Research and Reference Checking Before Sale

Research and reference checking is the stage where a seller turns an object from something owned into something that can be identified, described, priced and offered responsibly. It is not a search for the most valuable possible title. It is a controlled attempt to establish what the item probably is, which claims are supportable, what remains uncertain, whether the parts belong together, whether sale is lawful and which market results are genuinely comparable.

Good pre-sale research combines object evidence, specialist references, provenance, condition, completeness, legal checks and market evidence. No single catalogue number, certificate, archive entry or spectacular auction result should be allowed to carry the whole conclusion. The objective is a defensible listing: one whose language accurately reflects both the evidence found and the limits of the search.

The governing principle

Research should move in one direction: observation → comparison → verification → calibrated conclusion → defensible listing. Starting with a hoped-for identification reverses that sequence and invites confirmation bias.

Why pre-sale research changes the sale

Poor research can turn an ordinary issue into a supposed first edition, a later reproduction into an original, a faded surface into a rare colourway, or a replacement accessory into an original component. It can also produce subtler errors: using a premium-inclusive auction price as though it were the seller's return, describing a family story as documented provenance, or copying an unsold asking price from another seller.

Commercial consequence

Incorrect identity and unrealistic pricing produce failed listings, weak buyer trust, returns and disputes.

Evidential consequence

Unsupported claims enter the market and may be repeated by later sellers until speculation appears established.

Physical consequence

Misidentification can encourage cleaning, separation, repair or disposal of parts that should have remained together.

Legal and reputational consequence

Counterfeit, restricted or misdescribed material may trigger platform removal, regulatory problems or lasting damage to the seller's reputation.

Begin with the object, not the search result

The physical item is the primary source. Before opening a marketplace, catalogue or image search, record what is actually present. Transcribe exact wording rather than paraphrasing it; photograph markings before cleaning around them; measure rather than estimating; and record both associated components and obvious absences. This creates a neutral evidence base against which later interpretations can be tested.

Exact title, wording, logos, trademarks and names
Model, stock, catalogue, serial, matrix or product numbers
Copyright statements, patent information, date codes and addresses
Edition, printing, pressing or production statements
Signatures, initials, monograms, stamps and inscriptions
Dimensions, weight, materials, construction, colour and finish
Labels, stickers, price marks, packaging and inner packaging
Contents, accessories, paperwork and protective components
Repairs, alterations, substitutions, damage and signs of restoration
Invoices, certificates, correspondence and other provenance documents

Collector scenario: the valuable result appears first

A seller searches a character name and quickly finds a high auction result for a scarce early issue. Their own example looks broadly similar, so every ambiguous feature begins to feel like confirmation. A measured comparison later shows a different factory code, later safety text, revised packaging and a replacement accessory.

The disciplined response is not to preserve the valuable identification with cautious marketing language. It is to adopt the common later identification that fits the evidence, record the mismatched component and price the actual specimen.

Separate observation, conclusion and report

Direct evidence

Observed fact

Something visible, measurable or transcribable from the object itself.

The base is stamped ‘Made in Japan’.
The copyright page carries a particular number line.
The box measures 310 × 225 × 48 mm.

Interpreted evidence

Supported conclusion

A conclusion produced by comparing observations with reliable references.

The printer's key and publisher guide indicate a second printing.
The mould mark matches the issue catalogued for 1983.
The parts list confirms that the accessory belongs to this version.

Reported information

Unverified statement

A claim that may be recorded, but has not been independently demonstrated.

The previous owner called it a shop display model.
A forum post described it as a prototype.
Family recollection links it to a named collection.

Only observed facts and properly supported conclusions should normally appear as definite statements. An unverified history may be worth disclosing, but it must remain visibly unverified. Repeating it in a confident title, placing it beside a high estimate or issuing a seller-created certificate does not improve its evidential status.

Define the research question before choosing the source

“Find out about this” is too broad. Different questions demand different evidence. A publisher archive may identify a printing; a parts list may establish completeness; a conservator may identify repainting; government guidance may determine whether the item can be sold; completed sales may indicate current market behaviour.

