How to Find Out Who Owns an Alcohol Brand: TTB Guide

Executive Summary

To find out who owns an alcohol brand, start with the TTB public COLA registry, identify the label applicant and permit, reconcile that operator to a legal entity, and obtain separate evidence connecting the brand rights to an owner. A Certificate of Label Approval (COLA) authorizes specified labeling activity; it does not establish trademark ownership. The Alcohol and Tobacco Tax and Trade Bureau (TTB) expressly distinguishes its label authority from trademark protection. ^[1] The distinction matters because federal spirits rules expressly contemplate the same brand being bottled by distillers under different ownership. ^[2]

Treat same brand, same operator, and same owner as separate findings. The registry includes approved, expired, surrendered, and revoked certificates, so finding a record is not evidence that its approval remains usable. ^[3] Search brand variants and the applicant's permit separately. The live search interface limits date ranges to 15 years. ^[4] Its advanced permit-number search warns that results include only data after April 26, 2003. ^[5] Brand searches can also miss records because of data entry or application wording. ^[6] Preserve the query and its coverage rather than treating a blank result as proof that a brand does not exist.

The next step is a dated join to the permittee export, followed by a jurisdiction-specific entity search. A matching permit can support an operator connection. It cannot establish that the applicant, bottling plant, importer, trademark holder, and ultimate parent are one acquisition target. California explicitly says ownership information is not made of record with its Secretary of State; Oregon assigns an assumed business name its own registry number. ^{[7] [8]} Those are practical reasons to retain separate record types and relationship evidence. Trademark records add another check, but the United States Patent and Trademark Office (USPTO) warns that recording a change does not always update the trademark database automatically. ^[9]

For private equity (PE) screening, the defensible output is an evidence crosswalk with a clear merge decision and an unresolved ownership question where necessary. This report provides an editable table and a relationship diagram. Its quantitative example is explicitly hypothetical: it demonstrates counting distinct permit identifiers, not measuring permit holders, owners, or the market. Public beneficial ownership reporting is not an alternative open directory: the Financial Crimes Enforcement Network (FinCEN) describes the reported information as non-public. ^[10]

Introduction and Background

An origination analyst may begin with a bottle photograph, a retailer listing, or a brand website. The immediate task is navigation: find the label approval or licensed business. The investment question comes afterward: which legal entity would be approached, and what evidence shows that it controls the relevant brand rights? Mixing those questions makes a source useful for one purpose appear conclusive for another.

Label language records functions. For domestic wine, the federal rule distinguishes the bottler from another person for whom the wine was bottled. ^[11] For domestic spirits, the required identification describes the function performed by the named person. ^[2] For imported spirits, the importer definition connects to the importer's basic permit and the specified import relationship. ^[12] A producer name, a “bottled for” statement, and an importer statement therefore deserve different relationship labels in the research record.

This report uses owner narrowly: the legal person holding the brand rights relevant to the contemplated transaction. It uses operator for the entity performing the documented permitted activity. These are research definitions, and the deal team should specify whether its mandate seeks a brand asset, an operating business, a production facility, or a corporate group. A group-level target may contain several operators; a brand-focused target may rely on an independent producer.

The scope is U.S. wine and spirits screening, with current source checks anchored to October 2026. It is an evidence method rather than a title opinion. For a transaction, counsel should assess the operative rights and transfer requirements; basic permits themselves are subject to a rule against voluntary transfer. ^[13] The method stays specific to alcohol records rather than reproducing a general entity-search directory or a complete trademark-assignment guide.

The operational value of a label identifier is visible outside research workflows. Vinoshipper's product-onboarding instructions request a TTB COLA Label ID. ^[14] That shows why a COLA is a useful product reference. It does not expand the certificate into evidence of equity or brand ownership.

Key Changes in the Research Decision

The changes below are recommended changes to an analyst's decision process, not claims that the underlying regulations have recently changed.

