No one can honestly promise when AEO will produce citations, and no independent benchmark has been published. What is documented runs on three clocks. Access changes are fast: OpenAI says ChatGPT search adjusts to a robots.txt change in about 24 hours, and Google says crawling takes “a few days to a few weeks.” First citations, if a page earns them, often arrive within days or weeks: in a Semrush test of 81 new pages, Google AI Mode cited 29 the next day and ChatGPT search had cited 34 by day 30, though many were never cited. And citations churn: Ahrefs found 50% to 83% of them gone the next day, depending on the engine. So expect days to weeks for access fixes to register, and two or three monthly measurement cycles before you can say your visibility is really rising.

This page is about timing. For what to fix on each engine, see How to Get Cited by ChatGPT, Google AI Overviews and AI Mode, Perplexity and Claude; for the method, How to Measure AI Citations.

StageWhat’s documentedSource
Crawlers see an access changeChatGPT search: about 24 hours after a robots.txt update. Google: crawling takes a few days to a few weeksOpenAI; Google Search Central
A new page gets its first citationOften within days to a few weeks (Profound: median 6.81 days, 90% within 37); in Semrush’s 30-day test, many pages neverSemrush test (81 pages); Profound summary
A citation holds50–83% gone the next day; about 75% of dropped citations return within 20 daysAhrefs (37,500 citations, five tools)
You can call a trendAfter a change holds across two or three measurement cyclesSparkToro; Ahrefs; our measurement guide

Why there’s no single number

“How long does AEO take?” bundles three different things: mechanical delays the engines document (how fast a crawler sees a change), probabilistic choices that vary from run to run (whether an engine picks your page for an answer), and work with no fixed duration (earning coverage on the review sites and publications engines cite). Agency and vendor guides fill the gap with week-and-month ranges, but none we’ve found says how its ranges were measured (see below). This page sticks to what the engines document and what published tests measured.

Stage 1: Getting crawled and indexed

An engine has to fetch your page before it can cite it. This is the one stage with published latencies, and the one where you control the start date.

EngineWhat has to reach your pageWhat the company says about timing
ChatGPT searchOAI-SearchBot. Sites that opt out “will not be shown in ChatGPT search answers""It can take ~24 hours from a site’s robots.txt update for our systems to adjust” (OpenAI)
Google AI Overviews and AI ModeGooglebot. A page must be indexed and eligible to show with a snippet; there are “no additional technical requirements""Crawling can take anywhere from a few days to a few weeks” (Google). For preview-control changes in AI features, “anywhere from several days to several months” (Google)
Microsoft Copilot and Bing’s AI summariesBing’s crawler and indexNo timeline. Bing recommends IndexNow so AI answers use the current version of a page, but IndexNow says submitting a URL “does not guarantee immediate indexing”
PerplexityPerplexityBot and Perplexity’s own index of over 200 billion URLsNo timeline. A machine-learning model decides whether a URL needs indexing and when, based on its importance and how often it’s likely to change (Perplexity, September 2025)
ClaudeClaude-SearchBot, which indexes pages for Claude’s search resultsNo timeline published (Anthropic)

Fix access first. If robots.txt, a CDN or bot protection blocks a search crawler, no content work will earn citations from that engine; the AI Citation Readiness Checklist covers what to allow. You can ask, but not hurry: URL Inspection and sitemaps request a Google recrawl, but Google says requesting the same URL repeatedly “won’t get it crawled any faster,” and IndexNow reaches Bing and its other participants (Amazon, Naver, Seznam.cz, Yandex and Yep), not Google. Removal is faster than inclusion: Google says excluding a site with Search Console’s generative AI setting “generally takes a few days,” with most content gone within 1–2 days (Search Console Help). No engine publishes a clock for getting in, because inclusion depends on being chosen, not just crawled.

Stage 2: How fast new pages earn a first citation

Semrush’s controlled test (published December 10, 2025). Semrush published 81 FAQ-style test pages on its own blog in September 2025 and tracked related queries daily for 30 days in its own AI tracking tool.

