Etsy search reads your title, tags, categories and attributes together. The strongest titles put the most specific buyer phrase first, avoid repeating the same word, and read like something a human would say rather than a keyword list.
Front-load the specific phrase
Buyers do not search "necklace". They search "gold initial necklace for bridesmaid". Long, specific phrases carry less competition and far higher intent, and the first words of your title carry the most weight.
Write the exact phrase a buyer would type, put it first, then add distinguishing detail. Run the result through the Listing Optimizer, which flags length problems, repeated words and weak keyword placement.
Stop repeating yourself
Using the same word five times does not help you rank five times. It wastes character space that could carry a different phrase and it makes the title unreadable, which costs you the click even when you rank.
Use all thirteen tags, and make them different
Tags should extend your reach, not echo the title. If your title already says "gold initial necklace", a tag saying "gold necklace" adds nothing. Use tags for the phrasings you could not fit: occasion, recipient, style, material, and the descriptive terms buyers use that you would not.
Multi-word tags generally outperform single words, for the same reason specific titles do.
Attributes and categories are ranking signals
Sellers fill these in carelessly and then wonder why a well-written listing underperforms. Etsy uses them to filter search, so an incomplete attribute set removes you from results you would have won.
What matters beyond keywords
Etsy rewards listings that convert. Photography, price, review count and shipping speed all feed back into where you appear. A perfectly optimised title on a listing with weak photos will not hold its position. Check that your images meet each platform's expectations with the Image Resizer.
One compliance note
If you used generative AI on your images or descriptions, Etsy now requires disclosure and listings without it can be filtered from search. That looks exactly like an SEO problem while being a policy one. See the AI disclosure rules before you blame your keywords.