Separate the brand from the regulated activity

A label applicant can be the bottler in a contracted production arrangement. TTB's explanation applies the domestic bottler's approval responsibility even when another entity contracts for production. ^[15] The analyst should record that operating relationship without silently promoting it to “owns brand.” The spirits rule's explicit provision for a brand bottled by independently owned distillers reinforces the need for this distinction. ^[2]

Separate identity from economic ownership

An entity identifier is valuable because it distinguishes similarly named legal persons. California describes its entity number as an identifier issued on formation, qualification, registration, or conversion. ^[16] But the same agency says owner and shareholder information is not made of record. ^[7] Resolving a permittee to that entity answers an identity question. An ownership claim still needs evidence connecting the relevant rights to that person.

Separate source evidence from the research conclusion

Table 1 compares the evidence layers used in this report. Docket appears only in its position as the publishing research provider; the row states the evidence standard recommended here, without treating the publisher as an issuing authority.

The table is an evidence hierarchy, not a ranking of databases. An exact permit match and an exact entity identifier can make an operator link strong while the owner remains unresolved. Conversely, a trademark instrument may identify a rights holder without explaining who currently bottles the product.

Keep relationships reversible

Use separate nodes for brands, certificates, permits, entities, and premises. Attach the statement that supports each connection. The diagram shows the proposed research model; every ownership edge requires its own supporting evidence.

flowchart LR
    B[Brand text] -->|appears on| C[COLA record]
    C -->|names applicant and permit| P[Permittee record]
    P -->|identity evidence| E[State legal entity]
    C -->|label wording| F[Bottler or importer role]
    B -->|independent rights evidence required| R[Brand rights holder]
    R -->|separate corporate evidence required| O[Parent or equity owner]

This model permits a later correction without rewriting the original observation. “Same brand” is a label-text comparison. “Same operator” requires an operator join. “Same owner” requires an ownership link appropriate to the rights or equity being assessed.

Search the Public COLA Registry

Start with a reproducible lookup

Use the public registry for external research. Record the certificate's displayed status before considering the other fields. Do not confuse the existence of an earlier approval with the current status of the certificate or the commercial availability of the product. Record what the status field actually says.

The following is a recommended search sequence:

• Capture the starting evidence: retain the brand spelling and the exact producer, bottler, or importer wording visible on the label.
• Use a known identifier first: when a COLA ID is available, preserve it as text rather than recreating it from a product name. Vinoshipper's instructions specify checking that the ID has 14 digits. ^[14]
• Search the distinctive stem (Hypothetical Example): use the percent wildcard where needed, for example %Copper Orchard%. The interface permits the percent sign in non-date search fields. ^[4]
• Run spelling variants: test punctuation, shortened wording, and the distinctive part of a longer name. Record each variant as an additional query and compare the candidates.
• Constrain dates deliberately: divide longer investigations into dated searches, because the live basic interface caps each date range at 15 years. ^[4]
• Pivot to the permit: search the applicant's exact plant registry or basic permit number, preserving the interface's historical coverage warning. ^[5]
• Log the result set: retain the query, filters, retrieval date, record identifiers, and any limit or error displayed.

Copper Orchard is an invented brand used only to illustrate a query. A broad search result should be treated as a candidate set for inspection. Shared words, a similar design, or the same town do not justify merging applicants.

Read the applicant and the label separately

Record the applicant's name and permit exactly as displayed, then transcribe the relevant label wording separately. For wine, the rule permits identification of a person for whom wine was bottled or packed. ^[11] For spirits, identification wording describes the named person's production or bottling function. ^[2] The difference belongs in structured fields, not merely in an analyst's memory.

Where the label uses a trade or operating name, the wine rule requires that name to match a name on the basic permit or another qualifying document. ^[11] That can help reconcile a name variant. It should not be rewritten as evidence that every business using related wording has the same owner.

Preserve historical search limits

The live advanced interface separates permit-number results after April 26, 2003 from the older Vendor Code route. ^[5] An empty permit search therefore cannot exclude earlier activity. If the question concerns an unsuccessful application or historical production before the applicable search coverage, record the limitation explicitly.

A useful search log ends with a narrow conclusion: “Certificates located under these terms and dates,” or “No matching certificate located within these searches.” It should not end with “brand has no owner” or “business is inactive” solely because the lookup returned nothing.

Implementation Considerations and Process Changes

Join the permit before normalizing the company name

Record the update basis stated on the permittee landing page before downloading. Retain the original export, the landing-page date, and the retrieval date as separate evidence. The exact export headers are described in the data section below; a worksheet's preferred field names should not be mistaken for official source fields.

