Research brief: how long does it actually take a new website to move through Google's pipeline — a methodology-graded benchmark report (June 2026)

TL;DR

• Crawled → indexed typically lands somewhere between a day and a few weeks (Google's own range — see Realistic indexing timing — Google's own stated range is "several hours to several weeks"; Mueller suspects "most good content is picked up and indexed within about a week") with independent large-sample data putting the median in the 8–30 day band (Benchmark #3 (Single-source): IndexCheckr Feb 2025 — 14% indexed in 7 days, ~65% in 30 days, 85% by 90 days, 93.2% by 6 months, average 27.4 days (16M pages)) — but roughly 16% of even valuable, indexable pages are never indexed at all (Onely (Tomek Rudzki) — "on average, 83% of pages are indexed within the first week of publication; some pages have to wait up to eight weeks"; ~16% of valuable, indexable pages on popular sites NEVER get indexed — flag survivorship), and that failure tail is the single most important number in the whole brief.

• Index → first meaningful ranking (top 10) is brutally slow: only 1.74% of newly published pages crack the top 10 within a year (Benchmark #2 (Industry-consensus): Ahrefs May 2025 — only 1.74% of new pages reach top 10 within a year (6.11% non-empty English); 40.82% of those that rank top 10 do so within month 1; 2017 comparison was 5.7%); of the small minority that do, ~41% get there inside the first month. Stable ranking is rarer still — Semrush's 28,000-new-domain study found fewer than 5% held a first-page position for a full year (Benchmark #5 (Single-source): Semrush 2022 (28,000 new domains, 13 months) — 41% reached top 10 by month 6; 19% reached and HELD; 27% of top-10 reachers stayed all year; <5% held first page for a year; 92% failed to stay in top 100; 4.2% held a top-10 keyword all 13 months).

• Every figure here carries survivorship bias, and almost every large study is product-incentivized (Ahrefs, Semrush, IndexCheckr, Onely, Indexing Insight all sell tools whose value proposition is "indexing/ranking is hard, buy our help"). The right planning answer is a distribution with a fat failure tail, not a point estimate — see Rule: do not give clients precise "time to rank" numbers — there are none Google has confirmed; use Mueller's "couple of months to half a year+" as the only defensible anchor and Rule (R1): plan by distribution, not average — assume ~15–20% chance any given valuable page is NEVER indexed; assume reaching top 10 within year 1 is <10% per page.

What's new in this brief vs the sister lifecycle brief

The sister brief Research brief: the lifecycle of a website in Google Search — from launch to mature standing and the perpetual re-evaluation that follows (June 2026) covers the mechanism (how Google's pipeline works end-to-end). This brief covers the numbers and their bias flags — what the published distributions actually say, who measured them, and how to read them honestly.

New material here, not in the lifecycle brief:

• The 14-row methodology-graded benchmark table, decomposed into atomic entries each carrying its rigor grade and bias flags (Benchmark #1 (Verified): Google docs — "indexing isn't guaranteed"; "several days minimum… three days or more" to crawl/index for most sites (Search Central docs, last updated Dec 2025) through Benchmark #14 (Directional-Speculative): practitioner consensus 2024–2026 — new sites take 3–6 months (9–12 in YMYL) for meaningful traction; new page on an aged domain ranks in days–weeks).
• IndexCheckr 16M-page Feb 2025 study — the 27.4-day average, the 7/30/90/180-day distribution, and the full-snapshot 37.08% / 61.94% / 0.98% breakdown (IndexCheckr Feb 2025 16M-page snapshot — context, methodology disclosure, and how to read the two numbers (timing distribution vs full-snapshot) without conflating them).
• IndexCheckr submission-tool test — 33,930 previously-unindexed pages, 29.37% indexed via tools, 70.63% remained unindexed (Benchmark #11 (Single-source): IndexCheckr submission-tool test 2025 — 33,930 previously-unindexed pages submitted to indexing tools, 29.37% indexed / 70.63% remained unindexed).
• Indexing Insight March 2025 — 1.7M-page study, index-coverage scores by segment (97% news / <90% e-commerce / ~70% marketplace), 88% of not-indexed pages quality-driven (Indexing Insight March 2025 — index-coverage by segment (97% news / <90% e-commerce / ~70% marketplace); 88% of not-indexed pages quality-driven; 1.7M pages, 18 sites).
• Indexing Insight May 2025 de-indexing purge — ~25% of 2M monitored pages removed, individual sites lost 15–75% (Indexing Insight May 2025 de-indexing purge — ~25% of 2M monitored pages removed (highest-ever); sites lost 15–75%; broke the "130-day rule" of thumb).
• Cloudflare July 2025 "From Googlebot to GPTBot" — Googlebot up 96% May 2024→May 2025, peaked April 2025 at 145% above May 2024, >25% of verified bot traffic, ~4.5% of all HTML requests; same crawl pipeline serves search AND AI Overviews/AI Mode (Cloudflare July 2025 ("From Googlebot to GPTBot") — Googlebot up 96% May 2024–May 2025; peaked April 2025 at 145% above May 2024; >25% of verified bot traffic; ~4.5% of all HTML requests in 2025, Cloudflare July 2025 — Googlebot now serves BOTH search indexing AND AI Overviews/AI Mode from the same crawl pipeline; there is no separate "AI crawler" or "AI index" on the Google side).
• Semrush 28,000-domain study with its severe selection-bias flag (Benchmark #5 (Single-source): Semrush 2022 (28,000 new domains, 13 months) — 41% reached top 10 by month 6; 19% reached and HELD; 27% of top-10 reachers stayed all year; <5% held first page for a year; 92% failed to stay in top 100; 4.2% held a top-10 keyword all 13 months).
• Moz classic 76-link backlink-to-rank study — ~10 weeks per rank (Benchmark #13 (Directional-Speculative): Moz classic — single backlink takes ~10 weeks on average to move a page up one rank (76 links, very small sample, dated)).
• Search Engine Roundtable: December 2025 core update did not settle cleanly — 9 volatility waves in 7 weeks (Search Engine Roundtable: December 2025 core update did NOT settle cleanly — 9 separate volatility waves documented in 7 weeks (Dec 2025–Feb 2026)); SE Ranking 100,000-keyword analysis: ~15% of top-10 pages vanished from top-100 (SE Ranking 100,000-keyword analysis (post-Dec 2025 core update) — ~15% of pages previously in the top 10 vanished from the top 100 entirely).
• Google March 2024 update "45% less low-quality, unoriginal content" stat (Google March 2024 core update — Google's own statement: "45% less low-quality, unoriginal content in search results vs the 40% improvement we expected").
• 2025 GSC reporting glitches — Oct–Dec 2025 Performance report lag 50–70+ hours, Page Indexing froze ~Nov 21 to Dec 18 2025 (2025 GSC reporting glitches — Oct–Dec 2025 Performance report lagged 50–70+ hours; Page Indexing report froze ~Nov 21 → Dec 18 2025; Google confirmed reporting-only (didn't affect actual crawl/index/rank)).
• Indexing API + JobPosting-schema manual-action catalyst — Splitt and Mueller quotes (Indexing API restricted to JobPosting/BroadcastEvent + putting JobPosting schema on non-job pages is a manual-action catalyst — Splitt: "Pushing it to the API doesn't mean indexed right away or indexed at all"; Mueller: "Search is never guaranteed").