Google AI ModeChatGPT search
Cited the day after publishing29 pages (36%)8 pages
Day 745 pages (56%)14 pages (17%)
Day 1439 pages (48%)28 pages (35%)
Day 3021 pages (26%)34 pages (42%)

AI Mode picked pages up fast, then dropped more than half; ChatGPT was slower but kept what it cited. At their peaks the engines cited 59% and 42% of the pages, so many never became sources. Semrush’s blog has an Authority Score of 84, and Semrush says its fast AI Mode pickup “could be because” of that strength, so a newer site shouldn’t expect the same curve.

Profound’s time-to-citation figures (May 20, 2026). Profound, which sells AI visibility tracking, lists on its research hub a median of 6.81 days from publishing to citation in ChatGPT or Claude, with 75% of pages cited within 18.68 days and 90% within 37.10 days. In the LinkedIn post behind it, Profound’s Josh Blyskal says the figures come from about 900 newly published marketing pages pulled from Profound’s logs; he doesn’t say whether pages that were never cited are counted. That matters: if only cited pages are included, these numbers describe how fast winners get picked up, not how likely a new page is to be cited at all.

Both point the same way: if an engine is going to cite a page, it often does so within days to a few weeks. A crawlable, relevant page that has gone months without a first citation is less a reason for patience than a sign that something else is missing.

Stage 3: Citations come and go

Ahrefs’ 20-day stability study (published October 29, 2026). Ahrefs asked five AI tools the same questions every day for 20 days and followed roughly 37,500 cited websites.

AI toolQuestions trackedGone by day twoLasted all 20 days
Google AI Overviews56350%6.6%
Google AI Mode1,99658%6.8%
ChatGPT3,17359%4.0%
Gemini1,27681%0.2%
Copilot82383%0.5%

Most losses were temporary: about 75% of citations that disappeared came back within the 20 days, half of them within three days. Gemini and Copilot look worse partly because 64% and 53% of their answers cited nobody. Brand-new citations were the shakiest (in two Ahrefs tracking reports, around 70% vanished the next day and only 34–45% of those returned), while citations that had lasted eight to 14 days almost always came back.

AI Overviews rewrite themselves often. In an earlier Ahrefs study (November 11, 2025) of 43,000+ keywords tracked for a month, AI Overview content changed every 2.15 days on average and 45.5% of cited sources were new from one observation to the next, while the meaning held steady (cosine similarity 0.95).

Most citations are short bursts. Profound followed 883,000 pages across seven engines from September 2025 to September 2026 (September 30, 2026): the median page’s share of citations in its category fell to half its peak within 11 days of peaking (42 days from first citation). Profound averages shares over 14 days, so a page cited in a single burst can’t have a half-life longer than 14 days; the 11 days describes bursts, not a fixed shelf life. A page’s lifespan on one engine also said almost nothing about its lifespan on another.

So a citation that appears this week is a data point, not a result. Ahrefs suggests counting how many of the last seven days you were cited for each question, and not crediting or blaming a page edit for a single appearance or disappearance.

How long before you can tell whether it’s working

How soon you can know depends mostly on how you measure. In SparkToro’s study (published January 28, 2026), 600 volunteers ran 12 recommendation prompts through ChatGPT, Claude and Google’s AI Overviews or AI Mode 2,961 times in November and December 2025. There was less than a 1 in 100 chance that ChatGPT or Google’s AI returned the same list of brands twice. Rand Fishkin concluded that how often a brand appears across many runs is a reasonable metric and its list position is not (his co-researcher works for an AI tracking company, and the study isn’t peer-reviewed). Three consequences:

  • One check proves nothing. Run each prompt several times on each engine and report appearance rate (the share of answers that name you) and citation rate (the share that link to your site), not a rank.
  • Small samples hide real changes. At a 40% appearance rate, 150 sampled answers carry a margin of error of roughly ±8 percentage points, so a move from 38% to 44% isn’t yet a trend. The worked math is in How to Measure AI Citations.
  • Trends take cycles. Call a change only when it holds across two or three measurement cycles. Measured monthly, that’s two to three months after your baseline, whatever the engines did in between.