The recommended join order is:

• Exact permit: compare the full identifier, retaining letters, punctuation, and leading zeros in the original value.
• Category check: confirm that the downloaded list covers the activity under investigation rather than treating all permit lists as interchangeable. ^[20]
• Name corroboration: compare the legal or operating name, documenting an alias only when another record supports it.
• Premises corroboration: compare the relevant address while distinguishing a production premises from a correspondence or agent address.
• Outcome classification: use exact, ambiguous, or no join. These are workflow outcomes, not official regulatory statuses.
• Entity reconciliation: resolve the operator into a state record before assigning a jurisdiction-qualified legal entity identifier.

Exact join means the permit key agrees and the corroborating fields are consistent. Ambiguous join means competing candidates or conflicting details remain. No join means the selected dated export did not establish a match. None of these labels independently means “owner verified.”

Keep the evidence table editable

Table 2 is an editable COLA-to-permittee-to-entity worksheet. Every name, identifier, date, and result in its example rows is fictional. The EXAMPLE prefixes are deliberately non-operative placeholders, not registry IDs. Replace them with actual observed values before using the worksheet for a real target.

The purpose of the table is to make a partial result usable without making it appear complete. A blank importer field means the analyst has not recorded an importer; it does not prove that no import relationship exists. A filled state identifier establishes the matched legal record, not the answer to the final ownership question.

Add a separate relationship row when sources disagree. Keep the earlier observation, describe the conflicting name or date, and state what evidence would resolve it. The original source should remain readable alongside the analyst's normalization. That is especially useful where a trade-name registration has a distinct identifier: Oregon expressly gives assumed business names their own registry numbers. ^[8]

Reconcile the Legal Entity and Ownership Evidence

Use the jurisdiction's actual record types

The following routes are focused checks for alcohol operators, not a substitute for reviewing the matched filing:

• California: match the permittee to the entity number, then inspect recorded officer, director, member, and manager roles. The Secretary of State distinguishes these records from ownership information it does not maintain. ^{[16] [7]}
• Delaware: search by name or file number and obtain the needed status evidence separately. The free results include active and inactive entities; the result list does not itself supply current status. ^[21]
• Florida: check a fictitious-name registration as an alias lead. Its registration creates no presumption of rights to own or use the name, and duplicate names are not automatically rejected. ^[22]
• New York: check entity and assumed-name records together. Its database includes assumed-name filings for corporations, limited liability companies, and limited partnerships. ^[23]
• Oregon: distinguish the assumed business name registration from the entity holding it. The assumed name has its own registry number, and the agency warns that its records do not include every business using a name. ^[8]
• Washington: reconcile the tax account name with the trade name; the Department of Revenue warns they may differ. Its Unified Business Identifier (UBI) is a nine-digit identifier used across state agencies. ^[24]
• Kentucky: search the organization by name or identifier and inspect its filings. Current officers shown in the online services are those on the most recent annual report. ^[25]
• Texas: separate franchise-tax account status from alcohol licensing. The Comptroller describes the former as the entity's right to transact business; the Texas Alcoholic Beverage Commission (TABC) separately offers pending, active, and inactive license and permit searches. ^{[26] [27]}

These routes give an analyst explicit fields to reconcile. They do not justify assuming that an officer, registered agent, license holder, and shareholder are equivalent roles. Delaware, for example, says operating agreements are maintained by the entity rather than filed with its Division of Corporations. ^[21]

Check trademark ownership without overreading the record

Use the brand wording and the resolved entity name to investigate the relevant trademark record. Keep application or registration identity separate from the applicant's permit identity. The USPTO warns that an assignment recordation does not always automatically change the trademark database's owner display. ^[9] Its examination manual also describes recordation as a ministerial act and recognizes instruments that do not transfer title. ^[17]

Accordingly, a name discrepancy is a question to reconcile, not an automatic contradiction. Note the instrument type, parties, and relevant dates, and determine whether the evidence concerns the particular mark and goods being screened. This report stops at the escalation point; a full trademark chain analysis is a separate workstream.