•Who manufactured or issued this specific version?
•When was this production state made or released?
•Is it original, later, licensed, restored or counterfeit?
•Which edition, printing, pressing, state or regional issue is it?
•Is the colour a factory variant or later change?
•Did this accessory or packaging originally accompany this issue?
•Is the signature printed, stamped, mechanically reproduced or hand-applied?
•Was it sold individually or only as part of a set?
•Is the claimed rarity documented or merely repeated?
•Can it lawfully be advertised, sold and shipped through this venue?
•Which realised sales match the same identity and specimen characteristics?
•What evidence would disprove the preferred conclusion?

Build the identity all the way down to the specimen

Many errors occur because identification stops at the product or character. The sale, however, concerns a particular specimen. Two examples can share a title and basic design while differing materially in edition, factory, packaging, parts, condition, authenticity and value.

Object class

The broad category: book, coin, figure, poster, watch, game, ceramic or record.

Series or body of work

The franchise, range, issue family, artist's work or collecting field to which it belongs.

Maker or issuing authority

The publisher, manufacturer, mint, studio, artist, workshop or other responsible body.

Product or work

The particular title, model, design or named issue.

Edition or production state

The printing, regional issue, reissue, revised mould, pressing, state or packaging generation.

Variant

A documented colourway, cover, factory, language, retailer, code or production difference.

Individual specimen

This exact copy, including its condition, completeness, provenance, alterations and repairs.

Use a hierarchy of sources

Highest evidential value

Primary and governing sources

Use sources closest to production, publication, custody or the legal rule whenever they survive.

Maker, publisher, artist, mint or factory archives
Original catalogues, price lists, advertisements and trade literature
Official product documentation, records and registered identifiers
Museum, library and specialist institutional databases
Legislation, government guidance and official permit systems

Strong supporting value

Specialist secondary sources

These sources interpret and organise evidence, often making difficult fields usable for collectors.

Respected specialist books and revised catalogues
Scholarly articles and catalogues raisonnés
Major auction archives with full descriptions and reports
Long-running specialist dealers and collector organisations
Documented databases that explain their classifications

Lead-generating value

Market and community sources

Useful for terminology, photographs, recurring disputes and current market evidence, but rarely sufficient as sole proof.

Marketplace listings and dealer descriptions without citations
Collector forums, blogs and social-media posts
Crowdsourced databases and image-search matches
AI-generated summaries
Unattributed lists that repeat earlier listings

A listing is evidence of a claim, not proof of the claim

A marketplace page demonstrates that a seller used particular words and asked a particular price. It does not, by itself, prove the object's date, authenticity, rarity or value. Use listings to identify search terminology, recurring confusions, useful photographs and actual completed sales—not as an automatic substitute for references.

Read references critically

Printed books, online databases, institutional records and auction archives all carry context. Check when the source was produced, whether it has been revised, what evidence it cites, which regional or production scope it covers and whether its terminology is universal or specific to one classification system. A catalogue number should be tied to the named catalogue and edition; an old museum attribution may have been superseded; an auction description may be qualified by a condition report or later correction.

Read qualification words literally

“Attributed to”, “workshop of”, “school of”, “probably”, “possibly”, “circa”, “after” and “formerly attributed to” express different relationships and confidence levels.

Removing the qualifier changes the claim. A listing should not upgrade institutional caution into seller certainty.

Treat absence carefully

Absence from a surviving catalogue does not prove non-existence. Records may be incomplete, regional, provisional or lost.

Conversely, appearance in an advertisement does not prove that the exact version reached production. The defensible conclusion may remain narrower than either source first suggests.

A repeatable research sequence

Record the object

Photograph and transcribe what is present before forming a preferred identification. Include reverse, underside, edges, labels, seams, fasteners, inscriptions, defects and associated documents.

Define the question

Replace ‘find out what this is’ with precise questions about maker, date, edition, authenticity, colour, signature, packaging, completeness, legality or comparable sales.

Build the identity

Move from object class to exact issue, variant and specimen. Do not stop once the general product or character has been recognised.

Test references

Use the strongest available sources, check publication date and scope, and identify whether multiple sources are truly independent.

Seek disconfirming evidence

Ask what would show that the item is later, common, altered, incomplete, reproduced or mismatched. Valuable conclusions deserve more resistance, not less.

Translate evidence into language

Match the strength of the listing claim to the strength of the evidence. Record unresolved questions rather than disguising them.

Recheck after condition review

Condition, restoration and missing features can alter identification. Repeat key checks once the physical review is complete.

