Kagi keeps getting filed under “independent search” in privacy roundups. The finer print is more interesting. Coverage summarised by SearchEngines.Net notes Kagibot feeding Teclis (small web) and TinyGem (news), while typical result pages still lean on anonymised third-party commercial indexes — historically including Google, Yandex, Mojeek, and Brave. That hybrid is the product, not a scandal.
Kagi’s differentiator is ranking control: raise, lower, block domains permanently. Independence lives at the ranking layer more than at full web-scale crawl parity. For a paid, ad-free audience, that may be the correct trade — users buy a steering wheel, not a new library of every URL on earth. Compare with Brave (own global index + API) or Mojeek (smaller own crawl, less third-party blending): Kagi optimises for user agency over index purity narratives.
For SEOs chasing Kagi visibility, obsess less over “does it crawl me?” and more over whether power users have boosted or blocked your domain class. Subscription search audiences behave differently from ad-funded SERPs: they are fewer, fussier, and more willing to maintain personal ranking rules. A domain blocked by hundreds of Kagi users is a reputation problem your Google rank will not disclose.
Practical angles: ensure Kagibot can fetch your small-web and news content if you care about Teclis/TinyGem paths; avoid dark-pattern UX that power users hate-block; and do not sell “Kagi SEO retainers” that are really generic content marketing with a privacy sticker.
Caveats: third-party blend means some Kagi rankings still shadow larger indexes; user lenses fragment what “rank” means; and paid-search audiences are tiny versus Google. Treat Kagi as a reputational and niche-visibility channel, instrument what you can, and keep the hybrid index story honest in client decks.
Community and brand teams should listen for Kagi-specific complaints the same way they watch Reddit. A wave of blocks is a product-trust signal. Fix the underlying UX or claims issue; do not try to “SEO” your way around a subscription audience that paid specifically to silence you.
If you publish on the small web — independent blogs, niche docs, newsletters — Kagibot’s Teclis path may be one of the few Western crawls still interested. Make fetching easy. That is niche, and niche is allowed.
When pitching Kagi work, lead with ranking-control UX and hybrid crawl truth; clients who wanted purity myths will self-select out.
Competitive analysis tip: search your category on a stock Kagi account and on an account with heavy personal lenses if you can ethically access one. The delta teaches you how much “Kagi ranking” is really user configuration. That humility belongs in any proposal that prices Kagi optimisation as if it were Google position tracking.
Keep Kagibot in robots allow lists for sites that want small-web discovery; blocking it while asking for Teclis presence is self-sabotage.