Establish any parent relationship separately

Where a relevant filer exists, the Securities and Exchange Commission's (SEC) Electronic Data Gathering, Analysis, and Retrieval system (EDGAR) provides a distinct filer identifier. The SEC says a Central Index Key (CIK) remains unique and is not recycled. ^[18] Subsidiary-list requirements can provide entity and jurisdiction evidence, but the analyst must inspect the actual exhibit and applicable scope rather than treating a text search as a complete corporate chart. ^[19]

FinCEN's beneficial ownership information (BOI) system does not solve the public lookup problem: reported information is stored in a non-public database and is exempt from Freedom of Information Act disclosure. ^[10] If independent ownership evidence is absent, retain the verified operator and leave the owner unresolved.

Coverage Limits and Escalation

Explain absence before drawing a conclusion

An ownership-research record should distinguish an absent observation from a negative finding. Several coverage boundaries affect a TTB COLA brand owner lookup:

• Application coverage: rejected and returned applications are excluded from the public registry. ^[6]
• Image coverage: the current registry summary describes images from 1999 onward, generally available 48 hours after approval, and warns that searches before 1996 may be incomplete. ^[3]
• Labeling jurisdiction: FDA guidance addresses diluted wines and cider below 7% alcohol by volume under FDA-administered labeling requirements. A missing TTB COLA cannot be interpreted without checking the product category. ^[28]
• Exemption possibility: TTB describes circumstances in which domestic wine or spirits sold within the bottling state may qualify for a certificate of exemption instead of a COLA. The analyst should confirm the applicable conditions rather than assume every product requires the same approval. ^[15]
• Permit-list coverage: the public permittee lists omit specified categories, including brewers. A missing row must be interpreted against those exclusions. ^[20]

State coverage also matters. Oregon warns that its registration records do not include every business using a name. ^[8] Florida permits duplicate fictitious-name registrations. ^[22] A search that finds too little and a search that finds too much both require inspection of the underlying identity evidence.

Escalate the unresolved relationship, not the whole target

The following checklist is a recommended handoff structure:

• Applicant differs from brand entity: request evidence defining the production, bottling, import, or brand-rights relationship.
• Same brand, multiple permittees: retain the separate operators and investigate the role and period of each.
• Trade name differs from legal name: obtain the assumed-name or permit evidence that connects them.
• State officer is proposed as owner: ask for evidence supporting the equity claim rather than substituting the office held.
• Trademark owner differs from operator: inspect the relevant trademark instruments and any license or production arrangement.
• Transaction is contemplated: have counsel evaluate the asset and entity perimeter, rights, and permit implications.

The final item is substantive: the federal basic-permit rule prohibits voluntary transfer of the permit itself. ^[13] A proposed purchase of brand rights should therefore not be described as automatically transferring the operator's permit.

Use precise unresolved statements: “Brand rights holder not independently established,” “Importer relationship observed; equity relationship unverified,” or “No join in the selected export.” Those statements preserve useful work while preventing an unsupported merge. The target remains researchable even when ownership requires non-public documentation.

Data Analysis and Evidence

What the dated permit source can establish

The consolidated permittee page states data as of October 19, 2026 and says updates are posted weekly in CSV and JSON formats. ^[20] These dates and formats describe the public export, not the effective date of every underlying business relationship. Keep the export's date distinct from the COLA's completion date, a state filing's date, and the analyst's retrieval date.

The inspected export schema has ten fields: Permit_Number, Owner_Name, Operating_Name, Street, City, State, Prem_Zip, Prem_County, Industry_Type, and New_Permit_Flag. ^[29] The combined JSON uses Report Header, Definitions, and Permit Data, with positional rows that must be mapped using the definitions. Crucially, Owner_Name means the individual or business holding the permit, not the owner of the brand or of that business. Operating_Name means the operating or doing-business-as name. ^[29]

The older wine-producer explanatory page adds a specific exclusion for bonded wine cellars without Federal Alcohol Administration Act permits. ^[30] It also describes monthly updates, whereas the current consolidated page describes weekly updates. Use the current landing-page date for the downloaded consolidated export and preserve that source discrepancy rather than inventing a universal cadence. ^[20]

A state source can have a different update basis. New York says its available entity information is updated daily; Texas says its franchise-tax results reflect records at query time. ^{[31] [26]} These statements do not make the sources simultaneous. They explain why a crosswalk should retain the date associated with each observation.

