.
  • Search visual and spelling variants and deliberate misspellings (CALYXX, CA-LYX).
  • Search conceptual equivalents — the synonym/translation/idea, not just the letters (the third DuPont factor compares connotation and commercial impression).
  • Use the USPTO Trademark ID Manual to see how your goods are described and which adjacent categories may be deemed related.
  • Run a quick .com and social-handle availability check — a legally available name you cannot deploy online is a poor business choice.
  • Knock out any candidate showing an identical mark for identical goods, a highly similar mark for closely related goods, or ownership by a famously aggressive enforcer.

Why this matters and the traps. The knockout is triage, not diagnosis — half an hour, not a multi-day workup. The cardinal trap is searching exact matches only: infringement does not require identical marks; the test is likelihood of confusion, and an exact-match search manufactures false confidence. Remember the Live/Dead field is a lead, not a verdict — a dead registration is not an abandoned mark; the owner may still be using it with live common-law rights.

Phase 3 — The full search (candidates that survive)

  • Re-examine the federal register rigorously across all phonetic, visual, and conceptual variants and related classes — and include pending applications, not just registrations.
  • Flag any intent-to-use (ITU) applications under § 1(b): their priority can relate back to a filing date earlier than your first use (15 U.S.C. § 1057(c)). A similar ITU filed before your first use is a time bomb. See Intent-to-use trademark applications and Federal registration constructive use, 15 U.S.C. § 1057(c).
  • Pull file history and set status alerts in TSDR (Trademark Status and Document Retrieval) for any worrying application.
  • Search state trademark registers — targeted to states where you and known competitors operate, plus the big commercial states (CA, NY, TX, FL, IL). There is no unified national database.
  • Search common-law uses — the hardest and most important part: business directories, trade publications, news/court dockets, and above all the open internet (organic and paid results, plus reverse-image search for logos and trade dress).
  • Clear domain names (.com first, then .net/.org, relevant ccTLDs and gTLDs like .app/.health/.ai) via WHOIS, and social handles across major platforms.
  • Check industry-specific registries where relevant (USAN/INN for drug names, TTB for alcohol, FINRA/SEC for financial services, licensing boards for professional services).
  • Consider commissioning a professional vendor full search (Corsearch, CompuMark) for higher-stakes launches; expedited turnaround exists for a fee.

Why this matters. A search limited to the federal register misses most of the universe — common-law marks appear in no registry, the fifty state registers are separate, and the digital layer is invisible to formal databases. Common-law searching is where lazy clearance dies. If your real worry is selling a product without infringing a patent, that is a different inquiry — see Conducting freedom-to-operate analysis for new products and Freedom to operate search checklist.

Phase 4 — Analysis: apply the legal standards

  • Run each surfaced conflict through the likelihood-of-confusion factors — the DuPont factors the USPTO applies (In re E.I. DuPont DeNemours & Co., 476 F.2d 1357 (C.C.P.A. 1973)), plus the circuit tests (Sleekcraft, 9th Cir.; Polaroid, 2d Cir.). See Trademark likelihood of confusion analysis checklist and Navigating the maze of trademark confusion.
  • Weigh the two dominant factors hardest: similarity of the marks (sight, sound, meaning, compared in their entireties as a hurried consumer would encounter them) and strength of the senior mark (the distinctiveness spectrum sets the width of its moat).
  • Map relatedness of goods/services — confusion turns on relatedness in the consumer's mind, not class number. DELTA coexists for airlines, faucets, and dental supplies.
  • Triage each conflict into High / Moderate / Lower risk, weighing not just legal merit but the cost of opposition or litigation even where you would win.
  • Investigate the doubtful conflict from public sources (owner's website, USPTO file, Secretary of State, D&B) — but avoid pretextual or direct-contact investigation that raises ethics issues.

