Index bloat
Index bloat is when a search engine has indexed lots of your site's low-value pages, like filter results or empty tag pages, that nobody needs to find.
Index bloat is when a search engine has indexed far more of your site's pages than are worth finding. The extra pages are usually made by your site's software: every filter and sort order in your shop, internal search results, tag pages with one post, printer versions, old test pages.
It's a problem because those pages compete with your good ones. Crawlers spend time on them, and they can make your site look thinner and more repetitive than it is.
How to spot it
Compare how many pages Google has indexed with how many you'd actually want people to land on. Search Console's Page indexing report shows what's indexed. If you sell 200 products and Google has 9,000 of your pages, something is making pages you didn't mean to publish.
For example, Kestrel Roasters' shop lets you filter coffee by origin, roast and price, and sort by five orders. Each combination has its own address, so 40 coffees turn into thousands of pages that all show the same beans in a different order.
How to fix it
Work out which pages should exist in search, then deal with the rest. Add a noindex tag to pages people shouldn't land on from search, point near-duplicates at the main version with a canonical tag, stop crawlers from reaching endless filter combinations, and delete pages that have no reason to exist.
Index bloat and AI bots
AI bots crawl your site too, and they have the same problem with endless near-identical pages. Time a bot spends on your 3,000th filter page is time it isn't spending on your buying guide. And when pages repeat each other, it's less clear which one should be read and cited.
In AnswerOrbit
Connect your server logs and the Bot visits tab in Website shows which pages the AI bots read on your site. If most of their visits go to filter pages and old tags, that's a sign of bloat. See Website and AI traffic.