Amazon Search Keywords

I start with the product itself. If I am listing a stainless steel water bottle, I write down every word that describes it: material (stainless steel, BPA-free), size (32 oz, 1 litre), benefit (insulated, keeps cold 24 hours), use case (hiking, gym, office). That gives me a seed list of maybe 12 words. From there I open Amazon and type "water bottle" into the search bar. I copy every autocomplete suggestion that matches my product — "water bottle 32 oz", "insulated water bottle for hiking", "stainless steel water bottle no sweat" — and then scroll to the bottom of the search results page to grab the "Related searches" section. That gives me another 8-10 terms.

Next, I look at the top five competitor listings. I note the words they use in their titles and bullets, and — critical step — I view the page source (Ctrl+U) and search for "searchTerms" to see their backend keywords. If one competitor ranks for "vacuum insulated flask" and I do not have it, I add it to my list. I then throw all of these terms into a flat list and run a Sponsored Products campaign on broad match for a week.

After seven days I pull the [Search Term Report](/keyword-research/) from Amazon Advertising. I sort by clicks and sales, and highlight the terms that drove at least one sale. Those go straight into my title and first two bullet points. The rest — synonyms, misspellings, related uses — go into the backend search terms field. For example, my Search Term Report showed "thermos" as a high-click term even though my listing said "water bottle". I added "thermos" to my backend terms and organic impressions increased 18% in two weeks.

I repeat this every month for the first three months. After that I only revisit when sales drop or I add new variants. This workflow is not glamorous, but it is reliable because it relies on real shopper language, not on what I guess they might type.

Chasing high-volume keywords that are not relevant. A term like "water bottle" has 500k monthly searches, but the first five organic slots are held by brands with 50,000+ reviews. If you target it, you are paying for clicks from shoppers who will bounce. Use [long-tail keywords](/long-tail-keywords/) such as "32 oz stainless steel insulated water bottle with straw" instead. That term might have 500 searches, but the conversion rate is three times higher because the intent is specific.

Keyword stuffing in titles and bullets. Amazon’s algorithm recognises repetition. I once saw a listing that repeated "water bottle" four times in the title. Its click-through rate was 4% lower than the category average. More importantly, customers read your title. They do not want to see "Water Bottle 32 oz Water Bottle Insulated Water Bottle". One mention of the core term is enough; use the rest of the space for benefits.

Ignoring the backend search terms field. Many sellers either leave it blank or copy their front-end keywords. That is a missed opportunity. The backend field is for keywords that do not fit naturally on the page — synonyms, misspellings, foreign language variants, related terms. For instance, if your water bottle is not branded as a thermos but shoppers search "thermos" anyway, you can place it in the backend without cluttering your title. However, do not repeat terms already in your title or bullets; Amazon ignores duplicates.

Finally, watch out for [keyword difficulty](/keyword-difficulty/) on Amazon. The platform does not show a difficulty score natively, but you can approximate it by comparing your review count and price against the top ten organic results for a term. If the top five all have 10,000+ reviews and your product has 50, skip the term until you build authority.

When I started selling on Amazon, I packed my backend search terms with every synonym, misspelling, and variant I could find — including words already in my title like "bottle" and "water". I thought more keywords meant more visibility. In reality, Amazon ignores duplicates, so I wasted 60% of the available character limit. I also added terms like "sports drink container" which had zero relevance to my insulated bottle. That did not help because Amazon matches backend keywords to queries, and irrelevant terms either waste the space or attract the wrong audience.

My second mistake was ignoring the Search Term Report for two months. I assumed my broad-match campaign would tell Amazon what I sold. Instead, I spent £1,200 on clicks from shoppers searching "cheap plastic bottle" — a product I do not sell. When I finally ran the report, I discovered that 40% of my clicks came from "vacuum insulated flask" and "double wall bottle", terms I had not included anywhere. I added them to my backend field and saw organic impressions climb 35% within ten days.

I also learned that match type matters. I ran only broad match because I wanted maximum discovery. But broad match on Amazon interprets loosely; a query for "bottle" could trigger my ad for "mug". Switching to phrase match for my top ten converting terms cut my ACoS from 38% to 28% in three weeks. I now track term performance weekly using a simple spreadsheet and [tracking keywords](/tracking-keywords/) through Amazon's campaign manager. The lesson: data beats intuition every time on Amazon.