Why this matters. Doubts on confusion are resolved against the junior user, who bears a duty to avoid a confusing mark (Bridgestone v. Federal Corp., 673 F.3d 1330 (Fed. Cir. 2012); Nina Ricci v. E.T.F. Enterprises, 889 F.2d 1070 (Fed. Cir. 1989)). After Romag Fasteners, Inc. v. Fossil, Inc., 590 U.S. 212 (2020), willfulness is no longer a strict precondition to disgorging profits — making the good-faith record clearance produces more valuable, not less.

Phase 5 — Document the process

  • Record the date and scope of each search and capture screenshots of results (database contents are volatile and may be unreproducible later).
  • Write up the analysis of every significant hit, the risk categorization, and the recommendation.
  • Understand the privilege architecture: the third-party search report is generally discoverable raw data; the attorney's opinion built on it is what privilege protects. Discuss handling of any negative finding with counsel before it is drafted.
  • Where stakes are high, obtain a formal clearance opinion from qualified counsel.

Why this matters. Documentation converts clearance work into clearance value: it supports a good-faith-adoption defense, founds any formal opinion, and satisfies investors and insurers. A competent, genuinely-relied-upon opinion is the "shield of good faith" — see The shield of good faith. Note: asserting advice-of-counsel generally waives privilege over the opinion, so the document you may one day wield as a shield is one the other side will read.

Phase 6 — Decide: proceed, modify, or abandon

  • Proceed when no significant conflict surfaces, conflicts are clearly distinguishable, and the business case justifies the residual risk (clearance reduces risk; it never zeroes it).
  • If proceeding, weigh filing before launch to lock in constructive-use priority — then move to Filing a USPTO trademark application checklist.
  • Modify when conflicts are problematic but not fatal: add a distinctive element, adopt a logo/special-form mark, pair with a house mark, or narrow the goods (without merely dodging a class number).
  • Clear the obstacle through a consent/coexistence agreement, a cancellation petition (if the reference mark looks abandoned), or a purchase/license — but be ready to walk away before negotiating, because asking and then proceeding anyway can supply bad-faith evidence.
  • Abandon when the conflict is an identical mark for identical goods, the owner is a notorious litigant, or multiple independent conflicts pile up. Walking away during clearance costs only the time spent; walking away after launch costs the trucks.
  • Re-run the knockout immediately before committing serious brand spend to catch applications filed while you were developing.
  • For foreign expansion, clear country by country (many jurisdictions are first-to-file); the Madrid Protocol streamlines filing, not clearance.

Common mistakes

  • Searching too narrowly (exact matches only) — breeds dangerous false confidence.
  • Getting the goods analysis backward — treating any similar mark as fatal, or any non-identical mark as safe.
  • Over-relying on dead registrations — a lapsed registration is not a lapsed mark.
  • Underweighting common-law searching — the unregistered rights most likely to ambush you live in no database.
  • Failing to investigate a doubtful conflict, or failing to refresh the search before launch.
  • Consulting counsel too late, after the cheap fixes (modify, pick another name) are gone.

Primary authority

  • Statutes: Lanham Act, 15 U.S.C. §§ 1051 (filing bases, incl. § 1(b) ITU), 1052 (grounds for refusal), 1057(c) (constructive use), 1072 (constructive notice), 1114 / 1125(a) (infringement), 1117(a) (remedies).
  • Cases: In re E.I. DuPont (1973); AMF v. Sleekcraft (9th Cir. 1979); Polaroid v. Polarad (2d Cir. 1961); Bridgestone v. Federal Corp. (Fed. Cir. 2012); Nina Ricci v. E.T.F. Enterprises (Fed. Cir. 1989); Romag Fasteners v. Fossil, 590 U.S. 212 (2020).
  • Agency/tools: USPTO Trademark Search (uspto.gov; successor to TESS), Trademark ID Manual, TSDR, TMEP; WIPO Madrid Protocol (wipo.int); ICANN UDRP. Database contents and tools change — confirm current systems before relying on any search.

Related resources

This checklist is educational and not legal advice. Clearance involves subjective, fact-specific judgments; for a high-stakes adoption, obtain a clearance opinion from qualified counsel.