A Melbourne health-tech founder spends eight months in stealth building “Kindred.” The brand is chosen from a shortlist of forty; the domain is bought for $12,000 on secondary; the logo is commissioned from a Berlin studio for €18,000; a launch video is edited over Christmas. In late January the founder registers “Kindred Health” as a business name with ASIC, sees it accepted in two minutes, and takes that as the green light. In early February the marketing agency runs a $60,000 pre-launch campaign. In late February a cease-and-desist letter arrives from an Adelaide clinical-network company that has held a registered word mark “KINDRED” in class 44 (medical services) since 2019. Two weeks later a second letter arrives from a US telehealth company that has recently designated Australia under the Madrid Protocol. The founder has three problems at once: the brand cannot be used in the category it was chosen for, the domain is now a liability rather than an asset, and the Series A investor’s data-room checklist asks for evidence of freedom-to-operate on the wordmark. None of the work done to date can answer that question.
That is the ordinary shape of an Australian startup brand problem. The Corporations Act does not police brand collisions. ASIC’s business-name register does not check IP Australia’s trade-mark register. The two registers do not talk to each other. And Australia’s first-to-use common-law tradition sits alongside a first-to-file registration system in a way that is genuinely confusing until a founder has been on the wrong end of both.
The Three Registers That Aren’t the Same Thing
Australian founders regularly conflate three separate legal artefacts:
- A company name is issued by ASIC under the Corporations Act 2001 (Cth) and identifies the legal entity. “Kindred Health Pty Ltd” is a company name. It confers no brand rights whatsoever — a competitor can trade under an identical customer-facing name provided they hold a distinct company name of their own.
- A business name is registered on the ASIC business-names register under the Business Names Registration Act 2011 (Cth) and permits an entity to trade under a name other than its legal name. It is a disclosure obligation, not a property right. ASIC’s own guidance is explicit: business-name registration gives no exclusive right to use the name and is no defence to a trade-mark infringement claim.
- A trade mark is registered with IP Australia under the Trade Marks Act 1995 (Cth) in one or more classes of goods or services. Registration confers a statutory monopoly to use the mark in relation to the registered goods and services for an initial ten-year period, renewable indefinitely.
The registers do not cross-check. An available business name at ASIC routinely conflicts with a prior IP Australia registration, and vice versa. The dangerous inference — “ASIC accepted my business name, therefore I can trade under it” — is the single most common cause of forced rebrands in the Australian startup ecosystem.
First-to-Use Meets First-to-File
Australia is often described as a “first-to-use” jurisdiction. That is half true and misleads founders into complacency. What is accurate is that:
- The registered right is granted on a first-to-file basis. The applicant with the earliest priority date at IP Australia (or the earliest date of use recognised under section 44(4)) prevails in an examination or opposition.
- The unregistered right is protected by the common-law tort of passing off and by section 18 of the Australian Consumer Law (misleading or deceptive conduct). Both require the claimant to prove reputation in the mark at the relevant time, and reputation is expensive and slow to establish in evidence.
For a pre-launch startup, the practical translation is stark: reputation-based rights protect incumbents, not new entrants. A founder who has not yet launched has no reputation to assert. The only rights available to a pre-launch business are registration rights, and registration rights are decided by priority date. Filing early is not overcautious. Filing late is expensive.
Section 44 — What Examiners Actually Look For
Section 44 of the Trade Marks Act 1995 is the provision that decides whether a new application will be objected to on the basis of a prior mark. Two enquiries run in parallel: whether the applicant’s mark is substantially identical or deceptively similar to a prior mark, and whether the goods or services are similar or closely related.
“Deceptively similar” under section 10 is a lower threshold than “substantially identical.” It captures marks that so nearly resemble a prior mark that they are likely to deceive or cause confusion. IP Australia’s examiners apply a side-by-side visual test and an imperfect-recollection test, weigh phonetic similarity, and consider conceptual overlap. “KINDRED” and “KINDREDLY,” “Zenda” and “Zenba,” or two logos with different colours but identical dominant device elements will typically be flagged.
Section 44 objections can be overcome, but the standard pathways are narrow: prior continuous use before the cited mark’s priority date under section 44(4); honest concurrent use under section 44(3); consent from the cited-mark owner; or restricting the goods and services in a way that removes the overlap. None of these is a substitute for a proper clearance search at the front end.
The Clearance Search Founders Should Actually Run
A defensible pre-launch clearance search has four layers, and running only the first is the reason most rebrands happen.
- Identical-word search on the IP Australia register across the relevant classes and their nearest neighbours. Free, five minutes, catches obvious collisions. Insufficient by itself.
- Phonetic and structural similarity search using IP Australia’s advanced search or a trade-mark attorney’s proprietary tools, run against the classes the startup will operate in plus the classes into which the brand might reasonably extend within 24 months. A pure SaaS business will typically clear class 9 (downloadable software), class 42 (SaaS) and class 35 (advertising/business services); a fintech will add class 36; a health-tech will add class 44; a marketplace will add class 35 and often class 39.
- Common-law search for unregistered use — Google, Companies House equivalents in target export markets, industry directories, app stores, product-hunt archives. The point is to surface reputation-holders who could sue in passing off or under section 18 of the ACL even without a registration.
- Domain, social-handle and app-store availability check. Not legally decisive but commercially indispensable — a clear trade-mark is worth less if the .com is squatted by a live business.