TABC can generate license and permit lists by location, type, status, or original issue date. ^[27] That is useful supplementary evidence when the question concerns a Texas operating license. Its counts should not be blended with a federal export as if both measured the same population.

Count distinct permit identifiers, not inferred owners (Hypothetical Example)

Table 3 demonstrates a reproducible calculation for the invented brand Copper Orchard. All quantities are constructed for teaching; no live brand query or empirical market sample is represented. Assume the search log retains six COLA records within a stated example window. The records contain the example permit keys A, A, A, B, B, and C.

The calculation answers how many distinct permit identifiers are associated with the queried brand in the retained sample. It does not answer how many permit holders, operators, owners, brands, corporate groups, or available acquisition targets exist. Distinct permits need not represent distinct permit holders: TTB Ruling 54-544 requires an importer to obtain a separate basic permit for each location where it sells distilled spirits, wine, or malt beverages it imported. ^[32] In a real sample, report the exact query, filters, retrieval date, export date, inclusion rule, and unresolved keys alongside the result.

Deduplicate by the documented permit identifier for the permit metric. Do not collapse by normalized company name merely because the name is similar. If the research later establishes that several permits belong to the same legal entity, calculate a separate entity metric and preserve the crosswalk used to derive it. Count distinct permit holders or operators only after separately reconciling their identities. Keep permit, operator, and owner counts in separate columns, with “not established” available as a legitimate result.

Implications and Future Directions

For origination teams, the immediate benefit of this method is a better-defined target perimeter. A brand can remain linked to several documented operating roles without inflating the number of independently verified owners. The research database can retain each permit and premises relationship while presenting a separate, evidence-backed view of the legal entity proposed for outreach.

The practical next step is to make the acceptance rule field-specific. Brand identity should preserve the observed wording and certificate references. Operator identity should preserve the exact permit and category. Entity identity should include the registry jurisdiction and identifier. Ownership should specify whether the statement concerns a trademark, a brand asset, equity in an operator, or a parent relationship. These definitions are proposed data controls, not additional representations made by the registries.

Refresh the relationship that is likely to have changed rather than overwriting every connected record. The USPTO's warning about owner-display updates makes a trademark name discrepancy worth examining independently of a permit match. ^[9] Washington's warning about trade names and tax account names likewise supports retaining both original and normalized names. ^[24] A later record can refine the interpretation without erasing what the earlier source actually said.

A useful quality review asks whether another analyst can reproduce the accepted connection from the retained evidence. Can that analyst locate the certificate, identify the permitted function, find the same legal entity, and explain the remaining ownership gap? If not, the handoff needs a narrower conclusion or more evidence.

Automated joins can assist with candidate generation and exact identifier comparison. The decision to merge targets should remain tied to the relationship being asserted. A shared label name can support further research; a matched entity identifier can support entity continuity; neither should silently become a complete ownership conclusion. The explicit separation is what allows the file to remain useful through screening, outreach preparation, and a later legal diligence process.

Conclusion

Finding an alcohol brand's owner is a sequence of evidence decisions. Begin with the label and its certificate, identify the applicant's regulated role, join the documented permit to the appropriate dated list, and reconcile the operator to the relevant state entity. Then investigate the rights and corporate relationships needed to answer the actual investment question.

The central discipline is to preserve the difference between same brand, same operator, and same owner. Each statement needs its own evidence. A bottler, importer, trademark holder, and parent may be connected, but the analyst should describe the connection observed rather than merge their records on a shared name.

The editable worksheet gives a partial finding a clear place to live. Its ownership-question column prevents a useful regulatory match from appearing more conclusive than it is. The hypothetical calculation makes the same distinction numerically: distinct permits can be counted within a defined sample, while owners remain unmeasured until ownership evidence exists.

For a PE origination handoff, a strong result is a reproducible identity crosswalk, a documented target perimeter, and a precise list of unresolved relationships. An honest “ownership not established” preserves the work already completed and tells the next reviewer exactly what evidence is needed before a broader claim can be accepted.