Edition, printing and production-state research

The earliest copyright date is not necessarily the manufacture date. Designs, text, moulds and artwork often persist through later editions. Identification may depend on combinations of number lines, publisher addresses, prices, matrix codes, mould changes, logos, barcode formats, country markings, safety text, packaging typography, factory codes and component substitutions. The aim is to identify the specific production state, not merely the underlying design.

Made intentionally

Documented variants

Regional issues, languages, retailer exclusives, colourways, alternate covers, promotional versions, packaging revisions and different factories may be deliberate variants.

Made variably

Production variation

Paint batches, paper stock, printing tolerances, mould wear, hand-finishing, substitutions and assembly changes can create differences without creating a separately recognised variant.

Changed later

Post-production difference

Fading, yellowing, staining, repainting, restoration, replacement labels, swapped parts, deformation and environmental damage are changes to the specimen, not factory variants.

Myth versus reality: difference does not automatically mean variant

Myth

Any colour, paint, print or packaging difference creates a rare collectible variant.

Reality

A variant claim normally needs consistent examples, reliable production information or a recognised reference. Ageing, fading, tolerance and later alteration must be excluded first.

Completeness and component relationships

Research should establish not only what the main object is but what accompanied this exact issue. Instructions, inserts, trays, stands, cables, cards, tokens, maps, dust jackets, certificates and protective pieces may change between editions. A component can be genuine in itself yet wrong for the specimen. Watch for parts from another issue, later official replacements, aftermarket copies, photocopied paperwork, reproduction packaging and married sets assembled from multiple sources.

Boundary with completeness and condition

Research identifies what should be present and which production features are normal. The physical completeness and condition reviews establish what is actually present and how it has changed.

These are separate tasks, but they must be repeated against one another. A replaced binding, missing label, trimmed margin or altered colour may remove or distort the very feature used for identification.

Authentication, signatures and certificates

Research can support authenticity, but reading reference pages does not make every seller an authenticator. Strong authentication may combine provenance, correct construction, period-consistent materials, recognised marks, serial structures, maker records, comparison with unquestionably genuine examples, specialist examination and scientific testing. The more value depends on genuineness, the more important specialist competence becomes.

Signature pathway

Determine whether the mark is printed, stamped, facsimile, autopen, hand-signed or later inscribed.
Compare letter forms, pressure, stroke sequence, placement, ink interaction and known signing habits.
Check whether provenance and date fit the purported signer and object.

Certificate pathway

Assess the issuer's standing in that specific collecting field.
Confirm that the certificate is securely tied to the object and has not been copied or substituted.
Remember that a seller-issued certificate repeating the seller's unsupported claim adds little evidence.

Provenance, title and lawful ownership

Provenance is documented ownership or custody history, not simply the story attached to an object. Invoices, catalogues, inventories, labels, correspondence, photographs, wills, estate records and published references can strengthen attribution and buyer confidence. Ordinary collectibles often have gaps; a gap is not automatically suspicious. The essential discipline is to distinguish documented history from consignor report or family recollection.

Documented formulation

“From the collection of [named person], supported by a dated inventory and purchase invoice.”

Reported formulation

“Acquired from the original owner's family, according to the consignor; no independent record has been located.”

For higher-risk material, provenance research is also a title check. Consider disputed inheritance, consignment authority, stolen-property records, institutional restrictions, archaeological finds, loaned property, finance interests and cultural-property claims. A physically convincing object can still have a problematic ownership history.

Legality and platform eligibility

A seller should check legality before writing the listing, particularly where an object may contain ivory, tortoiseshell, coral, feathers, fur, reptile skin, protected wood, taxidermy, archaeological material, human remains, weapons or other regulated substances. “Antique” is not a universal exemption. The research may need to cover ownership, advertising, domestic sale, international movement, registration, certification and whether documents must transfer with the item.

Boundary with legal advice

This chapter describes the collector's due-diligence process; it does not determine legal eligibility for a particular object. Where the rules depend on species, age, material, jurisdiction, intended destination or exemption status, use current official guidance and obtain specialist legal or regulatory advice where necessary.

Legal ownership does not guarantee platform eligibility. Review the chosen venue's current rules for counterfeits, replicas, weapons, animal products, archaeological material, currency, medical items, offensive material and shipping restrictions. Platform suitability is part of research, not an afterthought.