AI referral traffic needs the same care: it can swing sharply month to month for reasons unrelated to your work, as The ROI of GEO shows.

What published timelines and case studies claim

Agency and vendor ranges. First Answer, an AEO agency for local businesses, says (February 23, 2025) to expect “initial measurable improvements in 4-8 weeks,” visibility gains in 3–6 months and “dominant AI authority” in competitive markets in 6–12 months, adding that the milestones “aren’t guaranteed — they’re typical.” Fogtrail, an AEO platform vendor, says (February 17, 2026) that B2B SaaS companies see first citations on Perplexity and Grok within 2–4 weeks, on Gemini and Claude within 4–8 weeks, and on ChatGPT for competitive queries in 2–4 months. Neither publishes the data or method behind its ranges, and First Answer’s quick wins still include adding FAQ schema, a rich result Google stopped showing on May 7, 2026 and has since removed from its documentation. Semrush’s measured test, where ChatGPT cited some new pages the day after publishing, is a reminder that any such range depends heavily on the site and the query.

Agency case studies. The public examples are agencies reporting on their own websites. The Rank Masters (September 12, 2025) reported 675 ChatGPT-referred page views from June 11 to September 8, 2025, against 8 in the previous 90 days, after relaunching its site; that’s the 8,337% in its headline, from a very small base. Go Fish Digital (September 24, 2025) reported a 43% rise in monthly AI-driven traffic and an 83.33% rise in monthly conversions from AI referrals over three months. Both are self-reported marketing with no control group. They show that movement within a quarter is possible, not that it’s typical.

We haven’t found an independent study that follows AEO programs against a control. Until one exists, treat any fixed timeline, including one quoted on a sales call, as a guess.

What changes how fast you’ll see movement

Crawl access. A blocked crawler stops the clock completely, and it’s the fastest thing to fix.

Whether you rank in Google, for Google’s AI features. In Ahrefs’ stability study, pages that ranked anywhere in Google for the same question were about seven times as likely to stay cited for a week or more: 29% against 4% in AI Overviews and 28% against 4% in AI Mode. The gap was smaller in ChatGPT (11% against 4%), and the ranking analysis covered only 230 questions. Domain Rating and backlinks showed little connection beyond the basics.

The kind of question. When a question named a brand, that brand’s own site stayed cited for all 20 days in 28–33% of cases on Google’s AI tools and ChatGPT, four to seven times the rate for the average cited site. Across question types, 64–76% of cited pages changed from one day to the next on ChatGPT and AI Mode, and “best X” questions were among the least stable, at about 75%. Expect steadier results on questions about your own products and pricing than on category recommendations.

Who the engines cite for comparisons. In a smaller Ahrefs sample of 22 questions about its own brand, Google’s AI tools almost never cited Ahrefs’ site for comparison questions (on 0% of days in AI Overviews and 5% in AI Mode for the typical question), while ChatGPT cited it on 71% of days. Where engines lean on third-party reviews and publications, your timeline depends on earning coverage there. Competitor Analysis for AI Citations shows how to find those sources.

A measurement-led first 90 days

Ninety days allows for crawling, a first round of content work and two re-measurements. Treat it as a diagnostic checkpoint, not a deadline for results.

Days 1–7: baseline. Lock a list of 20–50 prompts your buyers actually ask. Run each one several times on every engine that matters to you, such as Google AI Overviews and AI Mode, ChatGPT, Perplexity, Gemini, Copilot and Claude. Log whether you’re named, whether you’re linked, and which sources appear instead.

