Should I include filler words in backend keywords?
Filler words, also called stop words, are common words that carry little semantic meaning in isolation: prepositions such as 'for', 'in', 'on' and 'with'; articles such as 'the', 'a' and 'an'; conjunctions such as 'and' and 'or'. Amazon's indexing system is generally built to ignore these words in search queries because buyers rarely search for them in isolation and their inclusion does not meaningfully distinguish one product from another. Including filler words in your backend Search Terms field typically wastes bytes without improving your listing's indexation or ranking. Amazon's indexer maintains a list of stop words that are excluded from its indexation process. While Amazon does not publish this list officially, the category broadly matches the standard stop word lists used in information retrieval systems: articles ('a', 'an', 'the'), prepositions ('for', 'in', 'on', 'at', 'with', 'by', 'to', 'of', 'from'), conjunctions ('and', 'or', 'but', 'so'), and common auxiliary verbs ('is', 'are', 'was', 'be', 'have', 'has', 'do'). These words consume bytes when included in your Search Terms field, but the indexer skips over them when building the keyword index. A Search Terms field containing 'lamp for desk that is rechargeable and dimmable for home office or bedroom' is longer in bytes than 'lamp desk rechargeable dimmable home office bedroom' but provides no more indexation. The substantive words, those that describe the product in ways buyers search for, are the only words that contribute to indexation. Not every preposition or article is meaningless in every keyword context. Some phrases are searched by buyers with the preposition included because it changes the meaning or product context: 'gift for dad', 'organiser for under sink', 'hooks for shower curtain'. In these cases, the preposition is part of the natural buyer search phrase and may contribute to indexation for that specific multi-word query. The practical guidance is to include prepositions only when they are part of a complete phrase that buyers naturally search as a unit, and to exclude them when they are being used as connectors between otherwise unrelated keywords. In your Search Terms field, write keyword phrases in their natural buyer form where the preposition matters: 'organiser under sink' or 'gift for dad', but never 'lamp for home for desk for bedroom for office'. The phrase-level context is what matters, not the presence of the preposition itself. A useful exercise when reviewing your backend keywords is to go through the Search Terms field word by word and ask whether each word is substantive (a noun, adjective, verb or specific qualifier that a buyer would search for) or a filler connector. Remove every filler connector and replace it with a single space. Then use the bytes recovered to add one or two additional substantive keyword phrases at the end of the field. For a LED desk lamp, a bloated backend Search Terms field might look like: 'a lamp for your desk that is good for home and office use with dimmer and rechargeable battery for students'. A cleaned version might be: 'desk lamp rechargeable dimmable USB office bedroom study student reading night light portable'. The cleaned version is shorter in bytes but contains more unique indexable tokens and covers a wider range of search queries.
Filler words like 'for' and 'the' are largely ignored by Amazon's indexing system. Learn which words to skip to maximise your 250-byte backend keyword limit.