SEO built around product and collection pages,
not generic rankings.

How do you do keyword research for an ecommerce store?

Ecommerce keyword research starts from your catalogue, not a keyword tool. Map each collection to the category terms buyers use, each product to specific model, material or use-case searches, and guides to research-stage questions. Then check what Google already ranks for each term, because that shows whether it wants a collection, a product or an article.

• Pull the queries you already get impressions for from Search Console
• Group terms by intent: category, specific product, comparison and question
• Check the live results for each group before choosing the page type
• Give each keyword group one main page so your own pages don't compete
• Include the regional and Hinglish terms Indian shoppers actually type
• Recheck rankings and queries every quarter, since search terms shift with trends and seasons

The process I follow:

What makes ecommerce SEO go wrong?

Ecommerce SEO usually goes wrong in three ways: letting every filter and sort combination get indexed, writing blog posts while collection pages stay thin, and redesigning or migrating without a redirect map. Each one quietly costs rankings, and a migration without redirects can drop organic revenue almost overnight.

Filter indexing spreads Google's attention across thousands of near-identical URLs, so the pages you care about get crawled less. Thin collection pages are the most common gap I see, because collections usually target the biggest commercial searches and often carry nothing but a product grid. A short intro, buying notes and links to related collections give Google something to rank.

Migrations are where the most damage happens, and it's avoidable with some planning before the new site goes live. Every old URL that has traffic or links needs a mapped destination before launch, and the first weeks after go-live need daily checks on crawl errors and indexing.