Days 1–14: access. Check robots.txt, your CDN and bot protection for each search crawler. Verify the site in Google Search Console and Bing Webmaster Tools, confirm key pages are indexed with URL Inspection, and turn on IndexNow if your CMS or CDN supports it. Then watch your server logs for those crawlers’ visits.

Days 15–60: content and sources. Put the answer first on the pages that should answer your priority prompts, with specific, current facts. Keep pricing, product and policy pages accurate, since brand questions are where your own site is most likely to be cited. Start outreach to the third-party sites cited where you’re missing. Change a page’s date only when its content changes (AEO Content Refresh Playbook).

Days 30, 60 and 90: re-measure the same way. Same prompts, same engines, same number of runs. Call something a trend only if it holds across at least two cycles, and use the gaps to decide what to fix next.

Early signals before citations show up

Before your appearance rate moves, check that the groundwork is in place: search crawlers (OAI-SearchBot, Claude-SearchBot, PerplexityBot, Bingbot) fetching key pages in your server logs; pages confirmed as indexed in Search Console’s URL Inspection; impressions in Search Console’s generative AI performance report for AI Overviews and AI Mode; citations in Bing Webmaster Tools’ AI Performance report, which covers Copilot, Bing’s AI summaries and select partners; and visits in GA4’s AI Assistant channel. Setup for each is in How to Measure AI Citations.


Want a baseline before anyone quotes you a timeline? Our free AI Visibility Snapshot runs 10 prompts on ChatGPT, Google AI Overviews and Perplexity, with a short findings note. The Citation Audit tests 30–50 locked prompts across ChatGPT, Google AI Overviews, Perplexity, Gemini, Claude and the OpenAI Responses API, maps the sources cited instead of you, and delivers schema and content fixes with a prioritized 60–90 day roadmap in 5 business days. It costs $1,500, fully credited toward an Implementation Sprint or retainer signed within 30 days. We don’t guarantee citations or rankings, and we won’t promise a date; we re-run the same locked prompts the same way so you can judge the trend yourself.

Request a free Snapshot or a Citation Audit →


FAQ

How long does AEO take to work?

There’s no reliable benchmark. Access fixes register in days to weeks (about 24 hours for ChatGPT search after a robots.txt change, a few days to a few weeks for Google crawling), and new pages that get cited are often cited within days to a few weeks. Because answers vary between runs, expect two or three monthly measurement cycles before you can say your visibility has really changed.

How fast can a new page be cited by ChatGPT or Google AI Mode?

In Semrush’s December 2025 test of 81 new pages on its own high-authority blog, Google AI Mode cited 29 the day after publishing and ChatGPT search cited 8. By day 30, ChatGPT had cited 34, while AI Mode was down to 21 after peaking at 48. Many pages were never cited, and Semrush attributes its speed partly to its domain strength, so a newer site may see slower pickup.

Why did my AI citation disappear?

Probably because AI answers change constantly. In Ahrefs’ October 2026 study, 50% to 83% of citations were gone the next day depending on the tool, but about 75% of dropped citations came back within 20 days. Count how many of the last seven days you were cited before reacting, and look at the page only if a citation stays missing for two or three weeks.

Is AEO faster than SEO?

The access part can be. ChatGPT search adjusts to robots.txt changes in about a day, and new pages on strong domains have been cited the day after publishing. But AI citations churn daily, earning coverage on the third-party sites engines cite for comparisons is slow work, and in Google’s AI features, pages that already rank keep their citations much longer. AEO vs SEO vs GEO covers how the two overlap.

What should I do if I see no AI citations after 90 days?

Check access first: robots.txt, CDN and bot protection, crawler visits in your logs, and indexing in Search Console and Bing Webmaster Tools. Then check fit: does a page directly answer each prompt you track, and which sources appear instead of you? If competitors win through review sites or publications, that’s where the work is.


Sources:

All sources checked November 10, 2026.

Related reading: How to Measure AI Citations, The ROI of GEO, The AI Citation Readiness Checklist, What is Generative Engine Optimization?, AEO vs SEO vs GEO.