Comparable sales: compare the item that exists

Comparable-sales research asks what sufficiently similar items actually sold for under reasonably similar circumstances. Active listings can reveal seller expectations and market language, but completed sales are the stronger starting point. One exceptional result should not be allowed to define the range.

Exact identity, edition and variant
Authenticity or certification status
Completeness and original packaging
Condition, restoration and alterations
Provenance, signature and unusual associations
Size, format and grading status
Venue, location and sale date
Whether the lot sold, failed, was withdrawn or was relisted
Whether the result is hammer price, buyer total or seller net
Whether shipping, tax, VAT, premium or bundled items distort the figure

Normalise the price

Distinguish auction hammer, buyer total and seller net. Record premium, tax, VAT, shipping, currency, commission, bundled items and post-sale adjustments before comparing figures.

Measure market depth

Record how often the item is offered and sold, time to sale, failures, relistings, price spread and seasonal behaviour. A single high sale does not necessarily indicate a liquid market.

Value has more than one practical form

Research may support several different numbers: theoretical top value, achievable market value, patient retail value, quick-sale value and dealer-buying value. These should not be treated as synonyms. The suitable figure depends on the selling objective and evidence of market depth.

Rarity claims require precise definitions

“Rare” can mean few made, few surviving, few appearing for sale, few surviving in this condition, few currently listed or merely few encountered by the seller. Those are different propositions. Production records, censuses, grading populations, auction frequency, serial ranges and specialist registries can all contribute, but none should be interpreted without limits. Population reports may omit ungraded examples, include resubmissions, cover only one service and change over time. Rarity also does not create demand.

Prefer measurable wording

“Three auction appearances were located in the sources checked between 2016 and 2026” is more useful than “extremely rare”. It states the evidence, the search scope and the period without pretending to know the total surviving population.

Triangulation, negative findings and conflicting references

Important claims should ideally be supported through independent evidence: for example, an original catalogue confirms that the product existed, a specialist reference distinguishes the variant, a museum record confirms dimensions and completed sales show current market behaviour. Independence matters. Multiple websites repeating the same wording or photograph may represent one source, not several.

Record negative findings

State which catalogues, guides, archives and sales records were checked and what was not located.

Failure to find a claimed green factory version does not prove it never existed, but it defines the current evidential limit and prevents unsupported certainty.

Resolve disagreement transparently

Identify the exact disagreement, compare dates and citations, confirm that the sources concern the same regional issue and favour transparent reasoning over confident assertion.

Significant unresolved conflicts belong in the listing or research record rather than being hidden.

Translate research into calibrated listing language

Strongly established

State the fact directly

Use formulations such as ‘marked’, ‘dated’, ‘manufactured by’, ‘catalogued as’ or ‘identified in [named reference] as’ where the evidence is clear.

Reasonably supported

Qualify the conclusion

Use ‘appears to be’, ‘consistent with’, ‘probably’, ‘attributed to’ or ‘circa’ where the conclusion is credible but not absolute.

Unresolved

Disclose the limit

Use ‘possibly’, ‘reported by the previous owner’, ‘unverified’ or ‘no independent confirmation located’ only when the uncertain history is itself relevant.

Unsupported

Do not market speculation

Avoid ‘prototype’, ‘one of a kind’, ‘museum quality’, ‘investment grade’, ‘factory error’, ‘celebrity-owned’, ‘complete’ or ‘first edition’ unless the required evidence exists.

A disclaimer does not cancel a definite claim

“I am not an expert” does not neutralise “genuine first edition”. A seller should either support the definite statement, qualify it to match the available evidence or remove it. Uncertainty should be expressed in the claim itself, not placed in a general disclaimer elsewhere.

Research depth should follow risk

Routine threshold

Low-value, familiar item

Document identifying marks, confirm the exact issue, compare several completed sales, check platform restrictions and describe condition and completeness accurately.

Enhanced threshold

Moderate-value or easily confused item

Consult recognised references, compare variant indicators, investigate counterfeit warnings, document provenance and use sales from more than one source or venue.

Specialist threshold

High-value, rare or sensitive item

Create a research dossier, seek expert input, investigate title and provenance, consider authentication or grading, check legal and export rules and choose a specialist venue.

