indexnow works — for everyone but google

No. Google is not an IndexNow participant. The protocol's official roster lists seven engines — Bing, Yandex, Seznam, Naver, Yep, Internet Archive, Amazonbot — and Google is not on it. Our own logs agree: every IndexNow ping we ever sent came back HTTP 200 from the participating network, and Google Search Console shows zero rows tracing to any ping. IndexNow still earns its keep — it speeds Bing-family indexing (Bing, DuckDuckGo, Yahoo, Ecosia and Qwant all read that one index). For Google, the doors that work are the sitemap, internal links, and Google Search Console itself.

Measured on our own ping receipts, referral logs and the live IndexNow roster — last tested 2026-11-09. Every number below is a receipt we can point at: a roster fetch, a status code, a referral row. Nothing quoted from a forum thread.

This piece is the map for ground we already walked: why is google not indexing my site — the 199-page walk back covered the crawl-side diagnosis. This one covers the ping side — the part everyone assumes reaches Google. It does not, and the proof is short.

who actually reads an indexnow ping?

One fetch answers it. The protocol's participant list is public — no signup page to guess at, just the roster the engines themselves maintain:

curl -s https://www.indexnow.org/searchengines.json

verify: run it — seven engines come back, google is absent from every field · last tested 2026-11-09

The roster, sorted live:

every ping we sent came back 200

Since late October, each batch of new, changed or removed pages on this site went out as an IndexNow ping — including one batch of dead URLs we wanted gone from the Bing index fast. Nine pings, every single one answered HTTP 200:

10-25  435 urls  200
10-30  237 urls  200
10-31  241 urls  200
11-02   24 urls  200
11-03  274 urls  200   (new pages)
11-03  506 urls  200   (dead-url purge — 410s, deindex fast)
11-04  277 urls  200
11-05  278 urls  200
11-09    6 urls  200   (today's six)

total  2,278 urls accepted · 9 pings · 0 failures

A 200 does not promise a crawl — it promises acceptance by the ping network and a place in the queue. The queue belongs to the seven engines above. Google is not in that line at all: there is no address to send it, and there never was one. For completeness, the whole wire format fits on a napkin:

POST https://api.indexnow.org/indexnow
{
  "host": "example.com",
  "key": "your-indexnow-key",
  "urlList": ["https://example.com/new-page"]
}

verify: one POST carries up to 10,000 urls; the key file lives at your domain, engines fetch it once and trust the pings after

so — does indexnow work with google?

No — and the receipts say it three ways. One: Google is absent from the official roster; the front door of the protocol lists who signed up, and Google never did. Two: our Google Search Console export for the weeks those pings ran shows zero rows — no fetch that traces back to a ping. Three: the referral log. Thirty search sessions reached this site through search engines in the recent window. Here is who sent them:

Every frontend on that chart — DuckDuckGo, its no-AI mirror, Bing itself, Yahoo, cn.bing, Qwant — reads the same underlying Bing index. The pings feed that index and the readers follow within hours. Google's position on IndexNow has always been our position on the gym: fully aware it exists, no plans to join.

One more receipt for scale, because it makes the same point from the crawler side: over the same week, PetalBot — the crawler of a Huawei-backed search engine — hit this site 1,340 times while Googlebot managed 688. Engines nobody thinks about crawl more than the one everyone optimizes for, and the ping network had nothing to do with either.

what actually gets a page into google

IndexNow cannot carry a page to Google, so the working doors are the old ones — boring, and they still decide everything:

• sitemap.xml submitted in Google Search Console — and re-submitted after a big batch of new pages
• the page is linked from at least one page Google already crawls — internal links are the real ping
• robots.txt allows Googlebot, and the URL returns a plain 200 — a 410 or a noindex kills it
• URL Inspection ran once on the page in GSC — it both tests the render and asks for a crawl
• patience budget set: Google indexes on its own clock — days to weeks, not seconds like the Bing family

the door nobody can close

The uncomfortable number in this story is not Google's zero — it is how many readers ride one index. DuckDuckGo, Yahoo, Ecosia and Qwant are frontends; Bing is the engine underneath. Twenty-nine of our last thirty search sessions came through it. IndexNow makes that pipeline faster, not wider: if Bing ever hiccups, most of our search traffic hiccups with it.

The roster is a scoreboard anyone can read, and that is the quiet comfort in all this. If Google ever joins IndexNow, it will simply appear on that list — no press release required. We keep pinging, we keep the receipts, and the roster will tell us first. Until then the honest answer stays: indexnow works — for everyone but google.