The paid TM Headstart service ($200 for the Part 1 pre-application assessment, and an additional $130 per class to convert to a full application) provides an IP Australia examiner’s non-binding indication of registrability inside a week. For founders committing significant marketing spend to a brand, it is cheap insurance and often the difference between finding out at Part 1 and finding out after examination.
Choosing the Right Classes
Australia uses the Nice Classification — 45 international classes of goods and services, with classes 1-34 covering goods and classes 35-45 covering services. Trade-mark rights are class-specific. A registration in class 42 (technology services) does not stop a competitor from registering an identical mark in class 25 (clothing), which is exactly how the same “APPLE” wordmark can peaceably coexist across records companies, computer companies and grocery businesses.
For startups, class selection is one of the highest-leverage decisions in the whole IP program. Two errors are common:
- Over-narrow specification. A founder who registers only in class 42 for “SaaS platform for retail merchants” leaves class 9 (downloadable software), class 35 (retail services) and class 36 (payments) exposed. A competitor filing in one of those adjacent classes can force the startup to co-exist under a limitation, or worse, block a pivot.
- Over-broad specification with no supporting use. A class-9-through-42 blanket filing looks strong on the cap table diligence checklist and collapses under a non-use removal application after year three. Non-use applications under sections 92-93 of the Act are the standard tool competitors use to strip unused registrations, and IP Australia grants them regularly.
The disciplined pattern is: file in the classes covering current use, plus one or two adjacent classes covering credible near-term extensions, with specifications drawn from the IP Australia picklist to secure the $250-per-class fee rather than $400 for freeform text.
Ownership — Who Holds the Mark
The applicant of record on the trade-mark file is the legal owner of the registration. Founders regularly file in their personal name during a pre-incorporation stage and then never transfer the mark into the operating company. The consequences surface at diligence: the operating company does not own its own brand, and the founder-owner needs to execute an assignment (with stamp duty considerations, and often a small tax event under Division 100 of the ITAA 1997) before the round can close.
Two disciplines fix this at the outset. First, file the trade mark in the name of the operating company, not the individual. If the trade mark predates the company, execute a written assignment under section 106 and record it at IP Australia. Second, include an IP assignment clause in every founder accession deed, contractor agreement and employee contract from day one, expressly capturing trade marks along with copyright, designs and inventions. The clause is boilerplate. Its absence is a diligence red flag that has killed more term sheets than founders admit.
The 19 December 2025 Reforms
The Trade Marks Amendment (International Registrations, Hearings and Oppositions) Regulations 2025 took effect on 19 December 2025 and materially changed the opposition process. Two changes matter for founders:
- Notice of Intention to Defend now runs for two months, not one. Applicants whose accepted marks are opposed have a much less compressed window to instruct counsel and file a defence — a change that particularly benefits pre-Series A startups without in-house counsel. The extended period applies to oppositions where the acceptance was advertised on or after 19 December 2025.
- Automatic deferral of acceptance where a hearing is requested. Applicants no longer need to lodge a separate extension of time when a hearing is requested close to the acceptance deadline. Acceptance is automatically deferred until the hearing is resolved, removing a procedural trap that previously stripped applicants of rights for what were essentially administrative reasons.
The two-month opposition window itself — running from publication of acceptance under section 52 — has not changed. Any person may file a notice of intention to oppose within that window, and any startup filing a mark should assume that competitors and adjacent-class incumbents are monitoring the register for exactly the sort of collision that has just been accepted.
The International Layer — Madrid Protocol
Australia is a member of the Madrid Protocol. An Australian applicant with an existing IP Australia application or registration (the “base”) can file a single International Application designating any of the 130-plus Madrid member jurisdictions, with the mark and owner details tied to the Australian base for the first five years. Where a startup has credible near-term commercial activity in one or two overseas markets — usually the United States, the United Kingdom, Singapore, the EU or New Zealand — the Madrid route typically compresses filing costs by 40-60% relative to filing directly through local counsel in each jurisdiction.
Two structural cautions: the Madrid registration is dependent on the Australian base for five years — if the base is invalidated, the international designations fall with it (though they can be “transformed” into national applications). And the class specification cannot be expanded internationally beyond what is claimed in the Australian base. A startup filing internationally should therefore not shrink the Australian specification to the minimum required for local use — it needs enough breadth to support the international designations it intends to run through Madrid.
The Bottom Line
The founders who avoid the “Kindred” scenario treat trade-mark strategy as pre-launch product infrastructure, not a diligence-cleanup task. They run a proper four-layer clearance search before signing a brand-agency contract. They file in the operating company’s name across the classes covering current use and credible 24-month extension. They use TM Headstart for high-conviction brands and full applications for the rest. They file early to lock in a priority date rather than waiting until launch. They monitor the register on their own key words and their nearest neighbours. And they treat the two months after acceptance as a real risk window, not a formality — because the 19 December 2025 amendments have given opponents a longer runway to file, and applicants a longer runway to defend, and both sides are using the extra time. The startups that skip these steps discover — usually in the third week of a marketing campaign — that a brand is not the same thing as a name.
Viridian Lawyers advises Australian founders, technology companies and investors on trade-mark clearance, filing strategy, opposition and non-use proceedings, IP assignments through capital-raise diligence, and Madrid Protocol international programs. If you are scoping a new brand launch, cleaning up an IP portfolio ahead of Series A, or defending an accepted application in opposition, get in touch.