Proportionate does not mean casual. Even an inexpensive familiar item needs a correct issue, honest completeness statement, condition description, lawful venue and several real market comparisons. Proportionate research means increasing cost, formality and specialist input as financial, legal, authenticity and reputational consequences increase.

When expert input becomes necessary

The potential value is substantial or attribution materially changes value.
Authenticity, restoration or signature status is central to the sale.
The item lies outside the seller's competence or known counterfeits are sophisticated.
Legal restrictions, cultural sensitivity, title or provenance may be disputed.
The market normally expects recognised grading, certification or a formal opinion.

Give the specialist clear photographs, exact dimensions, marks, numbers, known history, condition information, references already checked and a precise question. A verbal opinion should not later be represented as written authentication, and third-party certification remains evidence rather than immunity from future challenge.

Keep a compact research dossier

Overall and detail photographs, including marks, damage, repairs and associated parts
Dimensions, weight, materials and a transcription of identifiers
Source list with author, title, edition, page number, link and access date where relevant
Saved catalogue records, screenshots or archive references where permitted
Comparable-sales record showing why each result is or is not comparable
Copies of provenance, certificates, invoices and expert correspondence
Legal, permit, export and platform-policy checks
Conclusions, confidence levels, negative findings and unresolved questions
The final wording used in the listing

The dossier makes the work reproducible. It supports buyer questions, corrections, insurance, disputes, relisting and later transfer of knowledge. Online pages and databases change, so record access dates and preserve sufficient detail to relocate the evidence without depending on memory.

Common research failures

Copying the first identification

Search visibility is not evidential quality. A chain of listings may repeat one unsupported description.

Using active listings as value evidence

An asking price records a seller's expectation. It does not show that a buyer accepted the price.

Comparing the wrong issue

A visually similar object can differ in production date, packaging, contents, authenticity, scarcity and value.

Treating repetition as corroboration

Four pages using identical wording may all derive from the same earlier listing rather than four independent sources.

Researching only for value

Price research built on uncertain identity, authenticity or completeness is unstable from the outset.

Using uncertainty as sales theatre

Phrases such as ‘possibly a rare prototype’ can plant a valuable claim while avoiding responsibility for proving it.

Sale-stop conditions

The item is not ready merely because a listing can be written

The object's identity remains materially uncertain.
Authenticity cannot be supported to a reasonable standard.
Ownership is disputed, consigned without authority or potentially stolen.
A permit, registration, exemption or export check may be required.
An essential component, signature, certificate or piece of packaging appears questionable.
Research reveals restoration, alteration or replacement parts that have not been assessed or disclosed.
The proposed listing depends on a claim that cannot be evidenced.
The intended platform prohibits or restricts the item.

The appropriate next step may be further documentation, specialist examination, legal advice, a more cautious description, a different venue or withdrawal from sale.

The final pre-listing research test

Identity

What exactly is it, who made it, which edition or variant is it, and which features establish that conclusion?

Authenticity

What supports genuineness, which reproductions or counterfeits are known, and is specialist authentication warranted?

Completeness

What was supplied with this exact issue, and are any present parts later, replaced, reproduced or from another version?

History

Which provenance is documented, which history is merely reported, and are there ownership concerns?

Condition

Has research distinguished normal production features from damage, restoration, alteration or substitution?

Market

Which actual sales are comparable, and have condition, venue, date, fees and sell-through been considered?

Legality

Can the item be advertised, sold, shipped and exported through the chosen venue, and are documents required?

Description

Can every important claim be supported, and are uncertainty and material adverse information expressed clearly?

Key takeaways

Begin with the object and a neutral record, not the most attractive online identification.
Separate observed facts, supported conclusions and unverified reports.
Identify the exact edition, production state, variant and individual specimen.
Use primary and specialist sources before relying on marketplace repetition.
Test valuable conclusions by actively looking for evidence that would disprove them.
Compare realised sales only after normalising identity, condition, completeness, venue and fees.
Make listing language no stronger than the evidence supporting it.
Pause the sale when identity, authenticity, ownership, legality or material disclosure remains unresolved.

Continue learning

Completeness and Included Components

Confirm what was originally supplied and whether the present components belong together.

→

Back to Preparing Items for Sale

Return to the full preparation sequence and its related topic pages.

→

Cleaning, Stabilising and Restoration Decisions

Decide what intervention is justified before sale and what must remain untouched or disclosed.

→