The honest planning frame

Stage 1 — Set expectations by distribution, not average. Indexing within days-to-weeks for most pages on a technically sound site; assume ~15–20% chance any given valuable page is never indexed; assume reaching the top 10 within the first year is <10% per page and treat any faster result as upside. See Rule (R1): plan by distribution, not average — assume ~15–20% chance any given valuable page is NEVER indexed; assume reaching top 10 within year 1 is <10% per page.

Stage 2 — Remove the controllable bottlenecks (weeks 0–8). Clean crawlability, prominent internal links, accurate sitemap. Distinguish the two "not indexed" states: Discovered – currently not indexed is a crawl-priority/quality-signal problem (A large "Discovered – currently not indexed" backlog is a SITE-WIDE quality signal, not a per-page problem — Google declines to spend crawl resources on URL patterns it predicts will be low-value); Crawled – currently not indexed is a page-level quality decision (Search Console "Crawled – currently not indexed" is a deliberate quality decision, not a queue state — Google fetched and evaluated the page and CHOSE not to index it). They are different triggers and need different fixes — see Rule (R3): "Discovered – not indexed" and "Crawled – not indexed" have DIFFERENT triggers — first is crawl-priority/site-wide quality signal, second is per-page quality decision; treat them differently.

Stage 3 — Compete on the real drivers (months 2–12). Lower-competition queries first; genuine authority/backlinks; raise content quality. Do not buy speed — submission tools indexed only 29.37% in independent testing (Benchmark #11 (Single-source): IndexCheckr submission-tool test 2025 — 33,930 previously-unindexed pages submitted to indexing tools, 29.37% indexed / 70.63% remained unindexed), and Google's Indexing API is restricted to JobPosting/BroadcastEvent (Google's Indexing API is restricted to JobPosting and BroadcastEvent content — Mueller has warned against using it for ordinary pages).

Stage 4 — Monitor with care. Threshold: if index coverage of intentional pages drops below ~85–90%, investigate quality/duplication first (Rule (R2): monitor index-coverage threshold at 85–90% — if it drops below, investigate quality/duplication BEFORE technical SEO; the Indexing Insight 88%-quality-driven finding makes this the right ordering). Do NOT spam-request indexing — duplicates are ignored within a crawl cycle (Rule (R4): do NOT spam-request indexing — duplicate Request-Indexing submissions are ignored within a crawl cycle; "a hint, not a command" (Google)). When GSC is known to be glitching (as in Oct–Dec 2025), cross-check with server logs + analytics + URL Inspection live tests (Rule (R5): cross-check GSC during known lag windows — server logs + analytics + URL Inspection live tests; do NOT make drastic site changes on stale data (Oct–Dec 2025 was the named instance)).

What is genuinely unknown

Google does not publish indexing- or ranking-time distributions; no party outside Google can measure true publish-to-index at scale (all proxies — crawler "first seen," tracking-start dates — introduce lag); the precise share of valuable pages never indexed is bounded only loosely (~16–20%); no controlled, vendor-independent, large-sample study of brand-new-domain vs established-domain timing exists. The ground-truth instrument (Google's internal logs) is not accessible. See Genuine unknowns in the Google Search pipeline — exact queue priority math, render-queue position, signal weightings, re-rendering triggers, whether/when a page will ever rank.

Source: compass_artifact research document, June 2026. Sources include Google Search Central documentation, John Mueller, Martin Splitt, Gary Illyes, Danny Sullivan, Ahrefs (Patrick Stox, 2025), Semrush (2022), IndexCheckr (2025), Onely (Tomek Rudzki), Indexing Insight (Adam Gent, 2025), Moz, Cloudflare, Search Engine Roundtable, SE Ranking.