Nobody remembers marks. People who identify them quickly do so because they know which reference covers which field and how each one is organised. Building that small library, physical and digital, is what turns mark research from guesswork into a reliable process.
What should you own in print?
A silver hallmark guide covering all the assay offices with full date letter tables, and a general ceramic mark encyclopedia. Those two cover the majority of what most people encounter.
Both are worth having physically rather than online, because you need to compare shield shapes and device outlines side by side while holding the object. A pocket sized hallmark guide is genuinely useful at fairs and viewings.
Why are printed mark books still better for some things?
Because they are organised by shape rather than by name. When you cannot read a mark, you cannot search for it, but you can flip through plates of similar devices until something matches.
Encyclopedias arrange marks by type: crowns, anchors, animals, monograms. That visual browsing is exactly what an unidentified mark needs and exactly what a text search box cannot provide.
Which online sources are reliable?
Museum collection databases first, because every object has been examined in person by a specialist and attributed properly. They are the closest thing to authoritative and they include high resolution images.
Assay office resources for silver, and collector society archives for individual factories and fields. Auction house archives are useful for comparables though their attributions vary in quality. Treat general image search results as suggestions rather than answers.
Why be cautious with image matching?
Because a visual match only tells you two marks look alike, which is a far weaker claim than that they are the same mark. Small factories copied the devices of larger ones constantly.
Use image results to generate candidates, then verify against a proper reference: did that factory use that method of application, in that colour, in that period, on that kind of object? An unverified match is a hypothesis rather than an identification.
What specialist references are worth adding?
Depends entirely on what you handle. If you buy clocks, a clockmaker directory. If you buy pewter, a touchmark register. If you buy books, a bibliography for your authors.
Buy specialist references after you know your field rather than before, since a shelf of books on categories you never encounter is wasted money. Second hand copies are cheap, and older editions remain useful for marks even where prices in them are obsolete.
What about trade directories?
Enormously useful and mostly not something to own. Local studies libraries and county record offices hold the fullest runs, and many are digitised and searchable.
A name and address from a mark leads to a business, a trade and a run of years through directories, which is frequently the most precise dating available. Learn where your local runs are held rather than trying to acquire them.
How should you organise your own findings?
Keep a file of marks you have researched, with photographs, the conclusion, and crucially the source for each attribution. A conclusion without a source is an opinion you will not be able to defend later.
Note the marks that defeated you as well, since attributions get established as archives are digitised and a mark that leads nowhere today may be traceable in a few years. Revisit unsolved marks periodically.
What is the most useful reference of all?
Other people. Collector societies for individual fields hold detailed unpublished knowledge, and members are generally generous with it. Saleroom viewings let you handle genuine marked material free of charge.
Handling a hundred marked pieces in a category teaches more than any book, because you learn what the body, the glaze, the weight and the workmanship should feel like. The books then confirm what the object has already told you.
What should you avoid buying?
Price guides that are more than a few years old, since figures date fast even where the mark information remains sound. Buy old editions for their marks and ignore their valuations entirely.
Avoid single volume guides claiming to cover every category, which are necessarily too thin to be useful in any of them.