How Long Does SEO Take? Milestone Tests and Payback Math

For a campaign, expect roughly 4 to 12 months before the benefit shows up in revenue. In a February 2017 video titled "How to hire an SEO," Google's Maile Ohye said that SEOs "need four months to a year to help your business first implement improvements and then see potential benefit" (video). That is a remark from nine years ago about hiring, not a current guarantee.

Google's current documentation points the same way without a number. Its SEO Starter Guide says some changes might "take effect in a few hours, others could take several months", and advises waiting "a few weeks" before judging whether work helped.

Our recommendation: plan for 6 months before expecting meaningful non-brand leads (leads from searches that do not contain your company or product name), and treat 12 months as the point at which you hold the program to account. Six months, because indexing can take up to a few weeks and most pages that reach the top 10 take longer than one month, so leads follow the early stages rather than accompany them.

About 1.74% of random new URLs reached the top 10 within a year in Ahrefs' study, and 6.11% of English pages with real content, because older pages already hold most top positions. Top 10 means the first page of Google's results.

The two figures come from different samples in the Ahrefs study (May 2025):

• 1.74% of 1 million random URLs seen by Ahrefs' crawler in September 2023.
• 6.11% of 2 million random URLs created in October 2023, counting only non-empty English content, which Ahrefs calls the fairer number for real content.

Here is the worked example. It is our arithmetic on Ahrefs' published percentages, not a result we measured:

• Start with 1,000 newly published pages.
• 1.74% of 1,000 is 17.4, so about 17 reach the top 10 within a year.
• 6.11% of 1,000 is 61.1, so about 61.
• Ahrefs reports that 40.82% of pages that reached the top 10 did so within one month. Applying that share to both groups is our assumption: 40.82% of 17.4 is 7.1 and of 61.1 is 24.9, so about 7 to 25 do it in month one.
• That leaves roughly 939 to 983 of the 1,000 outside the top 10 after a year.

Two readings follow, and both are our inference. Most winners take longer than a month, so a quiet month one is normal. And a campaign that reaches the top 10 depends on volume and on pages you already have, not on one launch.

The age of the winners supports the second reading. Ahrefs found, across the top 10 results for 1.3 million US keywords, that 72.9% of top-10 pages are more than three years old, and the average page in position one is five years old. A new page has to displace an established one. Our recommendation is to start with narrower queries, where fewer pages compete, and treat broad terms as year-two work.

Check four stages in order: indexed, impressions, clicks, leads. Each stage has a different clock, and a failure at an early stage explains silence at a later one.

Two terms first. Google Search Console is Google's free tool that reports your pages' indexing status and search performance. An impression is one appearance of your page in a search result, whether or not anyone clicked. Position is where the page sits in the results, with 1 at the top.

The pass lines below are our recommended thresholds, not Google rules. Scale them to your site's size. They build on Google's timing: crawling can take "anywhere from a few days to a few weeks".

Are my target pages indexed at month 1?

Pass: every target page shows as on Google in the URL Inspection tool, a Search Console feature that reports whether one page is in Google's index (the stored copy of the web that results come from). Fail: any target page is still missing after four weeks. Fix this before anything else.

Check three causes. A noindex tag is a page instruction that tells Google not to list the page. A robots.txt file is a site file that tells crawlers which pages to skip. And thin content, meaning too little original substance for Google to bother keeping. Our free technical and on-page tools include a robots.txt tester and a sitemap validator, with no sign-up.

Are impressions growing at month 3?

Pass: impressions on your indexed target pages are at least double month 1's, on at least 3 pages (or on all of them, if you have fewer). Find this in Search Console, Performance report, filtered by page. Fail: flat or zero. The page either does not match what searchers want, or the query is too competitive. Retarget a narrower query rather than waiting.

Are queries reaching positions 11 to 30 by month 6?

Pass: your target pages show at least 5 queries at positions 11 to 30, and more than at month 3. Positions 11 to 30 are pages two and three of Google's results, the zone where improvement turns into first-page rankings. Fail: the count is flat. Reassess topics, internal links and page quality.

Are clicks growing at month 6?

Pass: clicks from Search Console are at least double month 3's. Fail: impressions rose but clicks did not, so the page is being seen and skipped. Rewrite the page title and description shown in results.

Is SEO producing leads at month 12?

Pass: at least 1 non-brand lead in the last 30 days, attributed to an organic landing page that is not your home page. Fail: none. If every earlier test passed and leads are still zero, the keywords, the offer or the provider are wrong. This is the point to reset or stop.

Why can traffic rise while leads stay at zero?

Because visits and leads are different stages, and traffic can come from the wrong queries. Run two tests. First, in your analytics, list your top 5 organic landing pages by sessions and check conversions: pass if at least one produced a lead in the last 30 days, fail if none did.

Second, compare only Google organic sessions in your analytics with Search Console clicks for the same dates. A gap above about 30% is our rule of thumb for tracking or attribution being wrong. Ask any provider for read access to Search Console so these numbers can be checked.

Waiting costs the monthly fee multiplied by the months until benefit. Our payback model has three inputs (monthly fee, months, profit per customer) and two formulas:

• Spend before benefit = monthly fee × months.
• Break-even customers = spend before benefit ÷ profit per customer.

For the fee, Ahrefs' survey of 439 SEO providers (updated August 15, 2024) found an average of $2,917 a month, with agencies at $3,209 and freelancers at $1,348.63. That survey is two years old as of October 2026, so check current quotes.

Worked example, using the $2,917 average and the 4 to 12 month range used on this page:

• 4 months: $2,917 × 4 = $11,668.
• 12 months: $2,917 × 12 = $35,004.
• With an illustrative profit of $400 per customer (replace it with your own): $11,668 ÷ $400 = 29.17, so 30 customers; $35,004 ÷ $400 = 87.51, so 88 customers.
• To convert customers into leads, divide by your close rate. At an illustrative 20%, 30 customers need 150 leads and 88 need 440.

This is our model, not an official figure. If the keywords you are targeting cannot plausibly produce those lead counts, change the plan before spending, not after month 8.

Keep going while early stages pass, change course when one stage stalls, and stop only when every stage has been tested and nothing has moved. The tests above give the rules:

• Pages not indexed at month 1, or impressions not doubling by month 3: a technical or targeting fault. Diagnose; do not abandon.
• Impressions rising and clicks flat at month 6: a title and positioning problem. Fix it.
• Everything passing through month 6 but no non-brand lead by month 12: reset the keywords and offer, or the provider.

To shorten the wait, fix indexing first, target narrower queries, and improve pages that already earn impressions before writing new ones. Our content refresh triage table shows how to pick which existing pages to update first.

Yes. AI answers run on a separate clock and need a separate test. There is no fixed timeline for AI answers. Check by running your target prompts weekly after publishing and recording whether your page is cited. AI citation tracking over time explains how to keep that log consistent.

Generative engine optimization (GEO) is the practice of earning mentions and citations in AI-generated answers; our comparison of GEO and SEO covers where the work differs. To see where you stand today, run an AI visibility audit.

We track whether answer engines mention or cite your brand. As of October 2026, our Launch plan is $299 a month and tracks ChatGPT, Google AI Mode and Perplexity; Gemini is included from our Growth plan at $699 a month. All plans are on our pricing page. To see how this would work for your site, Book a demo.